Digital Techniques in Diagnostic of Aseptic Pleural Empyema in Patient After Cardiac Surgery

Editorial board

izet Masic (Editor-in-Chief), Mediha Zalihic (Secretary), Mirza Hamzic (technical editor), lejla Masic (lector)

Members: John bryden (Glasgow, United Kingdom), Zoran Hadziahmetovic (Sarajevo, b&H), Jacob Hofdijk (amsterdam, the Netherlands), Josipa Kern (Zagreb, Croatia), oleg Mayorov (Kiev, Ukraine), John Mantas (athens, Greece), ahmed Novo (Sarajevo, b&H), besim Prnjavorac (tesanj, b&H), Zlatko Puvacic (Sarajevo, b&H), Zoran ridjanovic (Sarajevo, b&H), drago rudel (ljubljana, Slovenia), Zekerijah Sabanovic (tuzla, b&H), armin Skrbo (Sarajevo, b&H), Selim toromanovic (Cazin, b&H)

address of the Editorial board: Sarajevo, Cekalusa 90, tel/fax 00387 33 444 714, E-mail: [email protected], [email protected]

Published by: “avicena” d.o.o., Sarajevo, Zaima Sarca 43

oriGiNal PaPErSWWW-based E-learning on demand: the learning resource Server Medicine (lrSMed) ................. 175Jürgen Stausberg, Martin Geueke

Evaluation of two Search Methods in PubMed; the regular Search and Search by MeSH terms ........... 180osman Saka, Kemal Hakan Gülkesen, birsen Gülden, oya deniz Koçgil

Medical informatics Education in bosnia and Herzegovina 184izet Masic, ahmed Novo

the Sensitivity and Specificity of Prolactin (Prl) in breast Cancer Patients .... 189Zlata Mujagic, Hamza Mujagic

analysis of β-thalassaemia by Using the Survival analysis techniques .............. 192Salahuddin Zaman, Qamruz Zaman

ProFESSioNal PaPErSontologies and information Models: Comparing the UMlS and the diCoM Content Mapping resource 198Jérémy lasbleiz, Valérie bertaud

Usage of international Nomenclatures and Metathesauruses in Shared Healthcare in the Czech republic ....................201Petra Preckova, Josef Spidlen, Jana Zvarova

information technologies and Quality of Menagement decision in University

Education .................206Mensura Kudumovic, Stevica Krsmanovic, dzafer Kudumovic

Classifications of deformities and developmental Malformations in the locomotor System 208Zoran Hadziahmetovic

rEViEWSinternet and Medicine: Yesterday – today – tomorrow .................211Gjuro dezelic

CaSE rEPortSValue of Ultrasound Examination in Case of Scrotum traumas ....................218Edin Herceglija, Fahrudin Smajlovic, Faruk dalagija

digital techniques in diagnostic of aseptic Pleural Empyema in Patient after Cardiac Surgery .....................220Sanko Pandur, Mehmed Kulic, Haris Vila, omer Perva, Safet Guska, Nermin Granov, Michele Musci, bedrudin banjanovic

off-pump bypass for left anterior descending aneurysm With Support of digital techniques ..............223Mirsad Kacila, Marco l.S. Matteucci, Stefano bevilacqua, Nermir Granov, Mattia Glauber

Endometriosis Externa treated by diagnostic laparoscopy 225Haris tanovic, ratko Juricic, Samir Muhovic

NEWS .........................228

ISSN 0353-8109

vol 13 no 4 DECEMBER 2005: 173-232

in this issue

acta informatica medicajournal of the society for medical informatics of b&h

background of dMib&H and aiM

The Society of Medical Informatics of Bosnia and Herze-govina (DMI B&H) was founded 1988, as member of for-mer Yugoslavian Association of Medical Informatics, found-ed also in 1988, and member of EFMI in 1990. The Society has now over 100 members. The Society become member of EFMI in 1994 (EFMI Council in Lisbon), and member of IMIA in 1994 (General Assembly in Dresden).

The Society carries out the following activities:• Promotion and improvement of informatics within the health-care system, health insurance and bio-medical research,• Engagement of experts in the field of medical informatics in B&H on develop-ment and establishment of health care information systems,• Assistance in research, development and professional work in the field of medi-cal informatics in B&H, • Distribution and development of technical information in the field of medical in-formatics in B&H,• Assistance in education of medical informatics experts,• Exchange of professional experience on national and international level,• Publishing activities in the field of medical informatics.

Acta Informatica Medica (AIM) is official journal of the Society for medical infor-matics of Bosnia and Herzegovina, founded in 1993. Journal discusses the basic methodology within field of medical informatics, like: fundamentals of processing data, information and data base knowledge, classification systems in medicine, health information systems, using of information technologies in health care sys-tems, etc. It publishes original papers, professional papers, reviews, case reports, news, etc.

Prof Izet Masic, MD, PhD,Editor-In-Chief

Web address: www.imasic.org/bhsmi/

acta informatica medicajournal of the society for medical informatics of b&h

Acta Informatica Medica

vol 13 no 4 DECEMBER 2005 175

oriG

iNal

PaPErS

original paperSummarydespite the lost enthusiasm concerning E-learning a lot of material is available on the World Wide Web (WWW) free of charge. this material is collected and systematically described by services like the learning resource Server Medicine (lrSMed) at http://www.lrsmed.de/. With the lrSMed, E-learning modules are made available for medical students by means of a metadata description that can be used for a catalogue search. the number of resources included has risen enormously from 100 in 1999 up to 1085 today. Es-pecially in 2004 there was an exponential increase in the lrSMed’s content. anatomy is still the field with the highest amount of available material, but gener-al medicine has improved its position over the years and is now the second one. technically and didacti-cally simple material as scripts, textbooks, and link lists (called info services) is still dominating. an eval-uation demonstrates that the retrieval accuracy of lrSMed and its competitors is quite well; four of five resources offered correspond with the search criteri-on. but, the completeness of WWW-based catalogues in this area is still not satisfying. With its application-programming interface, lrSMed could be integrat-ed in a service-oriented architecture easily. the scene of free E-learning modules on the WWW is ready to meet current challenges for efficient training of stu-dents and continuing education in medicine.Keywords: Computer assisted instruction; E-learn-ing; Evaluation; learning; Metadata; World Wide Web

1. introductionE-learning is a frequently used catchword in the IT-domain

and in the medical field. It is proposed that the market of E-learning software in Europe will grow up to 6 billion EURO until 2006 [1]. Nevertheless, E-learning has disappeared from the political agenda of strategically important topics [2]. This is due to the fact that expectations and promises have failed; un-realistic enthusiasm is followed by deep frustration.

In Germany, current regulations for medical education of-fer new chances for a revival of E-learning. On the one hand, conditions for education of medical students have been revised [3]. Education at medical faculties should be more oriented to health practice and health care system requirements. Ex-cathe-

dra teaching became less important, the role of seminars and teaching small groups of one to three students has grown. As consequence lecturers are confronted with a dramatically in-crease in hours needed for education, which have been gained from time used for research and patient treatment. On the oth-er hand, continuing medical education became mandatory for physicians in outpatient care [4]. Within 5 years, practitioners have to collect 250 points in certified courses to keep their li-cense and to avoid financial restrictions.

Both requirements, new regulations for medical educa-tion and mandatory continuing medical education, create new hopes and options for E-learning. Especially material offered on the World Wide Web (WWW) supports the required ef-ficiency in education and training through following advan-tages [5].

Independence from local connections because the Inter-net allows access from all over the world with a low cost technical infrastructure. Independence from time restrictions because the material is available 24 hours 7 days a week without any technical necessity for downtime. Independence from the availability of teachers at the time of learning because their knowledge is integrated in learn-ing modules. Easy and fast updating because all content is localized on the WWW-server.

A lot of useful material for WWW-based E-learning had been developed worldwide during the last 5 to 10 years of professional adoption. These could be used for self-learning, blended learning as well as distance learning curricula. But, to find suitable E-learning modules on demand reminds at the look for a needle in a haystack. Using an ordinary fulltext search engine like Google offers only poor chances of success. WWW-based catalogues are an important mean to overcome this problem. The providers of WWW-based catalogues take on the challenge to retrieve, collect, structure, and present re-sources for the domain they have competence.

Goal of this project is to support students in finding the right offers (E-learning modules on the WWW), anytime, any-where, on demand. To reach that goal the Learning Resource Server Medicine (LRSMed) was set-up. In the following we will introduce the LRSMed in chapter 2, present trends in WWW-based E-learning seen from the LRSMed in chapter 3, and end with results from an evaluation study about the LRSMed and its competitors.

•

•

•

•

WWW-based E-learning on demand: the learning resource Server Medicine (lrSMed)Jürgen Stausberg1, Martin Geueke2

institute for Medical informatics, biometry and Epidemiology, Medical Faculty, University of duisburg-Essen, Germany1

Clinresearch GmbH, Cologne, Germany2

AIM 2005, 13(4): 175-179

176 vol 13 no 4 DECEMBER 2005

Papers

2. lrSMedThe LRSMed is a multilingual ser-

vice that retrieves E-learning modules freely available on the WWW, describes the modules with a metadata standard, stores the metadata in a database and offers a user-interface for retrieval, commenting and authoring [6]. Main target group of LRSMed are medical students, but some material will also be useful for health care professionals as well as for the public. The LRSMed is available at http://www.lrsmed.de/. The first database of multimedia learn-ing resources in the WWW was creat-ed in 1997 [7], the implementation of the LRSMed started in 2001.

Figures 1 and 2 show screen-shots from the LRSMed. In figure 1 a search for E-learning modules is specified us-ing the criteria specialism (anatomy, general), application type (image at-las), and language (English). The result screen is shown in figure 2. Twenty-one modules were retrieved from the database and information is presented including the modules’ title, specialisms, application types, and languages. The user can call further information or can switch directly to the modules via the title’s hyperlink. As part of LRSMed’s quality assurance strategy, comments about a module can be entered here. The result set can be transformed into a docu-ment in Portable Document Format (PDF) and printed out.

Key features of LRSMed include the use of Learning Objects Meta-data (LOM) [8] in the implementation of the IMS Learning Re-source Meta-data Information Model [9] as metadata-specification, the eXtensible Mark-up Language (XML) as syntax for interfac-es and the Oracle suite for implementation. An application programming interface imple-mented with the Simple Object Access Proto-col (SOAP) enables the integration of LRSMed in other applications as hospital information systems, E-learning platforms, etc.

The IMS-specification for metadata in-cludes technical as well as non-technical as-pects organized in nine branches; each branch

consists of a hierarchical structure with the data at the bottom called leaves:

1. general – information about the resource as a whole

2. lifecycle – history and current state

3. metametadata – information about the description itself

4. technical – technical information, e. g. the required browser version

5. educational – characteristics con-cerning didactic and education

6. rights – information concerning li-cense restrictions

7. relation – connections to other re-sources

8. annotation – general remarks

9. classification – description of characteristics with clas-sifications

A critical assessment of the E-learning module’s quality is supported in two ways. Firstly, an interested user can read com-ments of others about that module. The comments are stored within the metadata and comprise free text as well as a simple score (How great was your satisfaction?). Secondly, LOM had

FiGUrE 1. User interface of the search for E-learning modules.

FiGUrE 2. Result set with further options and hyperlinks to the modules.

Quality ManagementEvaluationevidencetypedescriptionlocation

processstandard

locationupload filename upload filename

person

description

contextassessment

role

LOM-Standard

date

Annotation

Extensions

General

Lifecycle

Metametadata

Technical

Classification

Educational

description

FiGUrE 3. LRSMed’s extensions of LOM.



ogy; anatomy; neurology; radiology; dermatology; pathology. Medico-theoretical fields as biochemistry, biometry, and epi-demiology lost their position in the TOP 10-list. In 2004, num-ber 2 behind anatomy is general medicine, which improved its position from 8 in 2002. Firstly visible in the TOP 10 in 2004 is pediatrics at position 8. Histology (position 4), first aid (posi-tion 8), and otolaryngology (position 10) had an intermediate visit in the TOP 10-list 2002. Between 2002 and 2004 internal medicine had been divided into its sub-specializations due to a high number of respective modules. The most frequent com-binations of specialism and application type in 2004 are shown in table 1.tablE 1. Combinations of specialism and application type with 10 or more modules in 2004.Specialism application type Number of modules

anatomy image database 18

General medicine info service 17

Microbiology Script 16

Virology Script 16

General medicine Script 15

Cardiology textbook 13

Cardiology Script 13

radiology Case study 13

biochemistry Script 12

Hematology and oncology Script 12

anatomy Simulation 11

Gastroenterology Script 11

Pharmacology Script 11

Hematology and oncology info service 10

4. EvaluationWe compared the quality of retrieval by LRSMed with three

competitors. Computer aid learning reviews (CAL reviews) - http://axis.cbcu.cam.ac.uk/calreviews: CAL reviews is a project at the University of Cambridge, UK. Start of the proj-ect was September 1997. It is described as an ‘Internet-based catalogue of expert reviewed medical multimedia learning re-sources for use in the Cambridge clinical course’. Comment-ed database for E-learning: Medicine (KELDAmed) - http://www.ma.uni-heidelberg.de/bibl/KELDAmed/: KELDAmed is a service offered by the University Hospital of Mannheim

and the Heidelberg University, Germa-ny. It includes ‘interesting E-learning links for doctors, medical students and nursing staff ’ from the WWW. Learning modules in the Internet (LMU) - http://link.medinn.med.uni-muenchen.de/instruct/de/casus/wwwcbt.html: The working group for computer based med-ical education at the Munich downtown hospitals, Germany, maintains a HTML-page that lists E-learning modules in the Internet for medical students.

A detailed presentation of the evalu-ation study is given in [11]. We calculat-ed completeness (recall) as the propor-tion between the retrieved resources that are correct and all correct resources, and accuracy (precision) as the proportion between the retrieved resources that are correct and all retrieved resources.

It was decided to calculate complete-ness pair-wise and to pool a represen-

been extended with entities covering information about the module’s development process (Was the development based on specific standards for software development?) as well as infor-mation about evaluation (Has the module demonstrated its ed-ucational impact in a controlled study?) (cf. figure 3). Possible values for evidence are the following: I systematic review, II randomized study, III experimental study, IV not experimental study, V expert assessment. ISO 9000 is an example of a stan-dard in quality management.

Learning material has to fulfill several requirements to be accepted for inclusion in the LRSMed: availability (free of charge), technique (standard WWW-browser, only extend-ed by common plug-ins), target group (primarily medical stu-dents), language (at the moment German and English), and application type.

3. trends in frequency, specialism and application type

The number of E-learning modules available in LRSMed has increased dramatically. Having 100 modules in the first quarter of 1999, this number raises up to 267 until the second quarter in 2002 and exponential up to 439 the first quarter, 653 the second quarter, 763 the fourth quarter 2004 and 805 the first quarter 2005. The LRSMed offers currently 1085 active re-sources (status 2005-09-16). Further trends could be compared with the years 1999 [7], 2002 [6] and 2004 [10].

Figure 4 shows an overview of the relative frequency of modules categorized according to the application type. Because LRSMed allows multiple classifications of E-learning modules to different application types, some modules count twice in fig-ure 4. Similar to 1999, simple scripts are most commonly of-fered in 2004 (30 %). The type info service jumped from a bad position of 2 % in 1999 to position 3 with 12 %. We define info services as enhanced link lists concerning a medical topic. Other application types that received a higher relative frequen-cy in 2004 than in 1999 are textbook, drill and practice, and virtual presentation. On the opposite, the relevancy of simula-tion, image atlas, presentation, and audio database decreased. Video database and questionnaire were introduced later.

The six most frequent specialties in 1999 are also present in the TOP 10 of 2004: internal medicine divided in its special-ties cardiology, hematology and oncology, and gastroenterol-

0%

5%

10%

15%

20%

25%

30%

35%

Video D

ataba

se

Questi

onna

ireScri

pt

Simula

tion

Imag

e Atla

s

Case St

udy

Imag

e Data

base

Textbo

ok

Presen

tation

Info S

ervice

Audio

Databa

se

Drill an

d prac

tice

Virtual

Presen

tation

Dialog

FiGUrE 4. Relative frequency of E-learning modules 1999 (left column), 2002 (middle column) and fourth quarter in 2004 (right column) categorized according to the application type. From left to right the application types are ordered in descending relative frequency from 1999.


Papers

tative set of E-learning modules. Twenty re-sources were selected randomly from each of the four services on March 3rd, 2004. The availability of the 80 E-learning modules was checked in the following ten days with-out knowledge of the service, from which a module had been selected. For the calcula-tion of accuracy, it was decided to use criteria types, which are present in all three servic-es, as medical field, type of learning resource, and free-text search of diagnoses. Sampling was done between 7th and 15th June 2004. Gynecology and microbiology were select-ed randomly with 183 resources. The type ‘case study’ was chosen for the study and half of the learning resources randomly selected (106 from 211). As free-text diagnoses diabe-tes mellitus and AIDS were selected arbitrari-ly. Both diagnoses receive hits in all services leading to 93 enclosed modules. The accordance between cri-teria and E-learning modules was checked in the end of June 2004. In total 382 assessments were necessary. Information for the reviewers includes the URL, the title of the E-learning module within its source, and the criterion.

CompletenessSix hosts were not reachable for the assessment, 74 mod-

ules with 296 assessments remain for the calculation of com-pleteness, which yields to an availability of 93%. From the 74 remaining modules, 33 could be identified in CAL reviews (45.0%), 45 in KELDAmed (60.8%), 26 in LMU (35.1%), and 42 in LRSMed (56.8%). Table 2 shows the pair-wise compari-son between the services. The best mean completeness for for-eign E-learning modules shows KELDAmed with 46.7% as well as the best overall completeness with 60.8%. Sixteen E-learning resources could be retrieved in all German services, seven in all services.

AccuracyThe following results rely on 344 modules (127 from CAL

reviews, 116 from KELDAmed, 101 from LRSMed). Excluded were 33 modules as not available (8.6%). Another five mod-ules required identification with a password and were exclud-ed as well (1.5%).

Figure 5 shows a comparison of the services’ accuracy. The best values were reached for diabetes mellitus in CAL reviews and gynecology in LRSMed with 100.0%, the worst for AIDS with 34.8% in LRSMed and microbiology with 46.2% in KEL-DAmed closely followed by diabetes mellitus in LRSMed with 50.0%. CAL reviews wins in four of five criteria. Overall, 264 from 344 modules were rated as correct (76.7%). From 127 modules provided by CAL reviews 105 were rated as correct (82.7%), from 116 provided by KELDAmed 89 (76.6%), and from 101 provided by LRSMed 70 (69.3%). The highest accura-cy was achieved for case study with 88.3%, followed by medical

field with 72.6% and free-text diagnoses with 72.1%.

5. ConclusionsOn the WWW, the number of freely available E-learning

modules in medicine has increased enormously between 1999 and 2004. Some remarkable national and supranational fund-ing programs in the last decade might be one reason for this in-crease in learning material. For example, the German govern-ment funded 180 projects within its program “New Media for Education” between 2000 and 2003 with more than 180 million EURO, from which 16 projects in medicine received 34 million EURO [12]. In addition, the large number of scripts might in-dicate that many lecturers offer material they have developed for own purposes before. Meanwhile, the coverage of special-ties has changed. Image oriented specialties like anatomy, radi-ology, and pathology dominated in the late 90s as well as medi-co-theoretical fields like biochemistry. Nowadays clinical fields with a high common relevancy are present in the top-list of E-learning modules. Especially general medicine made its way into the WWW. General surgery is still underrepresented in comparison to its importance.

The quality of the E-learning modules is difficult to access. Published studies demonstrate a poor reliability in assessing

health teaching resources [13]. An impres-sion could be received by the trend in appli-cation types. Not the number of complex and sophisticating types had been raised - simple scripts are still the most frequent application type available in 2004. As mentioned in 1999 by Haag et al. [14], modules offering a tuto-rial dialog are still missing. So the WWW is used mainly for a distribution of classic ma-terial. The possibilities of the medium are still not fully exploited.

The presented results show that the LRSMed’s completeness is still not sufficiently on the one hand. On the other hand, the user can be sure that the resources re-trieved by means of this catalogue met his or her requirements concerning medical field, type of learning resource, and free-text diagnoses.

From overall 462 E-learning modules in this study, 39 could not be accessed with the provided URL (8.4% broken links). The LRSMed presents the best maintenance with only 5.4% broken links in comparison to 130 retrieved E-learning mod-ules. For the evaluated services, some reasons of shortcomings can be identified. An important reason for the incompleteness

availability Cal reviews KEldamed lMU lrSMed

Source N % N % N % N % N % p-value

Cal reviews 20 100.0 20 100.0 5 25.0 2 10.0 2 10.0 0.474

KEldamed 19 100.0 2 10.5 19 100.0 2 10.5 9 47.4 0.012*

lMU 15 100.0 6 40.0 6 40.0 15 100.0 11 73.3 0.108

lrSMed 20 100.0 5 25.0 15 75.0 7 35.0 20 100.0 0.003*

tablE 2. Absolute and relative frequency of the pair-wise evaluation of completeness. The source was excluded in the calculation of the p-value. Significant values with asterisk.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

100%

AIDS Diabetes mellitus

Case study Gynaecology Microbiology

(p=0,001)* (p=0,023)* (p=0,613) (p=0,053) (p=0,012)*

FiGUrE 5. Accuracy of the services according to the retrieval criterion (CAL reviews: left, black column; KELDAmed: middle, white column; LRSMed: right, blank column). P-values in parenthesis, significant values with asterisk.



of CAL reviews is the primary language. CAL reviews includes solely English modules as far as the authors know. Thus, the availability of German modules retrieved from the other ser-vices could not be expected. From the German services, KEL-DAmed achieved the best completeness with 60.8%. It includes also the highest number of resources. The low number of mod-ules from KELDAmed detected in the other services could be explained by many PDF-files, which are not accepted as E-learning modules by the competitors. The accuracy of the cat-alogues is quite better than their completeness. That mirrors the strengths and weaknesses of catalogues in comparison to search engines in general. A careful maintenance should lead to appropriate assignments of modules to rubrics. The weak re-sults in free-text search with LRSMed could be explained by an automatic indexing function. The LRSMed reads the homep-ages of every E-learning module regularly, scans the text, and builds-up a new index with every relevant word as an entry (without articles for example).

We do not know whether the situation seen in our LRSMed is representative for the scene. However, our analysis demon-strates that a lot of material is available which could be inte-grated into courses because of its simple structure. Today, lec-turers can utilize this material and combine it with face-to-face teaching to establish an ideal synergy of blended learning.

AcknowledgementsThe development of LRSMed was funded by the German

Federal Ministry of Education and Research (BMBF) within the project Vision 2003.

REfEREnCEs1. EU-Kommission – Vertretung in der Bundesrepublik

Deutschland, EU-Nachrichten. Themenheft Nr. 1. eEu-rope, 2002: 18.

2. The European ODL Liaison Committee. Distance Learn-ing and eLearning in European policy and practice: the vi-sion and the reality. 17 November 2004. URL: http://www.odl-liaison.org/ (accessed 2004-11-26).

3. Approbationsordnung für Ärzte vom 27. Juni 2002. Bundesgesetzblatt Teil 1, 2002: 2405-35.

4. Gesetz zur Modernisierung der gesetzlichen Kranken-versicherung (GKV-Modernisierungsgesetz - GMG). Bundesgesetzblatt Teil 1, 2003: 2190-258.

5. Lowe HJ, Lomax EC, Polonkey SE. The World Wide Web: a review of an emerging Internet-based technology for the distribution of biomedical information. JAMIA, 1996; 3: 1-14.

6. Geueke M, Stausberg J. A meta-data based Learning Re-source Server for Medicine. Computer Methods and Pro-grams in Biomedicine, 2003; 72: 197-208.

7. Stausberg J, Bündgens D, Prange J. Multimediale med-izinische Lehr- und Lernsoftware im World Wide Web: Eine Angebotsanalyse. In Victor N, Blettner M, Edler L, Haux R, Knaup-Gregori P, Pritsch M, Wahrendorf J, Win-deler J, Ziegler S, eds. Medical informatics, biostatistics and epidemiology for efficient health care and medical re-search. München: Urban & Vogel, 1999: 333-6.

8. IEEE Learning Technology Standards Committee (LTSC), Learning Object Model 6.1. URL: http://ltsc.ieee.org/doc/wg12/LOM_WD6-1_1_without_tracking.pdf (2002-05-30).

9. IMS Global Learning Consortium, IMS Learning Re-source Meta-data Information Model Version 1.2. Final Specification. 17 May 2001. URL: http://www.imsglobal.org/metadata/index.html (accessed 2002-06-26).

10. Stausberg J, Geueke M, Bludßat K. Trends in free WWW-based E-learning modules seen from the Learning Re-source Server Medicine (LRSMed). In: Engelbrecht R, Geissbühler A, Lovis C, Mihalas G, eds. Connecting Med-ical Informatics and Bio-Informatics. Proceedings of MIE2005. Amsterdam: IOS, 2005: 290-5.

11. Stausberg J, Bludßat K, Geueke M. Completeness and ac-curacy of WWW-based catalogues of medical E-learning modules. Medical Informatics and the Internet in Medi-cine (in press).

12. Projektträger Neue Medien in der Bildung + Fachinfor-mation des BMBF in der Fraunhofer Gesellschaft. Förder-programm Neue Medien in der Bildung. Förderbereich Hochschule. Aktuelle Fördervorhaben aus der Förder-bekanntmachung zum Einsatz Neuer Medien in der Hochschullehre. Sankt Augustin Dezember 2002. URL: http://www.gmd.de/PT-NMB/Projektdokus/Hochschul_Vorhaben.pdf (accessed 2003-05-05).

13. Darmoni SJ, le Duff F, Joubert M, le Beux P, Fieschi M, We-ber J, Benichou J. A preliminary study to assess a French code of ethics for health teaching resources on the Inter-net. In: Surján G, Engelbrecht R, McNair P, eds. Health data in the information society. Proceedings MIE2002. Amsterdam: IOS, 2002: 621-6.

14. Haag M, Maylein L, Leven FJ, Tönshoff B, Haux R. Web-based training: a new paradigm in computer-assisted in-struction in medicine. International Journal of Medical Informatics 1999; 53: 79-90.

Correspondence to: Priv.-doz. dr. med. Jürgen Stausberg, institute for Medical informatics, biometry and Epidemiology, Medical Fac-ulty, University of duisburg-Essen, Hufelandstr. 55, d-45122 Essen, Germany, tel.: +49 201 723 4512, E-Mail: [email protected]


Papers

original paperSummaryPubMed is the most frequently used medical in-dexing database by health and biological scienc-es professionals all over the world. MeSH (Medical Subject Headings) is NlM’s (National library of Med-icine) controlled vocabulary used for indexing arti-cles for PubMed. the users usually perform a reg-ular PubMed search, but they can also directly use MeSH terms in their search. Searching by MeSH terms may be more efficient than the regular PubMed search. the aim of this study is to compare the regu-lar PubMed search with search by MeSH terms. We searched five subjects using the two methods. the results of each search were evaluated by both quali-tative and quantitative methods. results belonging to each subject have been evaluated by five experts to give a satisfaction score and a relevancy score for each method. Mean satisfaction scores that are giv-en by experts were, 3.24 for the regular PubMed search, and 3.72 for the search by MeSH terms. Mean relevancy scores were 3.36 and 4.14 for the regu-lar PubMed and the search by MeSH terms respec-tively. the quantitative analysis showed that sensitiv-ity was 0.60 in the PubMed search, 0.50 by the MeSH search, whereas precision was 0.70 versus 0.68. the difficulty of designing such a study is clear, and fur-ther studies in this domain needs to be performed. a comparison by qualitative and quantitative analy-ses showed no dramatic difference for the two search methods. these results may be more easily explained for PubMed searches, but one of the main aims of the MeSH system is to prevent the complications of a routine PubMed search. our expectation was high-er sensitivity and precision as a result of the search by MeSH terms. the results suggest that, there is a problem with indexing procedures. authors must find appropriate keywords for their article, try to use stan-dard terms and design their abstracts carefully. re-viewing the indexing process in NlM to see if there is a problem may also be helpful. We think that the quality of MeSH indexing must be improved to pro-vide better functionality.Keywords: PubMed; Medical Subject Headings; in-formation Storage and retrieval.

1. introductionModern medical knowledge mainly grows with the help of

scientific research. The number of research papers increases every year and it is impossible to follow them without an in-dexing database. A few decades ago, the most widely known medical index was paper based Index Medicus. Computeriza-tion of this database has resulted in Medline, which has dra-matically decreased the time needed for searching. The real revolution was PubMed, which is the web counterpart of Med-line. PubMed evolves with additional features, and it is free and reachable from every access point on the Internet all over the world. Today, PubMed is the most frequently used medical in-dexing database by health and biological sciences profession-als.

MeSH (Medical Subject Headings) is NLM’s (National Li-brary of Medicine) controlled vocabulary used for indexing ar-ticles for PubMed. In other words, each article in the PubMed database is indexed according to the MeSH terms. The MeSH terminology provides a consistent way to retrieve informa-tion that may use different terminology for the same concepts. When the user performs a search in PubMed, synonyms and the MeSH headings are automatically added to original search terms, and the user is usually not aware of this process.

A user can also directly use the MeSH terms in his/her search. This type of search is a bit more complicated and most of the searchers do not know about searching directly by the MeSH terms. However, searching by the MeSH terms may be more efficient than the regular PubMed search.

The aim of this study is to compare the regular PubMed search with the search by MeSH terms. We tried to get infor-mation about the advantages and disadvantages of each search type, to guide researchers about their search strategy.

2. MethodsWe randomly selected ten researchers from our medical fac-

ulty. They were asked the subject of the last research that they have done using the PubMed, and the exact method and phras-es used in their search. As a result, we obtained ten subjects to search. By the help of the original search terms that were used by researchers, two different authors separately determined the optimum search terms for the subject for the search by the MeSH terms. If there was an inconsistency between the select-ed terms, a consensus was reached by the help of a third author. Five of the ten searches were eliminated because of a confusion in search terms, or very similar search subjects. If the regular PubMed search or the search by MeSH terms produced over 60 results, the number of results was reduced to less than 60 by us-

Evaluation of two Search Methods in PubMed; the regular Search and Search by MeSH termsosman Saka1, Kemal Hakan Gülkesen1, birsen Gülden2 oya deniz Koçgil2

akdeniz University, antalya, turkey1

Middle East technical University, ankara, turkey2

AIM 2005, 13(4): 180-183



ing a time limit. The reason of this operation was to obtain a reason-able number of titles for qualitative evaluation.

2.1. Quantitative evaluationEach article title in the list was

classified as relevant or non-rele-vant, based on the question of “If you would want to see the studies related to the subject, is this arti-cle may be accepted as relevant to the subject?” The classifications were made by three authors/ex-perts for each article and the ma-jority of opinions were accepted. Some additional searches with dif-ferent phrases were performed to find more titles that were missed by our searches. The results of all of the searches were examined, and the number of total relevant arti-cles was determined.

By these processes, we obtained some measures about features of each search subject (1); Number of total relevant articles (2). Num-ber of articles which were not in-dexed (the articles cannot be found by the MeSH search because index-ing procedure needs time after the article is entered to database, so some new articles are unin-dexed yet) (3). Number of total articles which were obtained by the regular PubMed search (4). Number of relevant arti-cles which were obtained by the regular PubMed search (5). Number of missed articles by the regular PubMed search (6). Number of total articles which were obtained by the search by MeSH terms (7). Number of relevant articles which were ob-tained by the search by MeSH terms. (8) Number of missed ar-ticles by the search by MeSH terms. As a result, some indexes, namely sensitivity and precision were obtained from these fig-ures (Table 1).tablE 1. Definition of “sensitivity” and “precision”.

PubMed MeSH

Sensitivity

Number of relevant articles which were obtained by the regular PubMed search/Number of total relevant articles

Number of relevant articles which were obtained by the search by MeSH terms/(Number of total relevant articles-number of articles that are not indexed)

Precision

number of relevant articles by the regular PubMed search/number of total articles obtained by the regular PubMed search

number of relevant articles by the search by MeSH terms/number of total articles by the search by MeSH terms

2.2. Qualitative evaluationThe results of the regular PubMed and MeSH term search-

es for each subject were printed. Results belonging to each sub-ject have been evaluated by five experts. The experts were not the same five people for every subject, and a total of 12 ex-perts evaluated the results. They were given a questionnaire to give a satisfaction score and a relevancy score of one to five for both the regular PubMed search and the MeSH terms search

results. 2.3. Statistical analysisThe reliability coefficients were calculated by an established

method [1] for qualitative scores of each subject. If the reliabil-ity was under 0.7, the expert with lowest part-whole correla-tion was excluded from the study, and an additional expert was asked to evaluate the results.

3. resultsDetails of selected five couple of searches are presented in

Table 2.3.1. Quantitative analysisThe results of quantitative analysis are presented in Table 3.

Using the data in this table, some indexes were produced (Ta-ble 4).tablE 3. The results of quantitative analysis.

Subject No

total number

Not indexed

regular PubMed search Search by MeSH terms

N of found

N of related titles

N of missed titles

N of found

N of related titles

N of missed titles

1 68 14 33 31 37 54 43 11

2 8 2 8 8 0 2 2 4

3 10 - 35 10 0 25 9 1

4 35 0 8 7 28 12 12 23

5 21 0 17 7 14 17 4 17

3.2. Qualitative analysisThe mean satisfaction scores that were given by experts

were, 3.24 for the regular PubMed search, and 3.72 for the search by the MeSH terms. The mean relevancy scores were 3.36 and 4.14 for the regular PubMed and the search by the MeSH terms respectively (Table 5).

4. discussionThe number of medical articles rises each year, and by the

help of the information technologies, the researchers can find the articles which they are interested in. National Library of Medicine (USA) has a free web service that assists searching

No the Subjectregular PubMed search

Search by MeSH terms Extra search(es)original search details

1Use of laparoscopy in prostatectomy

“laparoscopic radical prostatectomy” “prostate cancer”

(laparoscopic[all Fields] aNd radical[all Fields] aNd (“prostatectomy”[MeSH terms] or prostatectomy[text Word])) aNd “prostate cancer”[all Fields] aNd (“2003”[Pdat] : “3000”[Pdat])

“Prostatic Neoplasms”[MeSH] aNd “Prostatectomy”[MeSH] aNd “laparoscopy”[MeSH] aNd (“2003”[Pdat] : “3000”[Pdat])

laparoscop* radical prostatectomy

2Effect of melatonin on arteries in coronary bypass surgery

bypass melatoninbypass[all Fields] aNd (“melatonin”[MeSH terms] or melatonin[text Word])

“Coronary artery bypass”[MeSH] aNd “Melatonin”[MeSH]

1. pass melatonin2. coronary melatonin

3the current situation in treatment of pheochromocytoma

pheochromocytoma aNd treatment

(pheochromocytoma[text Word] or phaeochromocytoma[text Word] or “pheochromocytoma”[MeSH terms]) aNd (“therapy”[Subheading] or “therapeutics”[MeSH terms] or treatment[text Word]) aNd review[ptyp] aNd (“2003”[Pdat] : “3000”[Pdat])

“Pheochromocytoma” [MeSH] aNd “therapy”[ Subheading] aNd review[ptyp] aNd (“2003” [Pdat] : “3000”[Pdat])

pheochromocytoma therapy

4do cellular telephones cause brain cancer?

“brain cancer” phone

“brain cancer”[all Fields] aNd phone[all Fields]

“brain Neoplasms”[MeSH] aNd “Cellular Phone”[MeSH]

1. phone malignancy brain2. phone brain tumour

5

telemedicine applications for diabetes mellitus patients.

telemedicine applications for diabetes

(“telemedicine”[MeSH terms] or telemedicine[text Word]) aNd applications[all Fields] aNd (“diabetes mellitus”[MeSH terms] or “diabetes insipidus”[MeSH terms] or diabetes[text Word])

“telemedicine”[MeSH] aNd “diabetes Mellitus, type ii”[MeSH]

1. home care applications for diabetes2. diabetic aNd telemedicine Not “diabetic retinopathy”

tablE 2. Details of each PubMed and MeSH search.


Papers

the medical article database for several years. This web service is known as the PubMed, and widely known by a large number of researchers all over the world. It is possibly the most fre-quently used database for medical subject search.

Scientific literature is not disinterested in the PubMed [2], and some reports advising better search methods using the PubMed are published [3, 4]. However, we are aware of only one study that systematically evaluates PubMed searches [5]. We think that it is because of the difficulty of designing a study that evaluates an application that used by a wide spec-trum of people for a wide spectrum of purposes. The design of the present study was really troublesome trying to solve a lot of confusions about the study method, and led to long discus-sions between the authors.

A statistical comparison was not performed because of the low number of subjects. A comparison of qualitative and quan-titative analyses showed no dramatic difference for the two search methods (sensitivity: 0.60 versus 0.50, precision: 0.70 versus 0.68). However, these figures are not very satisfactory for researcher. A researcher can reach about half of the arti-cles by a search, and he/she must see one non-relevant article for two relevant articles. In other words, a regular search us-ing the PubMed or a MeSH term search can find 50 of 100 ar-ticles about the subject, and the researcher must see an addi-tional 25 article to find this 50 article. In another study search by MeSH terms provided higher precision than the regular PubMed search (0.66 and 0.47, respectively) with a lower sen-sitivity (0.78 versus 0.88) [5].

These results may be more easily explained for PubMed searches. A well known problem for text based databases is the presence of some synonyms, or use of different terms for the same entity. In some cases of PubMed search, an algorithm adds synonym terms to search spontaneously. For example, in the PubMed search of subject 3 (table II), the original search term is “treatment”, and the subheading of “therapy” was au-tomatically added to the search terms. This function is very valuable for the user, and prevents missing some articles. On the other hand, this function does not work for every term. The search with the subject 4 contains “brain cancer” term, and “brain neoplasms” term is not automatically included in the search.

The PubMed search also produces some undesired titles. This is a common problem for the text based search, because

the search is not semantically structured and the use of the same word in different context can produce positive results.

In fact, one of the main aims of the MeSH system is to prevent the complications of the routine PubMed search. Our expecta-tion was higher sensitivity and precision by the MeSH search. The results of the search by the MeSH terms are not complete-ly disappointing, but the presence of a functionality problem is clear. A search by the MeSH terms has no clear advantage over the regular PubMed search according to our results. The logic behind the design of the MeSH system is clear and defensible, but the results suggest that, there is a problem with indexing procedures. In a study, simultaneous use of textword and in-dexing terms retrieved only 82/107 (77%) papers [6]. Authors do not define their study with correct terms in their abstracts, and sometimes do not give appropriate keywords [6, 7]. Au-thors need to improve the quality of abstracts to make retrieval and screening of relevant papers more effectively and efficient-ly [8]. Over one million articles enter the database each year, and because of wide area of medicine and complexity of con-temporary scientific research, indexing the articles is difficult for the team of indexers in NLM. Authors must find appropri-ate keywords for their article, try to use standard terms and de-sign their abstracts carefully. Reviewing the indexing process to see if there is a problem may also be helpful. The workload of the people who perform indexing, and the workflow of in-dexing process may need to be checked for further improve-ment. However, because of the large number of articles that must be indexed, a higher quality indexing process may need a substantial investment. The ultimate and best solution may be the development of intelligent algorithms that analyses seman-tic relation in texts.

5. ConclusionIn this study, we tried to evaluate the sensitivity and preci-

sion in the regular PubMed and the MeSH term searches. A significant difference could not be observed by both qualita-tive and quantitative methods. These results may be more eas-ily explained for PubMed searches, but one of the main aims of the MeSH system is to prevent the complications of the rou-tine PubMed search. The MeSH indexing does not seem to be very reliable, and the situation must be evaluated by further studies in this domain. We think that the quality of MeSH in-dexing must be improved to have a wider acceptance and bet-ter functionality.

AcknowledgementsThe authors wish to thank Dr. Buket Cinemre and Dr. Uğur Bilge for revising the English of this article. We also wish to thank Dr. Sadi Özdem, Dr. Cahit Nacitarhan, Dr. Coşkun

Usta, Dr. Burhan Savaş, Dr. Hakan Bozcuk, Dr. Türker Kök-sal, Dr. Mustafa Usta, Dr. Erol Güntekin, Dr. Uğur Bilge, Mr. Özgür Tosun, Dr. Özlem Batu, Dr. Neşe Zayim for their as-

sistance in qualitative evaluation. This research is supported by Akdeniz University Research Foundation.

REfEREnCEs1. Friedman CP, Wyatt JC. Computing reliability coefficents.

In: Evauation Methods in Medical Informatics. New York: Springer-Verlag, 1997: 115-7.

2. Kostoffa RN, Blockb JA, Stumpa JA, Pfeil KM. Informa-tion content in Medline record fields. Int J Med Inform, 2004; 73: 515-27.

3. Bachmann LM, Coray R, Estermann P, Ter Riet G. Iden-tifying Diagnostic Studies in MEDLINE: Reducing the Number Needed to Read. J Am Med Inform Assoc, 2002; 9: 653–8.

4. Montori VM, Wilczynski NL, Morgan D, Haynes RB. Op-

Subject No

regular PubMed search Search by MeSH terms

Sensitivity Precision Sensitivity Precision

1 0.46 0.94 0.75 0.80

2 1.00 1.00 0.33 1.00

3 1.00 0.29 0.90 0.36

4 0.20 0.88 0.34 1.00

5 0.33 0.41 0.19 0.24

Mean 0.60 0.70 0.50 0.68

tablE 4. Sensitivity and precision of each search.

Subject No

regular PubMed search Search by MeSH terms reliability

Satisfaction relevancy Satisfaction relevancy

1 3.8 4.2 4.2 4.0 0.74

2 4.2 4.0 2.8 5.0 0.79

3 2.2 2.2 3.4 3.4 0.77

4 3.0 3.8 3.6 4.2 0.74

5 3.0 2.6 4.6 4.2 0.79

Mean 3.24 3.36 3.72 4.16 -

tablE 5- The mean scores given by experts on satisfaction and relevancy of the searches.



timal search strategies for retrieving systematic reviews from Medline: analytical survey. BMJ, 2005; 330: 68.

5. Jenuwine ES, Floyd JA. Comparison of Medical Subject Headings and text-word searches in MEDLINE to retrieve studies on sleep in healthy individuals. J Med Libr Assoc, 2004; 92: 349-53.

6. Derry S, Kong Loke Y, Aronson JK. Incomplete evidence: the inadequacy of databases in tracing published adverse drug reactions in clinical trials. BMC Med Res Methodol, 2001; 1: 7.

7. Murphy LS, Reinsch S, Najm WI, Dickerson VM, Seffin-ger MA, Adams A, Mishra SI. Searching biomedical da-tabases on complementary medicine: the use of control-led vocabulary among authors, indexers and investigators. BMC Complement Altern Med, 2003; 3: 3.

8. Dijkers MP. Searching the literature for information on traumatic spinal cord injury: the usefulness of abstracts. Spinal Cord, 2003; 41: 76-84.

Correspondence to: Prof dr osman Saka, akdeniz Universitesi tip Fakultesi, biyoistatistik ad, antalya, turkey. [email protected]

MEDICAL INFORMATICS INUNITED AND HEALTHIER EUROPE

LOCATION: Holiday Inn, Sarajevo,

Bosnia and HerzegovinaSociety for Medical Informaticsof Bosnia and Herezegovina

University of Sarajevo

European Federation forMedical Informatics

INTERNATIONAL CONGRESS OF

THE EUROPEAN FEDERATION FOR

MEDICAL INFORMATICS22nd

30.8. - 02.09.

2009

bridge

over

trouble

dw

ate

r

MIE‘09Sarajevo


Papers

original paperSUMMarYPurpose: Standardization of education process and al-most every aspect of life in EU moved authors of this paper to evaluate medical informatics education at medical faculties in bosnia and Herzegovina. Very complex political structure and existence of two enti-ties, one district and ten cantons in Federation of bos-nia and Herzegovina caused big difference in curric-ulums, teaching methods and quality of knowledge gained among medical faculties in the country. also, on the example of teaching process at Medical Fac-ulty, University of Sarajevo, authors proposed future united and integrated system in this area. Work meth-od: Method of the study is descriptive, compared edu-cation in medical informatics at five b&H medical fac-ulties. over 500 students answered questionnaires created at medical faculties in Sarajevo and tuzla. they respond on content of the subject of medical in-formatics, possibility of accepting knowledge from both practical and theoretical lessons, “good” and “bed” sides of curriculum, as well as students’ com-puter literacy. Work results: Subject of Medical infor-matics is being taught on at least 3-4 different ways. Medical schools in banja luka and Foca/Srbinje are under heavy influence of University belgrade, Serbia and Montenegro, teaching staff in Mostar are from Croatia, university of tuzla has its own way and Medi-cal faculty in Sarajevo which maintains high quality values and principles. things and events which make different Medical faculty University of Sarajevo is the fact that this is the only medical faculty in bosnia and Herzegovina to has web site of Cathedra for medi-cal informatics, organized number of events including distance learning course and with the most compe-tent teaching staff. Medical faculty in Sarajevo is the oldest medical school in bosnia and Herzegovina es-tablished in 1944. as obligatory subject, medical in-formatics is introduced in school year 1992/1993, and it is the only medical faculty in bosnia and Herzegov-ina where medical informatics is taught in two se-mesters, second and eleventh. discussion: three areas are important to be included in the discussion: Qual-ity of education in secondary schools should be im-proved, lack of multimedia equipment, good laN, high-speed connection to internet and well organized web design and issues in regard to maintenance of equipment and as the third one that students must have free access to computer rooms to enable them to broad their knowledge in spare time and general

information about health system must be available to students to allow them to find out about role and im-portance of medical informatics in “real life”. Natural-ly, we raised question about unique and systematized medical informatics education in the whole coun-try, unrespectable of entities, nationality or religion of students.Conclusion: Medical informatics education at Medical Faculty, University of Sarajevo is based on the same concept as is it on prestige universities all over the world and in accordance with recommendations of the working groups on education of EFMi and iMia. other medical faculties in bosnia and Herzegovina should be using the same working methodology and system of work in order to have standardized educa-tion in medical informatics and to achieve high qual-ity in education. to enable us to follow European and global achievements in this area, in education system and as well as in health system power of fact should be prevaling.Key words: medical informatics, education, distance learning

1. introductionMedical informatics focuses on the acquisition, storage and

use of information in health and biomedicine. Education of health workers and standardization of education process are essential for all stakeholders in healthcare systems and medi-cal staff to know how to control healthcare costs, patient safety and utilization of information technology. Very complex polit-ical structure in Bosnia and Herzegovina caused big difference in curriculums, teaching methods and quality of knowledge gained among medical faculties in the country what is con-sequent of the war (1992-1995). Dayton Agreement signed in November 1995 by all parties involved in Bosnian war and representatives from international communities, was the doc-ument that stopped killing in Bosnia. But, Dayton Agreement also, created very strange and bizarre environment in which is very difficult to enable progress and improvement in almost all areas including education and health system. Since 1995, Bos-nia and Herzegovina is consisted of two entities, Federation of Bosnia and Herzegovina and Republic of Srpska. Federation of Bosnia and Herzegovina is divided on ten cantons and each canton has separate government with ministry for education and health, among others, and unique politics in those areas. It created diversity of laws and very controversy situation in edu-cation and health system. In other hand Republic of Srpska is entity with very centralized government which inhibits prog-ress and local autonomy. Having in mind that Bosnia and Her-zegovina is country with less then 3.5 million citizens, it is clear that something should be changed. Every 70th citizen of Bos-

Medical informatics Education in bosnia and Herzegovinaizet Masic, ahmed Novo Medical Faculty University of Sarajevo, bosnia and Herzegovina

AIM 2005, 13(4): 184-188



nia and Herzegovina uses Internet (50.000 citizens of 3.5 million in Bosnia and Her-zegovina). Comparison with western world and European Union is horrifying. The aim of this paper is to point out dif-ferent education from Medical informat-ics at Bosnia and Herzegovina Medical faculties. Leading role of the Cathedra for Medical informatics at Medical faculty of University of Sarajevo is reflected in ar-eas as: Distance learning, tele-exam, final exam, publishing activities, organization of various events – workshops, conferenc-es, congresses etc (1,2,4,5,6,7).

2. research of students’ knowledge

Cathedra for Medical informatics was established before the war1, at three Med-ical faculties in Bosnia and Herzegovina, Sarajevo, Tuzla, Banja Luka, with adoption of common curri-cula in 1990. As obligatory subject Medical informatics was in-troduced in 1992 in above medical faculties and in additional two, in Mostar and Foca/Srbinje, as soon as the faculties were established after the war. (8,9,10,11,12,13,14,15)

One of the greatest challenges for teaching staff at Medical faculties in B&H is student’ knowledge they bring with them. On beginning and on the end of each school year at Cathe-dra for Medical informatics at Medical faculty of University of Sarajevo teaching staff conduct survey on students knowledge before and after the semester assessment of the level of medi-cal informatics education, and so far over 800 students were examined.

If we compare students enrolled in current school year 2004/2005 with “war and post-war generations” we can no-tice significant improvement. Students come from secondary schools with higher level of knowledge comparing with peri-od ten years ago.

Obvious example for that is the fact that 4.1% students do not know how many enter buttons are on keyboard versus 15.4% in period 1994-1997. Length of informatics education in secondary schools extended, so that we have more of 50% stu-dents who had 2 and more year’s education from this subject (1,2,3,4). It has influence on how students assess their knowl-edge, which is higher then in previous school years and general attitude to curriculum (figure 1). Standard of students also in-creased and now we have 76.6% of students who have personal computer and 47.5% have connection to Internet in compari-son with year 1997 when 55.8% of students had personal com-puter and 21% had connection to Internet.

These are some results from last school year: Theoretical part of curriculum (15 units) on medical informatics is difficult (43.9%), medium difficult (33.4%) and very difficult (22.7%). The curriculum is complex for 16.2% students asked. It is use-ful for 67.4% and for 21.6% is extremely useful. Practical part of the curriculum is difficult for 42.8%, very difficult for 21.2%, curriculum is not clear for 37.5%, 67.2 think that curriculum is good and 14% consider it very good. Satisfied with the theo-retical lectures were 47.2%, very satisfied 23.4% and highly sat-isfied 14.6%. Very satisfied with practical lectures were 72.3% students and 17.6% of they were satisfied. Students would like to have more lectures and practical training on internet appli-� Bosnian war commenced in March �992 and ended in Novem-ber�995.

cations, Tele-matics, electronic media and communications in-formation systems.

Good teachers and curriculum can improve general knowl-edge on informatics of medical student to the certain level, but academics are not able to influence on: non common curricula in entire Bosnia and Herzegovina, very bad economic situation and overall political environment.

3. Medical informatics in bosnia and Herzegovina

In the following chapter we compared comparable param-eters in all 5 Medical faculties in Bosnia and Herzegovina. Pa-rameters are: number of classes, faculty website and website of cathedra for Medical informatics, teaching staff, distance learning etc. (13,14,15,16,17,18,19,20,21).

3.1. Medical Faculty MostarMedical faculty of Croatian University of Mostar is edu-

cational institution established after the Bosnian war and the subject Medical Informatics is introduced from the beginning. Since the Medical faculty does not have own teaching staff thay use visiting professors from Croatia, Universities of Split and Zagreb. Website of Medical faculty is not created yet and ni-ether for Cathedra for Medical Informatics.

3.2. Medical Faculty banja lukaMedical faculty in Banja Luka is leading medical school in

smaller B&H entity (Republic of Srpska). It is established in 1978, and number of graduated student is 1107 so far. There is no cathedra for medical informatics as such and this sub-ject is been taught under subject Statistics and informatics in medicine. Subject Statistics and informatics in medicine is be-ing taught in second semester and students are obliged to have 30 hours of theoretical and 30 hours of practical work. There is faculty website (figure 2), but there is no cathedra’s one.

3.3. Medical Faculty Foca/SrbinjeMedical faculty in Foca/Srbinje, University of East Saraje-

vo is the “youngest” leading medical school in B&H. It is estab-lished in 1994 and they have 3 generation of graduate students. There is no cathedra for medical informatics and this subject is been taught under subject Medical statistics and informatics as well as in Banja Luka and with the same curricula, 30 hours

FiGUrE 1. Survey graphs – school year 2004/2005


Papers

of theoretical and 30 hours of practical work. There is faculty website (figure 3), but there is no cathedra’s one. Since school year 2000/2001 there is subject Medical statistics and informat-ics at postgraduate studies.

3.4. Medical Faculty tuzlaMedical faculty in Tuzla is established in 1976. There is no

cathedra for medical informatics as such and this subject is been taught under subject Basic computing and informatics. Subject Statistics and informatics in medicine is being taught in third and fourth semester and students are obliged to have 30 hours of theoretical and 30 hours of prac-tical work. There is faculty website (figure 2), but there is no cathedra’s one since there is no cathedra for medical informatics. It is specific of University of Tuzla that they have same university cathedra for subject informatics to be taught at all faculties.

3.5. Medical Faculty SarajevoMedical faculty in Sarajevo is the oldest

medical school in Bosnia and Herzegovi-na established in 1944 (12). As obligatory subject medical informatics is introduced in school year 1992/1993. Currently medical informatics is being taught in second semester (30 hours of theoretical and 30 hours of practical work) as Basic medical informatics and in eleventh semester as Applicative medical informatics from 2006. Since school year 1993/1994 there is subject Medical sta-tistics and informatics at four semester postgraduate studies for all three courses.

4. Medical Faculty University of Sarajevo

are leaders in the Medical informatics education at Med-ical faculties. In cooperation with UTIC (University Tele-in-formation Centre) teaching staff from Cathedra for Medical Informatics have organized first course in distance learning; also they have organized fist tele-exam in history of Medical faculty, prof Masic MD PhD is Editor-in-Chief of two profes-sional-scientific magazines Medicinski arhiv (Medical archive)

and Acta Informatica Med-ica (AIM) and chair person of B&H Society for Medi-cal Informatics and teach-ing assistant Ahmed Novo MD is secretary of B&H So-ciety for Medical Informat-ics and member of editorial board of Acta Informatica Medica. Also, numbers of various events – workshops, conferences, congresses etc. have been organized in sup-port of Cathedra for Medi-cal Informatics. Teaching staff is actively involved in preparation and organiza-tion of MIE 2009 to be held in Sarajevo from 31 August to 3 September 2009.

4.1. distance learning and website

In October 2003, University of Sarajevo began with Dis-tance learning education, opening University Distance Learn-ing Centre. Opening the University Distance Learning Centre, as coordination body and leader in all activities in connection to Distance learning, has provided opportunity for develop-ment and growth of this kind of lifelong education.

The project is conducted by the University Tele-informa-tion Centre (UTIC) and four faculties from University of Sara-jevo are involved: Electro-technical, Business and economy, and Medical faculty.

On UTIC web site, seven students enrolled from Medical faculty, for the subject Medical Informatics are able to learn from the distance location. So far, professor Masic and his as-sistants uploaded eleven lectures at the site: Hardware and soft-ware, Medical documentations, Medical informatics, Methods of data manipulation, Nomenclatures and classification sys-tems, Data organization, Data, information and knowledge, Lectures 1, System and communication, Structure and data or-ganization and Expert systems. Beside the materials it is pos-sible to upload and download the following: Practical works, Seminar work, Information, Recommended links, Plan and programs, Quiz, Schedule, Recommended readings, Examina-tion schedule and Examination results. Basically software ap-plication has two interfaces: teacher and student interface. Ac-cess from any of these is very simple and fast.

In year 2004, Cathedra for Medical informatics at Medi-cal faculty of University of Sarajevo got a web site with very broad and interesting content. By the end of the school year

FiGUrE 2. Web sites of Medical faculties in B&H

FiGUrE 3. Web site of Cathedra for Medical informatics Medical Faculty University of Sarajevo



2004/2005 we expect to have the following completely func-tioning at the cathedra’s web site (figures 8 and 9) and student will be able to:

Check terms for examsRegister for exam Check results of the examsAttend medical informatics education via MI web (distance learning).

Other operational stuffs on the web are: content and news from Cathedra for Medical informatics, or-ganization, teaching process, projects, publications, scientific events, visits and useful links.

4.2. First tele-examFirst Tele-exam in history of Sarajevo Universi-

ty was held at Medical faculty in 2002. Realization of the Tele-exam was composed of examination com-mission in Sarajevo, of three members – three teachers and on the other part, at the distance; in this case, the candidate was placed in Podgorica, Serbia and Montenegro with the presence of the special commission of Ministry of Education FB&H. In-ternet-web based communication system was established with the help of three small programs for the Internet Web commu-nication: MSN Messenger, Net Meeting and IP watcher and on the same time the candidate was asked to answer on the few questions in the front of web camera. The exam was success-fully made. Taking into consideration that this kind of exami-nation had been made for the first time in the history of Uni-versity of Sarajevo and generally in Bosnia and Herzegovina we have been faced with positive as well as negative reactions on this kind of examination. We believe that this is valuable con-tribution to student-teacher communication for the future and abundant world experience of the Internet – Web based com-munications are the real proof for it.

••••

4.3. EventsPresents of Distance learning is still modest, but in Decem-

ber, 2002 was made first big step forward. As first phase of the project: Possibilities of introduction of Distance learning in Medical curriculum, as part of the celebration of Ten years of Cathedra for Medical Informatics, was held workshop the project approved by the Federal Ministry of Education, BiH. Participants of the workshop, eminent experts from Sarajevo, Tuzla and Zagreb, Croatia, shared experience in application of distance learning. As a part of workshop we made direct com-munication with UTIC (University Tele-information Centre) in real time, speed 512Kbps (5).

University Tele-information Centre, established as part of University of Sarajevo and first ISP in Bosnia and Herzegov-ina in 1996 (www.utic.net.ba). It is scientific-organizational unit of the University of Sarajevo for improvement of scientif-ic-research work and through UTIC members of the Univer-sity can be gathered in the unique computer-communication structure. Objectives of UTIC are: to connect members of the University with similar institution in the country and abroad due to more efficient use of scientific, research and educational resources, use of educational data bases and other information for the needs of the University and its members. Also: Devel-opment an integration of informatics computer technologies in education Creation of flexible infrastructure which will en-able e-Learning to be accessible to all students and Universi-ty staff, Improvement of general digital literacy of academic population, Development of top quality educational content which could be integrated in the actual European processes of e-Learning revolution.

Second congress of Medical Informatics of Bosnia and Herzegovina with international participation, New trends of Health Informatics, was held in Sarajevo in the period 16-18

April 2004. The congress was opened by prof Masic and first presenter was chairman of the European Federation of Med-ical Informatics (EFMI), Assa Reichert. Among seventy four papers presented and over 150 participants were members of EFMI; Baud R, Bryden J, Engelbrecht R, Weber P, Rudel D and Hofdijk J. Venue of Second congress were Rectorat of Universi-ty of Sarajevo and the congress was held under the auspices of prof Boris Tihi, Rector of University of Sarajevo.

5. ConclusionsThe rise of IT as an artefact of everyday life in the modern

world has brought with it the dawn of a new era, often dubbed the “Age of Information”. These technologies are changing the way we perceive the world, how wee think and communicate with another. Established cultures are being transformed and new cultures are forming. New virtual environment affects the way we build our sense of who we are.

Some characteristics of the Internet:

FiGUrE 4. BHDMI host of MIE 2009 - Web site and LOC

FiGUrE 5. First tele-exam at University of Sarajevo


Papers

Large volume of users and potential users,Lack of physical boundaries which allows for the manipu-lation of time and space,Information can be accessed in a concurrent fashion us-ing different media,Concept of redundancy.

In the virtual environment we are applying for information in a way that is expanding our senses and one must to take into account that experience is occurring in the context of the vir-tual environment. Information without a context has no mean-ing.

Expected outcomes of the project Introduction and Imple-mentation of Distance learning in medicine are:

Development and integration of informatics-computer technologies in medical educationCreation of flexible infrastructure which will enable access to e-Learning by all students and teaching staffImprovement of digital literacy of academic populationEnsure high educational standards to students and teach-ing staff andTo help medical staff to develop “Lifelong learning way of life”.

The health sector is one of the most evident potential ben-eficiaries of the Internet revolution and World Wide Web re-source in the present and in the future, when the tools now available and the system’s reliability and efficacy as a whole will be further incremented and improved.

Low level of education in secondary schools is improving, but it is still low;Shortage of modern equipment;Lack of LAN, connection to Internet, organized web de-sign;Problem of the maintenance;Free access to computer rooms;Students are not informed about functioning of educations and health system in whole, especially at universities;All new measures and decisions must be formalized through legislation;In our system power of knowledge should be prevailing, instead of power of authorities.

REfEREnCEs1. Mangrulkar RS.Targeting and structuring information re-

source use: a path toward informed clinical decisions. J Contin Educ Health Prof. 2004 Fall; 24 Suppl 1: S13-21.

2. Tang S, Helmeste D. Digital psychiatry. Psychiatry Clin Neurosci, 2000 Feb; 54(1): 1-10.

3. Mašić I, Bilalovic N, Karčić S, Kudumović M, Pašić E. Telemedicine and telemetric in B&H in the war and Post war Times. European Journal of Medical Research, 2002; 7 (supl.1): 47.

••

•

•

•

•

••

•

•

••

•••

•

•

4. International Medical Informatics Association. Dostupno na URL adresi: http://www.imia.org. Datum pristupa in-formaciji 17. 01. 2005.

5. Branstetter BF 4th, Bartholmai BJ, Channin DS.Reviews in radiology informatics: establishing a core informatics curriculum.J Digit Imaging, 2004 Dec; 17(4): 244-8.

6. Hasman A, Haux R. Curricula in medical informatics. Stud Health Technol Inform, 2004; 109: 63-74.

7. Mantas J. Future trends in Health Informatics--theoreti-cal and practical. Stud Health Technol Inform, 2004; 109: 114-27.

8. Marques EP, Marin HF, Massad E, Fraser H, Ohno-Mach-ado L.Training in health informatics in Brazil.Stud Health Technol Inform, 2002; 90: 757-60.

9. Recommendations of the International Medical Informat-ics Association (IMIA) on education in health and medi-cal informatics. Methods Inf Med, 2000; 39: 267-77.

10. Kern J. Medical informatics in the medical curriculum--when? Stud Health Technol Inform, 1999; 68: 484-8.

11. Kern J. Medical Informatics Education – How to Tailor the Course to Given Stage of Student’s Knowledge? AIM, 2002; 10(2): 59-61.

12. Masic I, Kudumovic M, Pasic E. Education of Medical In-formatics at Medical Faculty of Sarajevo. Med Arh, 2004; 58 (1, supl. 2): 88-91.

13. Mašić I, Kudumović M, Pašić E. Deset godina edukacije iz Medicinske informatike na Medicinskom fakultetu u Sa-rajevu. Med Arh, 2002; 56 (4): 233-9.

14. Šabanović Z, Mujčinagić A. Edukacija medicinske infor-matike na Medicinskom fakultetu Tuzla. Med Arh, 2004; 58 (1, supl.2): 92-4.

15. Masic I, Novo A, Kudumovic M, Teleeducation at bio-medical faculties in Bosnia & Herzegovina. AIM, 2004; 58 (3-4): 15-8.

16. Mašić I, Riđanovic Z. Medicinska informatika, knjiga 1. Avicena, Sarajevo, 2000: 107-30.

17. Mašić I, Riđanovic Z. Medicinska informatika, knjiga 2. Avicena, Sarajevo, 2000: 176-90.

18. Mašić I, Riđanovic Z. Edukacija iz medicinske informa-tike i njene perspektive, AIM, 2001; 9 (3-4): 59-60.

19. Masic I, Ramic-Catak A, Kudumovic M, Pasic E. Distance learning in the medical education in B&H: E-Health & Education Proceedings, Zagreb, 2002: 17.

20. Ramic-Catak A, Masic I. Distance learning – ucenje s dis-tance u medicinskoj edukaciji. AIM, 2002; 10 (2): 63-6.

21. Mašić I. Korijeni medicine i zdravstva u Bosni i Hercego-vini. Avicena, Sarajevo, 2004.

Correspondence to: Prof dr izet Masic, Medical Faculty, University of Sarajevo, Cekalusa 90, 71000 Sarajevo, bosnia and Herzegovina,

Phone: +387 33 663 742, E-mail: [email protected]



original paperSummaryobjectives: in order to assess diagnostic value of pro-lactin (Prl) in breast cancer (bC), we examined its se-rum levels and frequencies of its increase in breast cancer patients (bCP), and compared them to those in two controls. We also determined circulating lev-els of Prl in localised and advanced bC and calcu-lated sensitivity and specificity of Prl in bC. Patients and methods: the main experimental group con-sisted of 47 female patients with histologically con-firmed diagnosis of bC. the obtained results were compared to those in two control groups: clinical-ly healthy women, and female patients with other types and locations of cancer. Serum levels of Prl were measured by means of radioimmunoassay. re-sults were processed by means of t-test and two way analysis of variance. results: the serum levels of Prl before treatment, as well as the frequencies of its in-crease, were significantly higher in bCP in compari-son to controls (p<0.01, 0.02). the average circulat-ing levels of Prl in patients with advanced bC were significantly higher (p<0.0001) in comparison to pa-tients with localised disease. Sensitivity for Prl in bC was 50%, and specificity was 100%. Conclusions: in-creased levels of Prl can be detected in the majority of patients with advanced bC. Prl has high specificity for bC, especially for metastatic bC, which leads to its diagnostic and prognostic importance in this disease.Key words: prolactin, localised breast cancer, advancer breast cancer, sensitivity and specificity of Prl.

1. introductionThe role of the peptide hormone prolactin (PRL) in either

the etiology or progression of human breast cancer is not clear. It is well established that PRL is involved in development and differentiation of the normal gland in mammalian species, and in rodent model systems it is a key player in the development of mammary cancer (1, 2). However, while literature suggests that PRL levels may influence human breast cancer, there is no clear correlation between circulating PRL levels and the etiol-ogy or prognosis of the disease (3, 4, 5).

With the goal of assessing of clinical usefulness of PRL de-termination in breast cancer patients, we examined its baseline levels and frequencies of its increase in breast cancer patients, and compared them with control values. We also measured se-rum PRL levels in patients with metastatic breast cancer and compared them with those in localised disease. Sensitivity and specificity for PRL have also been determined.

2. Patients and Methods

2.1. PatientsThere were two control groups of patients designated as

I, and II. Group I consisted of forty clinically healthy women with an age interval 30-65 years. All of them were non-smok-ers, occasionally coffee and alcohol drinkers, non-obese, and were not under any medication, including birth control. All of them had normal mammograms. The circulating levels of PRL were measured in all of them at least three times during the fol-low up period.

Group II consisted of thirty three patients having cancer of different histologic type and location with an age interval 18-77 years. Most of them had lung and colon cancer. The serum levels of PRL before any therapy (baseline levels) have been measured in all of them, and later at regular time intervals dur-ing the five years observation period. This group was further divided according to the presence of metastases into two sub-groups: patients with metastases, and those without metasta-ses.

The main experimental group consisted of forty seven fe-male patients with histologicaly proven diagnosis of breast cancer and with an age interval 38-82 years. The same data have been collected from them as in group II. This group was further divided according to the presence of metastases into two subgroups: patients with metastases, and patients without metastases.

2.2. the determination of serum levels of PrlBlood samples were drawn under sterile conditions at eight

in the morning each time, centrifuged at 3000 rpm for 10 min-utes under room temperature, and serum was stored at –20oC until processed. The circulating levels of PRL were measured by means of radioimmunoassay (RIA) method using commer-cially available kits from Serono, Switzerland. The major char-acteristics of this method are: principle – immunoradiomet-ric assay, separation method – double antibody RIA, standard control – WHO/RP 75/504 Standard Prolactin (1ng PRL=32.5 µ WHO 75/504), cross reaction with human gonadotropic hor-mons, GH – 0.15%, TSH – 0.39%, and range of normal values 52-520 mU/L.

2.3. Estimation of range of normal values for Prl We first estimated the range of normal values in 40 healthy

female subjects. Our referral values were mean +- 2 standard deviations. The values of normal ranges stretched between 308-738 mU/L. There were no age dependent significant dif-ferences in PRL concentrations. All cases with their values of PRL above 738 mU/L were judged as hyperprolactinemic, and all of them with their values below 310 mU/L were declared as

the Sensitivity and Specificity of Prolactin (Prl) in breast Cancer PatientsZlata Mujagić1, Hamza Mujagić2

department of biochemistry, Medical faculty, University of tuzla, bosnia and Herzegovina1

Clinical oncology, Medical faculty University of tuzla, bosnia and Herzegovina2

AIM 2005, 13(4): 189-191


Papers

hypoprolactinemic.The determination of sensitivity and specificity of PRL The sensitivity and specificity of PRL were calculated ac-

cording to the following formulas: Sensitivity = true positive results / (true positive + false negative results) Specificity = true negative results / (true negative + false positive results).

2.4. Statistical analysis of resultsThe results were evaluated using Student’s t-test with calcu-

lated two standard deviations and errors, and analysis of two way variance in F-test. The nature of distribution and linearity were checked using histogram presentation and tests of lineari-ty. The statistical significance of results was grouped according to the following criteria: p>0.05 (not significant), p<0.05-0.01 (significant), p<0.009-0.001 (very significant), and p<0.0009-0.0001 (highly significant).

3. results

3.1. the circulating levels of Prl

Table 1 shows the average se-rum levels of PRL and frequencies of its increase before treatment in breast cancer patients compared with healthy women and patients with other locations and histolog-ic types of cancer. The circulating levels and the frequency of PRL in-crease before treatment in breast cancer patients were significant-ly higher than in healthy women (p<0.01), and they also were higher than in patients with other locations of cancer (p<0.02). None of woman from healthy control group was hyperprolactinemic. There was also no correlation between age and menopausal

status and levels of PRL in this control group.

3.2. the circulating levels of Prl in patients with localised and advanced breast cancer

The initial PRL levels for patients who had strictly localised disease, and for those patients with widespread disease are shown in table 2. The average serum PRL level for the localised breast cancer group was 257.5 mU/L, and for advanced breast cancer group was 830.5 mU/L. Metastases developed in all hy-

perprolactinemic patients. All seven patients without metas-tases were hypoprolactinemic (table don’t show). The average serum levels of PRL in patients with advanced breast cancer were significantly higher (p< 0.0001) in comparison with pa-tients with localised breast cancer. Such a difference did not show up when patients with other histologic types and loca-tions of cancer were compared. The sensitivity/specificity at-test to these results.tablE 2. The circulating levels of prolactin (PRL) in localised and widespread breast cancer.

Group/number of patientsPrl (mU/l)mean +- S.E.

p value

breast cancer patients without metastases

257.5 +- 46.9a<0.0001b vs abreast cancer patients with

metastases830.5 +- 78.4b

other cancer patients without metastases

540.5 +- 87.4c<0.61d vs cother cancer patients with

metastases482.2 +- 69.9d

3.3. the sensitivity and specificity of Prl in breast cancer patients

Table 3 shows the results of calculated sensitivity/specifiity for PRL in breast cancer patients, and table 4 show the same for other cancer patients. The sensitivity and specificity for PRL in breast cancer patients were 50% and 100%, respectively, and in other cancer patients were 34.8% and 70%, respectively.

4. discussionThe number of hyperprolactinemic cases before treat-

ment was significantly higher in breast cancer patients than in healthy controls (p<0.01), and in patients with other histolog-ic types and locations of cancer (p<0.02) (table 1). This dif-ference points towards the specificity of PRL for breast can-cer which leads to its diagnostic and prognostic importance in this disease. However, measurements of serum PRL levels in breast cancer patients gave thus far controversial results. Some population studies showed mild but still significant increase of its circulating levels in breast cancer patients (6, 7), and even in daughters of these patients (8, 5) while other studies did not confirm this finding (9). The discrepancy in these results could at least in part be explained by the circadian rhythm of PRL secretion and with a number of other factors that influ-ence its secretion (estrogen/progesterone balance, stress, cer-tain drugs etc.). In addition, the circadian rhythm of PRL se-

Group/number of patients

Prl (mU/l) mean+-S.E.

< upper referent value (730 mU/l)

> upper referent value (730 mU/l) p value

Number/total number of patients (%)

Control group i 40

523.1+-16.9a 40/40 (100) 0/40 (0) <0.01 c vs a

Control group ii 33

524.1+-65.2b 24/33 (73) 9/33 (27) <0.02 c vs b

breast cancer Patients 47

689.8+-67.5c 27/47 (57) 20/47 (43) <0.5 b vs a

tablE 1. The average circulating levels of prolactin (PRL) in healthy women (control group I), patients with other types and locations of cancer (control group II), and in breast cancer patients.

Group Number of patients with hyperprolactinemia

Number of patients with normo and hypoprolactinemia

Sensitivity (%) Specificity (%)

Number of patients with metastases

tP (20) FN (20)tP/tP+FNx100= 20/40x100 = 50

tN/tN+FPx100 = 7/7x100 =100 Number of patients

without metastases FP (0) tN (7)

tablE 3. The sensitivity and specificity of prolactin (PRL) in breast cancer patients.

TP = true positives, number of patients with metastases correctly classified by the test.FN = false negatives, number of patients with metastases missclassified by the test.FP = false positives, number of patients without metastases missclassified by the test.



cretion from the pituitary differs between groups at high vs low risk of breast cancer (10), with no seasonal variations (11). In order for measurements to be reproducible and comparable it is important to exclude as much as possible all of these non-spe-cific factors.

Diagnostic usefulness of PRL in breast cancer was also as-sessed on the basis of its sensi-tivity and specificity in this disease. Sensitivity or true-posi-tive rate was in fact the frequency of elevated serum PRL levels in breast cancer patients with metastases and it was 50% (ta-ble 3), while specificity or true-negative rate was the frequen-cy of normal serum levels of PRL in non-metastatic breast can-cer patients. Since none of the patients without metastases had hyperprolactinemia the specificity of PRL for advanced breast cancer was judged as 100% (table 3). This is in accordance with other findings (12, 13).

It is clear that mean circulating levels of PRL measured be-fore therapy was significantly higher (p<0.0001) in patients with widespread breast cancer than in those with localised dis-ease (table 2). Such a difference did not exist in control group of patients with various other types and locations of cancer (ta-ble 2). All hyperprolactinemic patients developed the meta-static disease during the follow up period. Hypoprolactinemia has been detected in all non-metastatic patients. These results undoubtedly suggest the connection between the hyperprolac-tinemia and unfavorable prognostic factors of metastatic dis-ease. It can be concluded that breast cancer is more aggressive in hyperprolactinemic patients than in those with normal or below normal levels of this hormone. The results of some other studies support these findings (12, 14, 7).

However, it still remains to be clarified whether hyperpro-lactinemia is the result or cause of breast cancer. Further re-search to clarify the tight connection between PRL and the disorders of the immune system could possibly explain this problem (15).

REfEREnCEs1. Vonderhaar BK, Biswas R. Prolactin effects and regulation

of its receptors in human mammary tumor cells. In: Cellu-lar and Molecular Biology of Mammary Cancer, Eds Me-dina D, Kidwell W, Hepner G, Anderson E, pp 219-205, New York, Plenum Publishing Corp.

2. Vonderhaar BK. Prolactin involvement in breast cancer. Endocrinol Rel Cancer, 1999; 6: 398-404.

3. Wang DY, Hampson S, Kwa HG, Moore JW, Bulbrook RD, Fentiman IS, et al. Serum prolactin levels in women with breast cancer and their relationship to survival. Eur J Can-cer Clin Oncol, 1986; 22: 487-92.

4. Ingram DM, Nottage EM, Roberts AN. Prolactin and breast cancer risk. Med J Austral, 1990; 153: 469-73.

5. Love RR, Rose DR, Surawicz TS, Newcomb PA. Prolactin and growth hormone level in premenopausal women with breast cancer and healthy women with a strong family his-tory of breast cancer. Cancer, 1991; 68: 1401-05.

6. Hill P, Wynder EL, Kumar J, Helman P, Rona G, Kono K. Prolactin levels in populations at risk for breast cancer. Cancer Res, 1976; 36: 4102-06.

7. Patel DD, Bhatavdekar JM, Chikhlikar PR, Ghosh N, Suthar TP, Shah NG, et al. Node negative breast carcino-ma: hyperprolactinemia and/or overexpression of p 53 as an independent predictor of poor prognosis compared with newer and established prognosticators. J Surg Oncol, 1996; 62: 86-92.

8. Levin PA, Malarkey WB. Daughters of women with breast cancer have elevated mean 24-hour prolactin levels and partial resistance of PRL to dopamine suppression. J Clin Endocrinol Metabol, 1981; 53: 179-83.

9. Fishman J, Fukushima D, O’Connor J, Rosenfel RS, Lynch JF, Guirgis H, Maloney K. Plasma hormone profiles of young women at risk for familial breast cancer. Cancer Res 1978; 38: 4006-011.

10. Haus E, Lakatua DJ, Halberg F, Halberg E, Cornelissen G, Sackett LL, et al. Chronobiological studies of plasma pro-lactin in women in Kyushu, Japan and Minnesota, USA. J Clin Endocrinol Metabol, 1980; 51: 632-40.

11. Holdaway IM, Mason BH, Gibbs EE, Rajasoorya C, Letha-by A, Hopkins KD, et al. Seasonal variations in the secre-tion of mammotrophic hormones in normal women and women with previous breast cancer. Breast Cancer Res Treat 1997; 42: 15-22.

12. Holtkamp W, Nagel GA, Wander HE, Rauschecker HF, VonHeyden D. Hyperprolactinemia is an indicator of pro-gressive disease and poor prognosis in advanced breast cancer. Int J Cancer, 1984; 34: 323-28.

13. Lissoni P, Vaghi M, Ardizzoia A, Fumagalli E, Tancini G, Gardani G, Conti A, Maestroni GJ. Efficacy of monoche-motherapy with docetaxel (taxotere) in relation to prolac-tin secretion in heavily pretreated metastatic breast can-cer. Neuroendocrinol Lett, 2001; 22(1): 27-9.

14. Bhatavdekar JM, Patel DD, Vora HH, Ghosh N, Shah NG, Karelia NH, et al. Node-positive breast cancer: prognostic significance of the plasma prolactin compared with ste-roid receptors and clinicopathological features. Oncology Reports, 1994; 1: 841-45.

15. Prystowsky MB, Clevenger CV. Prolactin as a second mes-senger for interleukin 2. Immunomethods, 1994; 5(1): 49-55.

Correspondence to: Prof Zlata Mujagic, Medical faculty University of tuzla, Univerzitetska 1, 75000 tuzla, fax: ++ 387 35 300 601,

e-mail:[email protected]

Group Number of patients with hyperprolactinemia

Number of patients with normo and hypoprolactinemia

Sensitivity (%) Specificity (%)

Number of patients with metastases

tP (8) FN (15)tP/tP+FNx100 = 8/23x100 = 35

tN/tN+FPx100 = 7/10x100 =70 Number of patients

without metastases FP (3) tN (7)

tablE 4. The sensitivity and specificity of prolactin (PRL) in patients with other types and locations of cancer.

tP, FN, FP, TN – see legend in table 3.


Papers

original paperSUMMarYStatistical Methods play an increasingly important role in modern medical research [1]. in medical research, we often deal with different issues and diseases, as a result of which we collect data. For the analysis and in-terpretation of the data, the researcher needs Statis-tics. the tools of statistical inference and modeling are now essential aids to the researchers in the medical field. there are different statistical techniques used to analyze the data, one of these is called survival analy-sis. Survival analysis has found wider applications in the field of biostatistics.the aim of this article was to analyze the most frequently inherited blood disease in Pakistan i.e. β-thalassaemia. it has emerged as one of the major health problems and is one of the most causes of death. So for various studies have been con-ducted to study the incidence of β-thalassaemia and to determine the possible risk factors for the disease. the study was carried out along the following lines: to model the survival time of β-thalassaemia, a total of 320 patients records were examined in Fatimid Foun-dation. For each patient, the phenomena of β-thalas-saemia was studied in relation to different factors, namely, Sex, age of the patient, duration of treatment, Height of patient, Weight of patient, Mean Corpuscu-lar Haemoglobin, Family history, Status of the patient, relation of parents, blood group, Number of transfu-sion, Education level of patients. the response varia-ble in the study was survival time of patients in Fatimid Foundation. Kaplan-Meier survival function and Cox re-gression model were applied on to the data,. the initial consideration that Haemoglobin (Hb) may be one of the risk factors for the survival of thalassaemic patients was not supported by the model. Separate cox regres-sion models were fitted for each sex (considering inter-section). the analysis showed that number of transfu-sions, height of the patient, age of the patient * mean corpuscular haemoglobin, age of the patient * mean corpuscular haemoglobin * height of the patient, while in case of females significant factors were height of the patient, age of the patient * mean corpuscular haemglobin. important risk factors for the survival of thalassaemic patients are age of the patient, mean cor-puscular haemoglobin, number of transfusions and height of the patient. However, in case of males and fe-males some interactions of these four factors also play important role.Key words: β- thalassaemia, Survival anakysis, Kap-lan-Meier survival function, Cox regression.

1. introductionThalassaemia genes are remarkably widespread, and are be-

lieved to be most prevalent of all human genetic disease [2]. The word thalassaemia was derived from two Greek words - Thalassa meaning the sea and haima meaning blood, literally “sea water in the blood”. But Thalassaemia was recognized as a clinical entity by Dr Thomas Cooley and Dr Pearl Lee who de-scribed five cases of Thalassaemia in 1925 [3]. Thalassaemia is an inherited disorder in which there is an abnormality in one or more of the globin genes. In general thalassaemia can be di-vided into two groups, α-thalassaemia and β-thalassaemia.

α-thalassaemia :. People whose hemoglobin does not pro-duce enough alpha protein have α-thalassaemia. It is common-ly found in Africa, the Middle East, Southeast Asia, southern China, and occasionally the Mediterranean region.

β- thalassaemia:. People whose hemoglobin does not pro-duce enough beta protein have β- thalassaemia. It is found in people of Mediterranean descent, such as Italians and Greeks, and is also found in the Arabian Peninsula, Iran, Africa, Paki-stan, Southeast Asia and southern China. There are three types of β- thalassaemia that also range from mild to severe in their effect on the body.

Thalassemia Minor or Thalassemia Trait. In this condition, the lack of beta protein is not great enough to cause problems in the normal functioning of the haemoglobin. A person with this condition simply carries the genetic trait for thalassemia and will usually experience no health problems other than a possible mild anemia. As in mild α-thalassaemia, physicians often mistake the small red blood cells of the person with β- thalassaemia minor as a sign of iron-deficiency anemia and in-correctly prescribe iron supplements.

Thalassemia Intermedia. In this condition the lack of beta protein in the haemoglobin is great enough to cause a mod-erately severe anemia and significant health problems, includ-ing bone deformities and enlargement of the spleen. However, there is a wide range in the clinical severity of this condition, and the borderline between thalassemia intermedia and the most severe form, thalassemia major, can be confusing. The deciding factor seems to be the amount of blood transfusions required by the patient. The more dependent the patient is on blood transfusions, the more likely he or she is to be classified as thalassemia major. Generally speaking, patients with tha-lassemia intermedia need blood transfusions to improve their quality of life, but not in order to survive.

Thalassemia Major is also called Cooley’s Anemia, Ho-mozygous β-thalassaemia, Homozygous thalassaemia or Med-iterranean anaemia [4]. This is the most severe form of β- tha-lassaemia in which the complete lack of beta protein in the hemoglobin causes a life-threatening anemia that requires reg-ular blood transfusions and extensive on going medical care.

analysis of β - thalassaemia by Using the Survival analysis techniquesSalahuddin1 and Qamruz Zaman2

department of Statistics University of Peshawar, Pakistan1

department of Medical Statistics, informatics and Health Economics, innsbruck Medical University, austria2

AIM 2005, 13(4): 192-197



These extensive, life long blood transfusions lead to iron-over-load which must be treated with chelation therapy to prevent early death from organ failure.

Thalassaemia is the common disease in Pakistan, out of every 100 Pakistanis, 6 are carriers of Thalassaemia gene [5]. There are few research reports in Pakistan that deal with tha-lassaemia [6]. Constantoulakis et. al. [7] studied the strature and lognitudinal growth in thalassaemia major by taking 229 Greek patients. They showed that a high percentage of Greek children with homozygous β-thalassaemia were short and had a lag in growth, though some severely affected attained normal height. Bone age lagged significantly behined the chronolog-ic age, even in those with normal height for their age. A longi-tudinal study demonstrated an initial normal growth velocity with decreased after the age of six for males and eight for fe-males. No patient had a spurt of growth at adolescence. Height gain continued up to the age of 21 and then ceased. Statistical analysis with a multiple regression equation showed a signif-icant relationship between the age and height of the parents. Low haemoglobin levels and severity of the disease seemed to hinder normal growth, but neither relationship was statistically significant. Anichini et. al. [8] determined the predicted value of a new osmotic test in the screening of heterozygous β- tha-lassaemia. For this, they applied a new osmotic test ( time for 50% haemolysis in standard solution ) on patients affected with heterozygous β-thalassaemia in a population consisting of 19 thalassaemic patients, 15 sideropenic patients, and 52 controls. The same population was examined for heterozygous β-tha-lassaemia using electronic measurement of the erythrocyte in-dices. Sensitivity, specificity, predictive value and efficiency of the two tests were calculated. Statistical analysis did not show any significant differences in sensitivity and specificity be-tween the two methods. They conclude that the osmotic test is sensitive, reliable and rapid; it seems to a valid substitute for electronic haematology analysers in Countries where they are not available. Politis et. al. [9] have investigated the concepts of the average citizen of continental Greece and crete about the problem of thalassaemia, his awareness of the basic character-istics of the disease and his attitudes towards some ‘life-style’ parameters of the thalassaemic sufferer. They used the meth-od of stratified random sampling and interviewed 3500 people aged 15-65 years, carried out multivariate statistical analysis of responses to the items on thalassaemia in relation to sex, age, educational level, place of residence and marital status. The re-sults of this survey show that education was the factor with the greatest influence on the likelihood of being aware of thalas-saemia and of having accurate knowledge of the basic features of the disease. There was also a general effect of age with the highest awareness and the best knowledge being found in those aged 21-44 years, while place of residence was a significant fac-tor for two items. The features known best were that thalassae-mia is non-infectious, that it calls for blood transfusion, and is inherited. Education had the strongest influence on the prob-ability that a person holds the attitude that the sufferer from thalassaemia can play normal role in society’s regards work-ing and having a family, as well as generally living a normal life and, in the case of thalassaemia children, playing and going to school. The attitude that thalassaemics should work instead of receiving a state benefit depended on education, age and place of residence. Amrolia,-P-J [10] prospectively assessed the rela-tive contribution of host and donor to haemopoiesis following stem cell transplantation (SCT) in children with β- thalassae-mia major (n = 35), using karyotype analysis or Southern blot/polymerase chain reaction analysis of variable number tandem repeats on genomic DNA from peripheral blood. Early haemo-

poiesis was fully donor in origin in 24 out of 35 cases and re-mained so throughout the post-transplant course in all but one patient, who evolved to stable mixed chimaerism. The remain-ing 11 cases (31%) initially showed mixed chimaerism: four of these rejected, one eventually eradicated host haemopoiesis to become fully donor haemopoietic, and the remaining six had persistent mixed chimaerism, with 5--38% host haemopoiesis. The risk of graft rejection was high when > 15% host haemo-poiesis was present at 3 months post transplant: four out of six such patients rejected their grafts; conversely, zero out of 29 pa-tients with < 15% host haemopoiesis at 3 months rejected (P < 0.0001). There was a higher incidence of significant acute and chronic graft-versus-host disease in patients with full donor chimaerism. These studies confirm that the mixed chimaeric state is common following SCT for thalassaemia, often persists (with up to 4 years follow-up) and is compatible with long-term cure. Analysis of chimaerism in patients undergoing SCT for β- thalassaemia enables monitoring of engraftment in the early post-transplant period, provides insight into the biolo-gy of engraftment and may be useful in identifying patients at high risk of rejection.

Since thalassaemia is an inherited blood disease and it’s treatment is a life long process, which introduce the concept of censoring, which can not be solved by ordinary regression method. The purpose of this article is to analyse the β-thalas-saemia, which is common in Pakistan by considering the con-cept of censoring and using the survival analysis technique. Survival analysis is a tool for handling the censored data. The non-parametric and semi-parametric techniques of survival analysis are used to analyse the data collected from the fatimid foundation Peshawar [11].

2. data sourceThe data of 320 patients were collected between Novem-

ber 2002 and April 2003 from the Fatimid Foundation, Pesha-war. The objective of the study was to analyse the prevalence of Thalassaemia in the patient population registered with the organization. Fatimid Foundation, a non-profit charitable or-ganization is the pioneer of voluntary blood transfusion serv-ices in Pakistan. It is a symbol of hope for millions of blood disease carriers in Pakistan. Thalassaemia or β-thalassaemia is an inherited blood disease from which now-a-days almost every 500th child suffers. The present study was conducted to find the factors, which are the main causes of the disease. The data were taken from the complete records of the Foun-dation . Out of these 101 were censored and 219 were events. . Furthermore, out of these 320 patients 129 were females (85 events and 44 censored) and 191 were males (57 censored and 134 events). Relevant information was recorded on a special-ly designed Performa with the help of the foundation doctors. Complete medical history was taken from each patient includ-ing Time, Status, Sex, Age, Mean Corpuscular Haemoglobin, Height, Number of Transfusion, Family history.

3. Statistical analysisThe analysis is divided in to two separate parts based on sex

and is conducted by the following waysKaplan-Meier survival curve for femaleCox Regression for femaleKaplan-Meier survival curve for maleCox Regression for male

3.1 Kaplan-Meier survival function for femaleSurvival analysis plays an increasingly important role in

biostatistics and is based on different techniques i.e. paramet-

••••


Papers

ric, semi-parametric and non-parametric. Out of which the most reliable and frequently used method is a non-paramet-ric method [12].The popular method of non-parametric tech-nique is Kaplan-Meier [13]. Which is used for estimating the survival probabilities and curve. We applied the method to the female data and as a result, we obtained the Fig.1. Where * represents the events. If we look at the curve, we can say that it starts from 1, as the first event occurred on the first observed time and also the last observed time is event. Which means that in this situation, it completed the requirements of function and we can calculate the mean survival time. The mean surviv-al time for female is 112 months.

The results of Kaplan-Meier for females are summarized in the following table-1tablE 1.

Quantile Estimate [lower, Upper)

75% 143 134,156

50% 118 100, 128

25% 80 72, 90

Mean 112 104,120

The median survival time is 118 months, which means that half of these patients survived up to 118 months and half will enjoy the 118 months more. Similarly, third and first quartiles give also the same information’s. The last two columns give the lower and upper confidence limits of the corresponding quar-tiles.

3.2 Cox regression for femaleFor the regression modelling, we use the semi-parametric

approach i.e. Cox Regression [14].To fit the cox regression, the first and foremost step is to select the most significant factors out of various factors as discussed earlier. For this purpose, the stepwise procedure is used by taking p= .1 and considering the following five factors;

Age = Age of the Patient MCH = Mean Corpuscular HaemoglobinFh = Family HistoryHt = Height of the Patient N = Number of transfusionWith these five explanatory variables, we shall construct

possible models including interaction or powers of the ex-planatory variables. On the basis of global chi-squared score value, the best eight models selected at six different stages are given below:

To select the best model among these models, the parame-ter estimates and stander errors of these models, obtained from SAS are given below

Model 1. Variable dF Parameter

EstimateStandard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

age *Ht 1 -0.000301 0.0000346 75.45856 0.0001 1.000

Model 1 indicates that among the one variable model, Age*Ht is the best model. As the p value is very small, reveals that age is highly associated with the height of the female pa-tient.

Model 2.

Variable dFParameterEstimate

Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

Ht 1 -0.037752 0.00829 20.71486 0.0001 0.963

age*MCH 1 -0.000957 0.0001579 36.73400 0.0001 0.999

In model 2, there are two factors i.e. height and the interac-tion of age and mean corpuscular haemoglobin. As p values of both these show that both are significant and both play the im-portant part in the life of thalassaemia patient.

Model 3.Variable dF Parameter

EstimateStandard Error

WaldChi-Square

Pr > Chi-Square

riskratio

Ht 1 -0.011976 0.01472 0.66174 0.4159 0.988

age *MCH 1 -0.000687 0.0001599 18.45172 0.0001 0.999

age*MCH*Ht 1 -0.000009216 4.22583E-6 4.75629 0.0292 1.000

A three factor model 3, involves height having p value 0.4159 which is very high. The interaction of age and mean corpuscular haemoglobin is significant in the model. While a three factor interaction of age, mean corpuscular haemoglobin does not play any important part.

Model 4.


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

N 1 -0.008270 0.00375 4.86360 0.0274 0.992

Ht 1 -0.040771 0.01009 16.32982 0.0001 0.960

age * MCH 1 -0.000887 0.0001929 21.13204 0.0001 0.999

age*MCH* N*Ht 1 7.3145351E-9 1.7634E-8 0.17206 0.6783 1.000

This model includes number of transfusion, height , inter-action of age and mean corpuscular haemoglobin and the in-teraction of age, mean corpuscular haemoglobin, number of transfusion and height. Among these four factors, the middle two factors play the significant part, while the p-vales of first and the last reveal that they are not so important.

Model 5.


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

age 1 -0.051943 0.00946 30.13152 0.0001 0.949

N 1 -0.023460 0.00695 11.38351 0.0007 0.977

age * N 1 0.000149 0.0000412 13.11003 0.0003 1.000

MCH *Ht 1 -0.001005 0.0005708 3.09934 0.0783 0.999

MCH*N*Ht 1 -0.000002363 3.45227E-6 0.46861 0.4936 1.000

Model 5 , which is a model of five variables depends most-ly on age, number of transfusion ,the interaction of age and the number of transfusion and the interaction of meancorpuscular

Number of Variables

Variables

Score value included in models

1 71.1244 age*Ht

2 117.9534 Ht age*MCH

3 137.9325 Ht age*MCH age*MCH*Ht

4 161.1322 N Ht age*MCH age*MCH*N*Ht

5 176.3637 age N age*N MCH*Ht MCH*N*Ht

6 187.3652Fh N Ht age*MCH age*MCH*N age*MCH*Fh*Ht

FiGUrE 1 Kaplan-Meier curve for female



haemoglobin and height due to their small p-values. While the last factors have the high p-values.

Model 6.


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

Fh 1 -0.347936 0.43403 0.64262 0.4228 0.706

N 1 -0.025757 0.00557 21.40668 0.0001 0.975

Ht 1 -0.030980 0.00957 10.47096 0.0012 0.969

age*MCH 1 -0.001653 0.0002789 35.11663 0.0001 0.998

age*MCH*N 1 0.000005115 1.33743E-6 14.62913 0.0001 1.000

age*MCH*Fh*Ht 1 -0.000002561 2.47714E-6 1.06863 0.3013 1.000

In model 6, the first factor is family history which has the highest p-value and is a clear indication that it is not signifi-cant. A four factors interaction is also not significant.

By adopting the procedure of Kleinbaum [15], the best model is selected from the above eight models. Model 3-6 give similar information involving some insignificant factors. While the model 1 and 2 involve only the significant factors. Between these two , the model 2 is the best as it gives more in-formation than model 1.

Once again the model along with the other in formations is given below:


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

Ht 1 -0.037752 0.00829 20.71486 0.0001 0.963

age*MCH 1 -0.000957 0.0001579 36.73400 0.0001 0.999

In terms of PH model h(t, X) = h0(t) exp [(-0.037752 ) Ht + ( -0.000957) Age*MCH ]

Except the small p-value of Ht, the model shows that e-0.037752 = 0.963 times Ht effect the model adjusted for interaction of Age and MCH. While e-0.000957 =0.999 times Age*MCH effect this adjusted for the other factor Ht. It means that height (Ht) and the interaction of age and mean corpuscular haemoglob-in play the important part for the survival of female thalassae-mia patients.

3.3 Kaplan-Meier survival function for maleThe Kaplan-Meier survival function for the 191 male pa-

tients is calculated and the results are described in Fig. 2. It also

covers the whole range of survival function i.e. 0 ≤S ≤1. So the mean survival times for male patients is 110 months.

By applying this method on the data, the summary of the three quartiles is given in the following Table-2. tablE 2.

Quantile Estimate [lower, Upper)

75% 145 134,163

50% 97 94, 110

25% 74 71, 84

Mean 110 102,118

Table-2 contains the quantile Kaplan Meier estimate with 95% confidence interval.

The median survival time is 97 months, which means that half of these male patients survived to 97 months and half will enjoy the same more. Which is comparatively smaller than the female survival time. While the survival times of third quartile is two months greater than the female times. Similarly, the first quartile gives the smaller survival times than that of female.

3.4 Cox regression for maleThe following factors were considered for applying the cox

regression to male patients data:Age = Age of the Patient MCH = Mean Corpuscular Haemoglobin Ht = Height of the Patient N = Number of transfusionWith these four explanatory variables, we constructed

models including interaction or powers of the explanatory variables. By adopting the same procedure global chi-squared score value, the best six models selected at six different stages are given below

For finding the best model, the parameter estimates of these models are given below:

Model 1.


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

age *MCH 1 -0.001236 0.0001190 107.86744 0.0001 .999

Age and mean corpuscular are strongly associated. In oth-er words, in case of one factor model the interaction of age and mean corpuscular haemoglobin play the important part for the survivor of male patient.

Model 2.


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

age 1 -0.027958 0.00375 55.65362 0.0001 0.972

MCH* Ht 1 -0.001570 0.0002020 60.38062 0.0001 0.998

Both the factors in model are very important for male pa-tients. As both have the low p-value of 0.0001. The individu-al role of age and the interaction of mean corpuscular haemo-globin can not be ignored.

FiGUrE 2 Kaplan-Meier curve for male

Without With

Criterion Covariates Covariates Model Chi-Square

-2 loG l 681.819 571.256 110.563 with 2 dF (p=0.0001)

Score 117.953 with 2 dF (p=0.0001)

Wald 85.843 with 2dF (p=0.0001)

Number of Variables

Variables

Score value included in models

1 109.5148 age*MCH

2 197.5202 age MCH*Ht

3 257.4915 Ht age*MCH age*MCH*Ht

4 270.4915 N Ht age*MCH age*MCH*Ht

5 284.3661N Ht age*MCH age*MCH*N age*MCH*Ht

6 289.5780N Ht age*MCH age*Ht age*MCH*N MCH*N*Ht


Papers

Model 3.


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

Ht 1 -0.056997 0.01260 20.46343 0.0001 0.945

age *MCH 1 -0.001348 0.0001982 46.23942 0.0001 0.999

age*MCH*Ht 1 0.000007144 3.96568E-6 3.24495 0.0716 1.000

In model 3 again all three factors are significant due to p<.1. So, these factors show their importance in the disease.

Model 4.


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

N 1 -0.008904 0.00165 29.04701 0.0001 0.991

Ht 1 -0.052185 0.01295 16.24189 0.0001 0.949

age * MCH 1 -0.001184 0.0001987 35.48681 0.0001 0.999

age*MCH*Ht 1 0.000005723 3.96974E-6 2.07846 1.4994 1.000

Model 4 shows that the last factor i.e., the interaction of age, mean corpuscular haemoglobin and height does not show the presence in the model as p= 1.4994 >.1. While the first three factors are very important in the model.

Model 5.


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

N 1 -0.018456 0.00415 19.80984 0.0001 0.982

Ht 1 -0.043542 0.01299 11.23279 0.0008 0.957

age * MCH 1 -0.001640 0.0002655 38.15890 0.0001 0.998

age*MCH *N 1 0.000002694 1.05121E-6 6.56717 0.0104 1.000

age*MCH*Ht 1 0.000003931 3.64889E-6 1.16084 0.2813 1.000

Number of transfusion, height of the patient, interaction of the age of patient and mean corpuscular haemoglobin and the three factors interaction of age, mean corpuscular haemoglob-in and number of transfusion are significant, while the last fac-tor is not.

Model 6.


Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

N 1 -0.019763 0.00418 22.31602 0.0001 0.980

Ht 1 -0.042384 0.01350 9.86389 0.0017 0.959

age*MCH 1 -0.001460 0.0002610 31.31103 0.0001 0.999

age*Ht 1 0.000003776 0.0000849 0.00198 0.9645 1.000

age*MCH*N 1 0.000002237 1.1685E-6 3.66428 0.0556 1.000

MCH*N*Ht 1 0.000002634 1.86155E-6 2.00167 0.1571 1.000

The fourth and the last interaction factors are not signifi-cant, while the rest are not.

Again using the comparison method of Kleinbaum, it re-veals that models 1-3 involve significant factors, while models 4-6 involve insignificant factors too. So, the best model is se-lected from models 1-3. As model 3 gives more information, it is selected to be best.

Once again the model along with the other information is given below:

Variable dF ParameterEstimate

Standard Error

Wald Chi-Square

Pr > Chi-Square

riskratio

Ht 1 -0.056997 0.01260 20.46343 0.0001 0.945age *MCH 1 -0.001348 0.0001982 46.23942 0.0001 0.999age*MCH*Ht 1 0.000007144 3.96568E-6 3.24495 0.0716 1.000

In terms of PH model h(t, X) = h0(t) exp[(-0.056997) Ht + (-0.001348) Age*MCH + (0.000007144) Age*MCH*Ht ]

Except the small p-value of Ht, the model shows that e-0.056997 = 0.945 times Ht effect the model adjusted for interaction of Age and MCH and the interaction of Age, MCH and Ht. While e-0.001348 =0.999 times Age*MCH effect this model adjusted for the other factors. While Age*MCH*Ht e-0.000007144 = 1.00001 times effect the model adjusted for the other factors. It means that height (Ht), the interaction of age and mean corpuscu-lar haemoglobin (Age*MCH) and the interaction of Age, Mean corpuscular Haemoglobin and height play the important part for the survival of male patients.

4. ConclusionThe objective of the study was to model the incidence of

Thalassaemia in population of patients attending the Fatim-id Foundation. A total of 320 patient’s records were examined. For each patient, the phenomena of survival time (i.e. a time of treatment of a patient in Fatimid and Current Status) was examined in relation to different risk factors, namely Age of patients (Age), Family History (Fh), Height of patients (Ht), Number of transfusions (N), Mean Corpuscular Haemoglobin (MCH), Sex (S), Other variables under consideration were Re-lation of parents, Educational level, family size etc. Out of 320 patients 101 were censored and 219 were events. Furthermore, out of these 320 patients 129 were females ( 85 events and 44 censored) and 191 were males (57 censored and 134 events).

The study were divided into two stages:i) Survival time of female ii) Survival time of male Kaplan-Meier estimates of both the categories were ob-

tained , which gave the survival probabilities and curves of the males and females survival times. The quantiles table of male gave the median survival times of 118 months for females and 97 months were the median survival times of males, which is slighter smaller than females survival times.

In order to fit the cox regression for female, the covariates were selected by stepwise procedure at p=.1, which gave the five most important factors i.e. Age, Ht, MCH, Fh and N. All possible models were drawn with their corresponding global score chi-squared statistic from these factors. In this case we considered the interactions of these factors. On the basis of largest score value and parsimony approach, the best model se-lected was Ht and Age*MCH. Similarly, for males, the factors selected were Age, MCH, Ht and N, while the model selected was Ht, Age*MCH and Age*MCH*Ht.

In the light of analysis and results, following conclusions could be drawn. Thalassaemia is an inherited blood disease, in which family history and inter-family marriages play an impor-tant part. The gender of person has no significant effect on the presence of thalassaemia. In short, for the survival time of pa-tients, the important factors are Age of the patient, Mean Cor-puscular Haemoglobin, Number of transfusions and Height of the patients. These factors play the key role in the life of a thal-assaemia patient.

From these we can conclude that, there is no cure for thal-assaemia major, and it’s treatment of regular blood transfu-sion is a life lasting process, to boost haemoglobin levels in the blood. However, this cause effects such as diabetes, heart fail-ure. The only chance of a cure is the bone marrow transplanta-tion, but the risks are still considerable. This is more successful in young children who do not suffer from iron overload.

To prevent the disease different suggestions are recom-mended Couples from communities with a higher than aver-age risk of being carriers should be tested for β- thalassaemia, especially before starting a family.

Without With

Criterion Covariates Covariates Model Chi-Square

-2 loG l 1186.789 1000.519 186.270 with 3 dF (p=0.0001)

Score 257.491 with 3 dF (p=0.0001)

wald 153.803 with 3 dF (p=0.0001)



General awareness of the people regarding the features and complications of thalassaemia which can be carried out through different medias like newspaper, television, radio etc.

This is not only the responsibility of foundation and gov-ernment to control the disease, but it is also the responsibility of every carrier and especially, it is the duty of carrier marriage women to visit the doctors regularly during pregnancy.

REfEREnCEs1. Zaman Q, Pfeiffer KP. Statistics in Medical Research. Acta

Informatica Medica, 2005: 13(3): 147-8.2. Hong GR. Hemoglobin disorders. In: Behrman RE, Klieg-

man RM, Jenson HB, eds. Nelson textbook of pediatrics, 16th ed. Philadelphia, WB Saznders Company, 2000.

3. Oscar A. Thorup Jr, Bithell Thomas C., Hess Charles E., Keeling Richard P., Mohler Daniel N., Niskane Eero, Quesenberry Peter J., Roy E. Smith and Wheby Munseys Leavell and Thorup’s Fundamentals of clinical Haematol-ogy (fifth Edition). Library of Congress Cataloging -in- Publication Data, 1987.

4. Rino V, Bernadette M, Evgentia G. What is Thalassaemia, 2nd ed. Thalassaemia International Foundation, 1995.

5. Thalassaemia Society of Pakistan. Established, 1995.6. Ahmed S, Saleem M, Rashid Y, Abbas N, Malik A. Prena-

tal diagnosis of thalassaemia in Pakistan. First case report. Pakistan journal of Pathology, 1994: 5(1).

7. Constantoulakis M, Panagopoulos G, Augoustaki O. Stat-ure and longitudinal growth in Thalassaemia major. A study of 229 Greek patients. Clin-pediatr-(Phila), 1975: 14(4): 355-7, 362-8.

8. Anichini-P, Reynaudo G, Pacini G. Predictive value of a new Osmotic test in the screening of heterozygous beta-Thalassaemia. Scand-J-Clin-Lab-Invest, 1983: 43(2): 171-3.

9. Politis C, Richardson C, Yfantopoulos JG. Public knowl-edge of Thalassaemia in Greece and current concepts of the social status of the thalassaemic patients. Soc-Sci-Med, 1991; 32(1): 59-64.

10. Amrolia PJ, Vulliamy T, Vassiliou G, Lawson S, Bryon J, Kaeda J, Dokal I, Johnston R, Veys P, Darbyshire P, Rob-erts IA. Analysis of chimaerism in thalassaemic children undergoing stem cell transplantation.Br-J-Haematol, 2001; 114(1): 219-25.

11. Fatimid Foundation, A symbol of hope Pakistan, estab-lished, 1981.

12. Fleming TR. and Harrington DP. `Nonparametric estima-tion of the survival distribution in censored data`, Com-munication in Statistics – Simulation and Computation, 1984; 13; 1-26.

13. Kaplan EL, Meier P. Nonparametric estimation from in-complete observations. Journal of the American Statistical Association, 1958: 53(282): 457-81.

14. Cox DR. Regression models and life-tables. Journal of the Royal Statistical Society, B, 1972: 34; 187-220.

15. Kleinbaum DG. Survival Ananlysis A self-learning text. NewYork: Springer, 1995.

Correspondence to: dr Qamruz Zaman, phone: +43-512-507-3217, Fax #: +43-512-507-2711, e-mail: [email protected]


Papers

Professional paperSummaryEven if ontologies and information models can be distinguished by many points, they also have many data in common. Nowadays, we can consider diCoM (digital image Communication in Medicine) as a ref-erence information model for radiology and Unified Medical language System as a reference ontology for biomedicine. the goal of the study is to know if com-paring the diCoM Content Mapping resource and the UMlS, could help to build a medical imaging on-tology. the methodology has consisted in analysing and comparing the diCoM Structured reporting and the UMlS Semantic Network: the diCoM network is rather limited but efficient, while the UMlS is a huge database well organized. there is a lack of terms cor-responding to image analysis and technical vocabu-lary in the UMlS and a too complex redaction of Con-text Group id terms in diCoM. this shows that the UMlS should include more technical and semiologi-cal concepts to be suitable for radiology. it also sug-gests that diCoM could build its new Context Group id and template id respecting, if possible, UMlS Se-mantic types.Keywords: digital image Communication in Medicine; ontolo-gy; information model; Unified Medical language System

1. introduction Rector distinguishes between ontologies and information

models (1). He assimilates ontologies to terminology or con-cept models. He argues that, while the purpose of ontologies is essentially to represent meanings, information models are more concerned with the structure of the information. How-ever, information models may be populated by concepts drawn from terminologies and ontologies.

DICOM (Digital Image Communication in Medicine) of-fers an example of a reference information model for radiol-ogy. DICOM part 16 (2) has been created in 2001 and is in quick mutation. This part includes all codes and terminolo-gies appearing in DICOM objects (Images, structured report-ing, Physiological signals…). Thus it is called DICOM Content Mapping Resource (DCMR). In this document, terms are sort-ed in domains (Context Group ID: CID), from which are elab-orated templates (Template ID: TID). Those TID and CID are DICOM Structured Reporting (DICOM SR) roots.

The Unified Medical Language System® (UMLS®) has been developed and maintained by the U.S. National Library of Medicine since 1990. It comprises two major inter-related components: the Metathesaurus®, a huge repository of con-cepts, and the Semantic Network, a limited network of 135 Se-mantic Types, and 54 relations. The latter is a high-level repre-sentation of the biomedical domain based on Semantic Types under which all the Metathesaurus concepts are categorized, and which is intended to provide a basic ontology for the bio-medical domain (3).

We have to consider those two representations as references to create a medical imaging ontology. Indeed, DICOM in im-aging data management and UMLS as a reference repository of medical concepts are both indisputables.

The goal of this study is twofold:To know if coherences between DICOM and UMLS infor-mation representation could help to complement or create DICOM SR templates.To know if differences between the two representations could help to enrich DICOM and UMLS with new rela-tions and terms.

2. Materials and Methods DICOM SR information model network analysis: starting

from the elaboration of a generic standardized report, we ana-lyzed DICOM SR main concepts hierarchy and their relations.

UMLS Semantic Network analysis: we isolated the Seman-tic Types which have interest in imaging and then we analysed possible relations between Semantic Types (is-a and others)

Comparison between DICOM SR network and UMLS Se-mantic Network: Thanks to the two previous analyses, we made a comparison.

Comparison between DICOM SR 2004 CID content and UMLS Semantic type content: In order to compare the DICOM DCMR content with the UMLS ontology, we used NLM lexi-cal resource: MetaMap 2.3.C version (4). This software finds UMLS Metathesaurus concepts in different kind of texts. All CID contents have been analyzed with MetaMap. In this pa-per, only the MetaMap results for the main CID, identified in generic DICOM SR report elaboration, are presented. For re-sults analysis, only high probable terms have been considered (MAP), i.e. those with high mapping scores. The MetaMap re-sults are expressed in: percentage of CID terms correspond-ing to UMLS concept, percentage of CID terms correspond-ing to more than one UMLS concept, percentage of CID terms

•

•

ontologies and information Models: Comparing the UMlS and the diCoM Content Mapping resource Jérémy lasbleiz1,2, Valérie bertaud1, Fleur Mougin1, anita burgun1, régis duvauferrier1,2 Ea 3888, liM, Faculté de Médecine, Université rennes, France1

département de radiologie et imagerie Médicale , CHU de rennes, France2

ProFESSio

Nal

PaPErS

AIM 2005, 13(4): 198-200



which cannot be described by at least one UMLS con-cept. Moreover, we indicate the number of UMLS terms and the number of Semantic Types found for each CID

3. resultsDICOM SR information system network analysis

results:All the relations existing between CIDs in an imag-

ing report are summarized in figure 1. In DICOM SR, a report includes several different parts. First of all, an observation context (HAS OBS CONTEXT) which concerns the observer (radiologist, technologist…). The CID used are:

Observer : CID 7452 (Organisational Roles), CI-DS 7453 (Performing Role).Subject: CID 7455 (sex), CID 7457 (Units of mea-sure for age). Procedure: the reference is CID 29 (Acquisition Modality). Those procedures can need to speci-fy some acquisition contexts (HAS ACQ CON-TEXT). It can be the anatomic region explored (CID 4 and 4031 (Anatomic Region)) or radio-graphic projection CID 4009 (DX Anatomy Im-age). They are often related to concept modifi-er (HAS CONCEPT MOD) as CID 4010 (DX View) with CID 4011 (DX View Modifier). An-other example concerns sonography, « DX Anat-omy image » contained in CID 5 (Transducer ap-

•

•

•

proach). Another possible acquisition context for procedure is the patient positioning, speci-fied in CID 20 (Patient Orientation Modifier). Contrast injection is an important acquisition context specified in CID 13 (Contrast Agent Ingredient) and CID 11 (Route of administra-tion). Finally for MRI, the acquisition modal-ity is essential and specified in CID 7203 (Im-age derivation).

Beyond those elements, structured report-ing takes into account patient background (his-tory), test results (finding), general impression about the test, recommendation for the clini-cian and conclusion. CID uses are:Patient history is essential, it includes proce-dure reason. For mammography, it is speci-fied in CID 6051 (Breast procedure reason), CID 6055 (Breast clinical finding or indicat-ed problem) and CID 6080 (Breast cancer risk factors).General impression is specified in CID 6027 (Assessment from BIRAD) with six levels.Recommendations for the clinician are con-tained in CID 6029 (Recommended follow-up from Bi-Rads)Findings are well represented in DICOM lexi-con, the majority are “lesion finding” but can also be more precise with finding modifiers.

UMLS Semantic Network analysis results:Semantic Types regarding imaging report

are few. Only 46 types on 134 are useful. The 134 UMLS Semantic Types have been incor-porated into 15 groups, expressing better some notions as Disorders which contained every

•

•

•

•

FiG.1. rElatioNS between DICOM CID in a structured report (Relations: 1 - HAS OBS CONTEXT, 2 - HAS ACQ CONTEXT, 3 - CONTAINS, 4 - HAS PROPERTIES, 5 - HAS CONC MOD)

FiG.2. HiErarCHY of UMLS semantic Types that are the most frequent in medical imaging. Italic terms represent arborescence root (is-a type) which are too high level to be used in imaging reports. The four semantic groups that are the most represented in radiology are: anatomy ANAT, Chemicals & drugs CHEM, Disorders DISO, Living beings LIVB, Physiology PHYS, Concepts & ideas CONC, Procedures PROC.


Papers

Network relations of UMLS could be included in one DICOM SR relation (ex: HAS PROPERTIES (DICOM) = location_of, disrupted_by, mea-sured_by, affected_by, associated_with…(UMLS)). Short CID terms are well recognized by MetaMap, es-pecially in anatomy, pathology, ther-apeutic procedures and medical de-vice fields. Composed terms are a special problem, they should be con-sidered as not recognized. Indeed the lack of syntax in the UMLS pre-vents from representing the whole information. Finally, CID terms that are not recognized underline UMLS defects, particularly in semeiologic and technical domains.

In response to the two questions that motivated this study, we can make suggestions: It would be bet-ter, for CID terms corresponding to more than one UMLS concept, to segment those CID in UMLS Se-mantic Types and to link each other with the DICOM SR syntax as “has properties” and “has concept mod-ifier”. All CID are not homogenous and could earn to be, it would give them more meaning as ontology does. It would be interesting to take into account those remarks to elabo-rate the new DICOM SR templates. This could help the SR template conception using the entire UMLS knowledge database.The UMLS does not represent all the notions considered to be essential in medi-cal imaging. For example, its cover-

age in terms of descriptions and image analyses vocabulary is not sufficient for our purpose.

5. ConclusionThe UMLS and DCMR do not serve the same objectives.

The UMLS is an ontology/terminology whereas the DCMR of DICOM is an information model. Giving the Dicom informa-tion model an ontologic dimension would be of great value. Moreover, DICOM SR is a part of HL7 V3, i.e. of Clinical Doc-ument Architecture (CDA), and it would be very useful to have possible links with the UMLS.

REfEREnCEs 1. Rector AL. The interface between information, terminolo-

gy, and inference models. Medinfo, 2001; 10(Pt 1): 246-50.2. National Electrical Manufacturers Association. Digital Im-

aging and Communication in Medicine (DICOM), part 16: Content Mapping Ressource. Rosslyn, Va : NEMA, 2001.Available at : http://medical.nema.org/dicom/2003.html

3. Burgun A, Bodenreider O. Mapping the UMLS Semantic Network into general ontologies - Proc AMIA Symp, 2001: 81-5.

4. Aronson AR. Effective mapping of biomedical text to the UMLS Metathesaurus: the MetaMap program. Proc AMIA Symp, 2001: 17-21.

Correspondence to: Prof Jérémy lasbleiz, [email protected], département d’imagerie médicale – Hôpital Sud – 16, bd de bulgarie – bP 90347 – 35203 rennes cedex 2 - France

pathology and semeiologic types. If we analyze the 15 groups, only 9 have are interesting. Hierarchy of Semantic Types (is-a relations) represents the ontology in the field of medical imag-ing (fig. 2).

Comparison between DICOM SR network and UMLS Se-mantic Network results:

The possible relations between the seven UMLS Semantic Types are numerous. The main ones are : ANAT-CHEM : con-tain, ingredients_of ; ANAT-DISO : location_of, disrupt_by ; ANAT-PHYS : location_of ; ANAT-CONC : location_of ; ANAT-PROC : location_of, analysed_by ; CHEM-PROC : assesses_ef-fect_of ; CHEM-CONC : measure ; CHEM-PHYS : disrupts, af-fects ; CHEM-DISO : treats, diagnoses, causes ; DISO-PROC : diagnoses, results_of ; DISO-PHYS : affected_by ; DISO-LIVB : associated_with, occurs_in ; LIVB-PROC : performs, diagnoses ; LIVB-PHYS : affects ; PHYS-PROC : measures_by, assessed_for_effect_by ; PHYS-CONC : measurement_of ; CONC-PROC : evaluation_of, conceptual_part_of.

The DICOM CID and TID relations could be: CONTAINS, HAS OBSERVER CONTEXT, HAS CONCEPT MODIFI-ER, HAS PROPERTIES, HAS ACQUISTION CONTEXT, IN-FERRED FROM, SELECTED FROM

Comparison between DICOM SR 2004 CID and UMLS se-mantic type contents results:

Table I shows main CID statements implicated in observa-tion context examination. The analysis shows that anatomical and pharmacological CIDs correspond to UMLS concepts. On the other hand, terms about procedure and image characteristics doesn’t have correspondence with UMLS concepts. In between, there are clinical or radiological CID, which have the particulari-ty to include several UMLS terms and multiple Semantic Types.

4. discussionIt appears that no exact similarity between DICOM and

UMLS Semantic Networks. On the other hand, several Semantic

CID number of term

CID terms corresponding to UMLS concepts (percentage)

CID terms corresponding to more than one UMLS concepts (percentage)

CID terms can’t be described by at least one UMLS concepts (percentage)

Number of UMLS interested Semantic Types

Cid 4 anatomic region 130 80,77 6,92 12,31 3Cid 5 transducer approach 35 74,28 11,43 14,28 4Cid 11 route of administration

18 94,44 5,56 0 1

Cid 12 radiographic Contrast agent

34 94,12 0 5,88 4

Cid 20 Patient orientation Modifier

17 29,41 17,65 52,94 5

Cid 4009 dX anatomy images

73 94,52 4,11 1,37 4

Cid 4010 dX View 38 21,05 34,21 44,74 1Cid 4011 dX View Modifier 7 57,14 0 42,86 2Cid 6027 assessment from bi-radS®

6 16,67 33,33 50 5

Cid 6029 recommended Follow-up from bi-radS®

17 23,53 11,76 64,71 10

Cid 6041 Mammography image Quality Finding

44 6,82 11,36 81,82 17

Cid 6051 breast Procedure reason

13 46,15 38,46 15,38 12

Cid 6055 breast Clinical Finding or indicated Problem

14 35,71 42,86 21,43 12

Cid 6080 Gynaecological Hormones

5 80 20 0 4

Cid 6135 Chest image Quality Finding

14 7,14 42,86 50 5

Cid 7203 image derivation 32 3,12 28,12 68,75 12Cid 7452 organizational roles

13 46,15 0 53,85 2

Cid 7453 Performing roles 8 62,5 0 37,5 4Cid 7455 Sex 10 30 50 20 7Cid 7460 Units of linear Measurement

3 100 0 0 1

tablE 1. CID terms and UMLS concepts correspondence for the main CID. (Table total: terms =531, CID terms corresponding to UMLS concepts = 58%, CID terms corresponding to more than one UMLS concepts=14%, CID terms can’t be described by at least one UMLS concepts = 28%)



Professional paperSUMMarYUsing international nomenclatures and metathesau-ruses for coding of terminology in healthcare is the first and essential step for interoperability of hetero-geneous health record systems, which are the key-stones for shared medical care leading not only to effectiveness in healthcare but also to financial sav-ings and reduction of patients’ stress. this article de-scribes various international nomenclatures and me-tathesauruses used in healthcare. the main emphasis is put on the Unified Medical language System and mainly on the UMlS Metathesaurus, which helps us mostly in mapping of the professional healthcare ter-minology. in our work we try to verify practical ap-plicability of internationally used terminological dic-tionaries, thesauruses, ontologies, and classifications in attributes of the Minimal data Model for Cardiol-ogy, in the data Standard of Ministry of Health of the Czech republic, and in several chosen modules of commercial hospital information systems. the article describes problems appearing during the mapping process and it outlines their solutions.Keywords: metathesaurus, ontology, classification, nomen-clature, electronic health record.

1. introductionDetermination, denomination, and classification of medi-

cal terms are not optimal in comparison with other natural sci-ences. The proof is that for one term we can often meet with more than ten synonyms. Understanding of a more specific definition of a clinical unit (symptom, diagnosis) is different in different fields of medical schools, even in a national scale. In-ternationally accepted conventions are not very frequent. More rules stand for example in biology and zoology. In these fields there is a rule that the definition is valid according to the au-thor who has described a category as first. It prevents from re-petitive description of the same category with various names and hereby synonyms.

Let us show a negative example. In medicine there may ap-pear a situation when an effect of a new drug for a given diag-nosis is described in two publications. If the understanding of the diagnosis is in each publication slightly transferred and it stands for various groups of patients, then we can also meet with controversial results, which reduce the value of the final information.

This problem has intensified with introduction of a com-

puter technology to healthcare. Using computers means high-er uniqueness of data feeding, of term definitions, their precise denomination, etc., thereby the significant drawback becomes more noticeable.

Generally, in the scientific terminology it is more advanta-geous to use only one expression for one term. Computers are able to learn synonyms but it enlarges dictionary databases and the number of necessary operations grows. Moreover, synony-my in the scientific terminology leads to inaccuracy and mis-understanding. In current medicine we can meet with many synonyms for one single disease.

2. Classification SystemsClassification systems are coding systems based on creating

classes. The classes form aggregated terms, which correspond, at least, in one classification attribute. The classes of a classifi-cation must cover totally the defined field and they must not overlap. The formation of classification systems has been mo-tivated mostly by their practical usability in registration, sort-ing, and statistical processing of medical information. The first interest has been to register incidence of diseases and causes of deaths.

2.1. iCd – international Classification of diseasesThe foundation of the International Classification of Dis-

eases [1] was laid by William Farr in the year 1855. The World Health Organization took it over in the year 1948. At that time it was its 6th revision. The basic drawback of ICD lies in its low-er level of hierarchy. ICD is convenient for purposes of diag-nosis statistics but not for further coding of complex medical information as e.g. terms for symptoms and therapies are miss-ing. The last revision made an effort to classify in as much de-tail as possible (instead of the first digit there is a letter from the Latin alphabet, further places are digits).

Since 1994 the 10th revision of ICD is in use and it contains 22 chapters: Certain infectious and parasitic diseases (A00-B99); Neoplasms (C00-D48); Diseases of the blood and blood-forming organs and certain disorders involving the immune mechanism (D50-D89); Endocrine, nutritional and metabolic diseases (E00-E90); Mental and behavioural disorders (F00-F99); Diseases of the nervous system (G00-G99); Diseases of the eye and adnexa (H00-H59); Diseases of the ear and mastoid process (H60-H95); Diseases of the circulatory system (I00-I99); Diseases of the re-spiratory system (J00-J99); Diseases of the digestive system (K00-K93); Diseases of the skin and subcutaneous tissue (L00-L99); Diseases of the musculoskeletal system and connective tissue (M00-M99); Diseases of the genitourinary system (N00-N99);

Usage of international Nomenclatures and Metathesauruses in Shared Healthcare in the Czech republicPetra Preckova, Josef Spidlen, Jana ZvarovaEuroMiSE Centre of Charles University and the academy of Sciences Cr, department of Medical informatics, institute of Computer Science aS Cr, Prague, Czech republic

AIM 2005, 13(4): 201-205


Papers

Pregnancy, childbirth and the puerperium (O00-O99); Certain conditions originating in the perinatal period (P00-P96); Con-genital malformations, deformations and chromosomal abnor-malities (Q00-Q99); Symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified (R00-R99); Injury, poisoning and certain other consequences of external cause (S00-T98); External causes of morbidity and mortality (V01-Y98); Factors influencing health status and contact with health servic-es (Z00-Z99); and Codes for special purposes (U00-U99).

2.2. SNoMEdThe acronym SNOMED [2] stands for Systematized Nomen-

clature of MEDicine. SNOMED was published for the first time in the year 1965. It is a detailed reference terminology based on coding. It consists of more than 300 thousands of terms re-ferring to healthcare and it enables to use medical informa-tion whenever and wherever it is needed. SNOMED provides a “common language” enabling a consistent way of acquiring, sharing, and collecting healthcare data from various clinical groups among which we can rank nursing, medicine, labora-tory, pharmacies, and veterinary medicine. This classification system is used in more than 40 countries worldwide. SNOMED enables to describe any situations in medicine by means of 11 axes – dimensions: Topography; Morphology; Function; Living Organisms; Physical Agents, Activities and Forces; Chemicals, Drugs, and Biological Products; Procedures; Occupations; Social Context; Diseases/Diagnoses; General Linkage/Modifiers. Indi-vidual terms are determined by an abbreviation of a dimension followed by a hierarchical numerical code.

2.3. MeSHMedical Subject Headings (MeSH) [3] is a vocabulary con-

trolled by the National Library of Medicine (NLM) in the USA. It is composed of terms, which denominate keywords hierar-chically and this hierarchy helps with searching on various lev-els of specificity. Keywords are arranged not only alphabetical-ly but also hierarchically. On the most general level there are broad terms such as “anatomy” or “mental diseases”. NLM uses MeSH for indexing of papers from 4600 world best biomedical journals for the MEDLINE/PubMED® database. MeSH is used also for a database cataloguing books, documents, and audio-visual materials. Each bibliographical reference is connected with a class of terms in the MeSH classification system. Search-ing inquiries use also the MeSH vocabulary to find papers with required topics. The MeSH vocabulary is updated continuous-ly and it is also controlled by specialists creating it. They collect new terms appearing in scientific literature or in the arising fields of research. They define these terms in the frame of the contents of the existing vocabulary and they recommend their adding to the MeSH vocabulary. There exists also the Czech translation of MeSH. Unfortunately, the Czech translation is not complete and its quality is very low.

2.4. loiNC®

The Logical Observations Identifiers, Names, Codes - LOINC® [4] classification system is a clinical terminology, which is im-portant for laboratory tests and laboratory results. In the year 1999 the HL7 organisation accepted LOINC® as a preferred coding system for names of laboratory tests and clinical obser-vations. This classification system contains more than 30 000 various terms. The mapping programme called the Regenstrief LOINC Mapping Assistant (RELMATM) helps with mapping of local codes of various tests to the LOINC codes.

2.5. iCd-oThe ICD-O [5] classification system is an extension of the

International classification of diseases for oncology coding. This classification was firstly published by WHO in the year 1976. It is a four-dimensional system. These dimensions are Topography, Morphology, Progress and Differentiation. The dimensions are appointed to classify morphological kinds of tumours. The third version of ICD-O is used nowadays.

2.6. tNM ClassificationThe TNM classification [6] is a clinical classification of ma-

lignant tumours used for comparison of therapeutic studies. It proceeds from the knowledge that, for the disease prognosis, the localization and spread of a tumour is the most important.

2.7. dSM iii.DSM III. belongs to psychiatric nomenclatures. It contains

also definitions of individual terms. It is a very elaborate no-menclature. Unfortunately, it is a closed system without any link to other fields of medicine.

2.8. other Classification SystemsCurrently, there are more than one hundred of various clas-

sification systems. These are for example AI/RHEUM; Alterna-tive Billing Concepts; Alcohol and Other Drug Thesaurus; Beth Israel Vocabulary; Canonical Clinical Problem Statement Sys-temCurrent Dental Terminology 2005 (CDT-5); COSTAR; Med-ical Entities Dictionary; Physicians’ Current Procedural Termi-nology; International Classification of Primary Care; McMaster University Epidemiology Terms; Physicians’ Current Procedural Terminology; CRISP Thesaurus; COSTART; Diseases Database; DSM-III-R; DSM-IV; DXplain; Gene Ontology; HCPCS Version of Current Dental Terminology 2005 (CDT-5), 5; Healthcare Common Procedure Coding System; Home Health Care Classifi-cation; Health Level Seven Vocabulary; Master Drug Data Base; Medical Dictionary for Regulatory Activities Terminology (Med-DRA); MEDLINE; Multum MediSource Lexicon; NANDA nurs-ing diagnoses: definitions & classification; NCBI Taxonomy and many others.

3. tools for Sharing information from More Sources

The increasing number of classification systems and no-menclatures requires designing of various conversion tools for transfer between main classification systems and for record-ing of relations among terms in these systems. Extensive on-tologies and semantic networks are modelled for information transfer among various databases. Metathesauruses are de-signed to monitor and connect information from various het-erogeneous sources. UMLS is the most extensive project now-adays.

3.1. UMlSThe Unified Medical Language System (UMLS) [7] was initi-

ated in the year 1986 in the National Library of Medicine in the USA as a “long-term R&D project”. UMLS knowledge sourc-es are universal. It means they are not optimized for individ-ual applications. UMLS contains more than 730 000 biomedi-cal terms from more than 50 biomedical thesauruses. It is an intelligent automated system, which “understands” biomedi-cal terms and their relations and it uses this understanding for reading and organisation of information from machine pro-cessed sources. Its aim is to compensate terminological and coding differences of these non-homogeneous systems and



also language varieties of users. It is a multilingual thesaurus of classification systems such as MeSH, ICD, DSM, SNOMED and others on a high-capacity medium, which enables to trans-fer coded terms among various classification systems.

UMLS is based on three knowledge sources: Metathesau-rus®, Semantic Network, and SPECIALIST Lexicon. The Seman-tic Network contains information about semantic types and their relations. The SPECIALIST Lexicon records syntactic, morphologic, and orthographic information of each word or a term.

The UMLS Metathesaurus is an extensive, multi-purpose, and multilingual database. It contains information about bio-medical, healthcare and their relative terms, their various ex-pressions and relations among them. The UMLS Metathe-saurus has been developed from electronic versions of many various thesauruses, classifications or collections of codes, such as SNOMED, MeSH, AOD, Read Codes, ICD-10, and others. The main aim is to connect alternative expressions of the same terms and to identify useful relations among various terms. If thesauruses use the same expressions for different terms, then both meanings are present in the Metathesaurus and we can also see which meaning is used in which thesaurus. If the same term is used in different hierarchical contexts in various the-sauruses, then the Metathesaurus keeps all these hierarchies. The Metathesurus does not give one consistent view but it keeps many views, which are present in source thesauruses.

The computer application providing Internet access to knowledge and relative sources is called the UMLS Knowledge Source Server. Its aim is to make UMLS data accessible to users. The system architecture enables for remote users to send a que-ry to the National Library of Medicine. The UMLS Knowledge Source can be found at http://umlsks.nlm.nih.gov/. To enter the UMLS Knowledge Source Server users must register. After logging in we have to choose a version we would like to work with. The most recent is the 2005AB version. Then we enter a studied term. The identification number of a term, seman-tic type, definition, and synonyms will appear. As it was men-tioned hereinbefore, in medicine there are a lot of synonyms for one term. The UMLS Knowledge Source Server will show us in which classification systems the entered term appear. The information about similar, narrower or broader terms, seman-tic relations with other terms, and other detail information are available.

The most important for our work from the point of view of the first analysis of usability of these classification systems for needs of clinical contents description of some systems used in healthcare in the Czech Republic is to find out whether a giv-en term appears in the SNOMED CT classification system and to find out its identification number in this system. This and possibly identifiers in other systems can be later used in mod-elling of archetypes – basic building blocks of electronic health records.

3.1. SNoMEd CtSNOMED Clinical Terms (SNOMED CT) [8] originated

from two terminologies: SNOMED RT and Clinical Terms Ver-sion 3 (Read Codes CTV3). SNOMED CT represents the Sys-tematized Nomenclature of Medicine Reference Terminology de-veloped by the College of American Pathologists. It serves as a common reference terminology for gathering and acquiring health data recorded by organizations or individuals. The Clin-ical Terms Version 3 was developed by the United Kingdom’s National Health Service in the year 1980 as a mechanism for storing structured information on primary care in Great Brit-ain.

These two terminologies united in the year 1999 and a high-ly complex terminology SNOMED CT arose. Around 50 physi-cians, nurses, assistants, pharmacists, computer professionals, and other health professionals from the USA and Great Britain participate in its development. Special terminological groups were created for specific terminological fields, such as nurs-ing or pharmacy. SNOMED CT covers 364 400 health terms, 984 000 English descriptions and synonyms, and 1 450 000 se-mantic relations.

Among fields of SNOMED CT belong finding, procedure and intervention, observable entity, body structure, organism, substance, pharmaceutical/biological product, specimen, physi-cal object, physical force, events, environments and geographi-cal locations, social context, context-dependent categories, stag-ing and scales, attribute, and qualifier value. Nowadays we can meet with American, British, Spanish, and German versions of SNOMED CT.

4. Practical Usability of internationally developed Methods and tools in the Czech Healthcare

4.1. application of Classification Systems for Shared Health Care

Terminology mapping presented in applications of elec-tronic health record to internationally used terminological the-sauruses, ontologies, and classifications is the basis for interop-erability of heterogeneous systems of electronic health record. Understanding on the level of terminological terms is the basis for ensuring interoperability, however, it is not sufficient by it-self. Harmonization of a clinical content of a record is impor-tant. This harmonization does not have to be absolute; yet, it is possible to share only data, which are common among appli-cations. If so called reference information models of individual applications of health records correspond, it facilitates interop-erability. Of course, there are possibilities of mutual mapping between these models; however, it is difficult when considering different approaches of individual models.

For example the HL7 Reference Information Model (HL7 RIM) [9] represents a model of a closed world defined by means of classes, their attributes, and relations among class-es. The Domain Information Model (D-MIM) is derived from the HL7 RIM for further applications in a specific field. To get from this model to a record carrying information about a pa-tient health record we will use the Refined Message Information Model (R-MIM), which is a subset of D-MIM used for express-ing information contents of one or more abstract structures of records called also Hierarchical Message Descriptions (HMD).

CEN TC 251 is another example defining contents of elec-tronic health records in the European preliminary standard ENV 13606 (Electronic healthcare record communication, Part 4 – Messages for information exchange) by means of a rel-atively rough model specifying 4 basic components: Folder – describing bigger parts of an electronic health record of a given subject, Composition – representing one identifiable contribu-tion to the health record of a given subject, Headed Section – containing data sets on a more finer level than a Composition, and Cluster – identifying data sets, which should be kept clus-tered together if the lost of context is endangered.

The NEMA (National Electrical Manufactures Associa-tion) association in its DICOM SR (DICOM Structured Report-ing) specification uses an absolutely different approach. It ex-tends the Digital Imaging and Communication in Medicine for modelling of specifications for generation, presentation, ex-


Papers

change, and storage of DICOM medical images for modelling of the whole health record of a patient. The main idea is to use the existing DICOM infrastructure for exchange of struc-tured records represented as a hierarchical tree document with end nodes to store structured concepts. Semantics of individ-ual nodes is described by coding systems such as ICD-10 or SNOMED.

The reference model Synapses Object Model (SynOM) de-veloped in the frame of the Synapses, resp. SynEx (Synergy on the Extranet) project [10] is very similar to the model defined in CEN ENV 13606. Archetypes – definitions of structured col-lected attributes in a particular domain containing specified restrictions ensuring integrity of the whole record – are used as types of collected values. The project continued under the pa-tronage of the non-profit openEHR Foundation and it defined the Good European Health Record (GEHR) [11]. The specialists

of the project specify requirements of electronic health records with the main aim to support possibilities of integra-tion and cooperation of heterogeneous EHR applications. A formal model spec-ifying the GEHR architecture (GEHR Object Model, GOM) and a knowledge model specifying the clinical structure of a record by means of archetypes were developed for this purpose. Nowadays, results of the openEHR project can be considered as a significant competition to standards directed at the implemen-tation aspects of EHR systems.

4.2. terminology Mapping and development of archetypes

It is advisable for computer scientists to have the right terminology on mind from the beginning of archetypes proposing and developing of other basic elements in different model types of the healthcare record architecture.

The Ocean Informatics’ archetype editor [12] represented in the Figure 1 can be used to demonstrate how to refer the right terminology from the beginning of archetypes development.

It is possible to add an arbitrary number of languages in which we describe a term. It is also possible to choose from available terminologies. Those we will use to define the right

meaning of individual terms. In other tab-panels in this editor we define individual terms and in the Term Bindings panel we create relevant mapping of our terms to terms in terminologi-cal thesauruses in a way how it is displayed in the Figure 2.

4.3. Standardization of Clinical ContentsThe analysis of suitability and utilizability of individu-

al terminological thesauruses has been started by mapping of clinical contents of the Minimal Data Model for Cardiology (MDMC) [13] to various terminological classification systems. MDMC is a set of approximately 150 attributes, their mutu-al relations, integrity restrictions, units, etc. Prominent profes-sionals in the field of Czech cardiology agreed on these attri-butes as on the basic data necessary for an examination of a patient in cardiology.

During the analysis we have found out that approximately 85 % of MDMC attributes are includ-ed in, at least, one classification system. Most of them (more than 50 %) are in-cluded in the SNOMED CT system. At-tributes, from the point of view of pos-sibilities of their mapping to standard coding systems, can be classified in the following way:

•Trouble-free attributes – i.e. attri-butes, which can be mapped in a direct way, so only one possibility of mapping exists, possibly there are only synonyms with exactly same meanings and there-fore the same classification code (e.g. patient first name, current smoker, mo-tility, height of a patient, etc.).

•Partially problematic attributes – i.e. attributes, which can be mapped in a way that there are several possibili-ties of mapping to different synonyms,

which differ slightly in their meanings and usually in their clas-sification codes (e.g. ischemic cerebro-vascular stroke, angina pectoris, hypertension, congestive cardiac failure, etc.).

•Attributes with a too small granularity, i.e. attributes de-scribing certain characteristics on a too general level so that classification systems contain only terms of a narrower mean-ing (e.g. e-mail in MDMC versus e-mail to work / e-mail to home / e-mail of a physician and so on in classification sys-tems).

FiG. 1. Using terminologies in the Ocean Informatics’ archetype editor.

FiG. 2. Mapping of used terminology to standard coding systems.



•Attributes with a too big granularity, i.e. attributes describ-ing certain characteristics on such a narrow level so that clas-sification systems contain only a term of a more general mean-ing (e.g. symmetrical pulse of carotids, etc.).

•Attributes, which cannot be found in classification systems, e.g. dyslipidemy, etc.

Similar results were achieved when analyzing standardiza-tion possibilities of attributes of the Data Standard of Ministry of Health of the Czech Republic (DASTA) [14]. However, struc-tured attributes in this standard are limited to a large degree to administrative and laboratory data. The results of adminis-trative data mapping were similar to the results of administra-tive data mapping in MDMC. Laboratory data in this standard are specified in big details by means of the National Classifi-cation of Laboratory Items [15], on which analysis we are still working.

We also try to map attributes of chosen clinical modules of commercial hospital information systems. As an example we can show results from mapping of a specialized ECG module in the WinMedicalc clinical information system [16]. Because of the big specialization of this module we managed to map ap-proximately 60 % of attributes to various classification systems. Prevailing classification problems are connected with a too big granularity of attributes in this model (e.g. ejection fraction 1, ejection fraction 2, septum of left ventricle, etc.).

Close cooperation with physicians is essential for solving of such mapping problems. It is often needed to choose the right synonym substituting a certain technical term. It is necessary to do it very carefully not to lose information or not to misin-terpret it. In case it is not possible to do it without any lost of information, the better way is to describe a non-coded term by means of a set of several coded terms, possibly with showing mutual semantic relations. If this is not possible, we can po-lemicize with specialists whether these “indescribable” terms (attributes) can be replaced by other more equivalent or more standard ones. In special cases it is possible to add a certain term to an upcoming new version of a certain coding system. In case it is not possible to use any of the above mentioned pos-sibilities of solving mapping problems, it is necessary to cope with the fact that mapping will never be 100%. The insufficient mapping process limits the interoperability of heterogeneous systems used for various purposes in healthcare. Restricted in-teroperability is often inevitable from the very root of the prob-lem, e.g. insufficient harmonization of clinical contents of het-erogeneous systems of electronic health records.

5. ConclusionWe try to verify practical usability of internationally used

terminological thesauruses, ontologies, and classifications, specifically by studying attributes of the Minimal Data Mod-el for Cardiology, Data Standard of Ministry of Health of the Czech Republic and some chosen models of commercial hospi-tal information systems, which are sought out primarily in the SNOMED CT classification, secondary in other classifications. SNOMED CT is used in the HL7 version 3 and this is the rea-son why we try to map primarily to this classification system. In case of absence of a term we try other available terminolo-gies. The UMLS Metathesaurus is used to find appropriate re-lations to terms in other classification systems.

While mapping we face several problems, e.g. ambiguity in mapping and impossibility to map because of absence of a cor-responding term in classification systems. The big problem in using nomenclatures and metathesauruses in healthcare in the Czech Republic is the non-existence of Czech terminological systems or their appropriate Czech translations.

Despite some problems using international nomenclatures and metathesauruses in healthcare in the Czech Republic re-main, their using is the first essential step towards interoper-ability of heterogeneous systems of healthcare records. Suffi-cient interoperability of these systems is the basis for shared medical care leading to effectiveness in health care, financial savings and also to reduction of patients’ stress and this is the reason why we try to analyze how to use international classi-fication systems as best as possible for the needs of the Czech healthcare.

AcknowledgmentsThe work was supported by the project 1ET200300413 of

the Academy of Sciences of the Czech Republic.REfEREnCEs

1. World Health Organization©, International Classification of Diseases, 2005, http://www.who.int/classifications/icd/en/.

2. SNOMED International®, Systematized Nomenclature of Medicine, 2004, http://www.snomed.org/.

3. National Library of Medicine, Medical Subject Headings, http://www.nlm.nih.gov/mesh/MBrowser.html.

4. Regenstrief Institute, Inc., Logical Observation Identifiers Names and Codes – LOINC®, http://www.regenstrief.org/loinc/.

5. World Health Organization©, International Classification of Diseases for Oncology, 1990, http://www.cog.ufl.edu/publ/apps/icdo/.

6. Woxbridge Solutions Ltd©, General Practice Notebook – a UK Medical Encyclopaedia on the World Wide Web, 2005, http://www.gpnotebook.co.uk/simplepage.cfm?ID=1134166031.

7. United States National Library of Medicine, National Insti-tute of Health, Unified Medical Language System, http://www.nlm.nih.gov/research/umls/.

8. SNOMED International®, Systematized Nomenclature of Medicine – Clinical Terms, http://www.snomed.org/snomedct/.

9. Health Level Seven, Inc., HL7 Version 3 Standards, 2005, http://www.hl7.cz/.

10. Jung B., Grimson J. Synapses/SynEx goes XML, Studies in Health Technology and Informatics, 1999; Vol. 68: 906-11.

11. Centre for Health Informatics & Multiprofessional Education (CHIME), The Good European Health Record, http://www.chime.ucl.ac.uk/work-areas/ehrs/GEHR/.

12. Miro International Pty Ltd.©, Ocean Informatics, 2000-2004, http://oceaninformatics.biz/CMS/index.php.

13. Tomeckova M.: Minimal Data Model for Cardiology – Selec-tion of Data (in Czech). In: Cor et Vasa, 2002; Vol. 44, No. 4 Suppl.: 123.

14. Lipka J, Mukensnabl Z, Horacek F, Bures V. Current Commu-nication Standard DASTA of the Czech Healthcare (in Czech). In: Zvarova J., Preckova P. (eds.): Information Technology in Health Care, EuroMISE s.r.o., Praha, 2004: 52-9.

15. Ministry of Health of the Czech Republic, Data Standard of the Ministry of Health of CR and National Classifica-tion of Laboratory Items, 2004, http://www.mzcr.cz/index.php?kategorie=31.

16. Subrt D., Raska J., Bures V.: Structuring of Information in the WinMedicalc Hospital System (in Czech). In: Zvarova J., Preckova P. (eds.): Information Technology in Health Care, Eu-roMISE s.r.o., Praha, 2004: 33-51.

Correspondence to: Prof Petra Preckova, Phd, EuroMiSE Centre, department of Medical informatics, institute of Computer Science aS Cr. Pod Vodarenskou vezi, 2, 182 07 Prague 8, Czech repub-lic, e-mail: [email protected]. phone: +420 266 053 620,fax:

+420 286 581 453


Papers

Professional paperSummaryModern education supported by informational tech-nologies (it), represents revolutional transformation in further educations as well as bussines that follows modern education. the aim of this research is to prove the exsistence of close connection along with it’s quality, which is presented in correlation scheme. the coeficient significant testings of it correlations are providing efficient management and reengener-ing and new organisation of university functions, along with modern information systems brings and promotes the quality of education together with eco-nomic development of country.(1,2,3,4)Key words: information technologies, Education, Quality, Uni-versity.

1. introductionIn turbulent electronic environment, settings up of strate-

gy for further reform development of education by using mo-dern information technologies, presents mayor issue. We cam freely say that the Internet is not just a new exstension of uni-versity bussines, but Internet applications must be connected with standard ones which are already being used by university or other educational institution in its work. Informational te-chnologies and internet are making possible the egzistence of better connection of university /faculty with student and envi-ronment. (5,6,7).

2. Metodology of researchResearch study is performed on five university samples ((Sa-

rajevo, Banja Luka, Mostar, Mostar Sv., Tuzla), on 30 testers.Methodology of systematic approach and systematically analysis Statistics methods, sample firstly, questioner as some of information in empirical research. Equipment and tools that were used in work are: PC Pen-tium IV - Internet and its protocols , MS Office; MS EX-CEL; MS WORD; MS ACCESS, SPSS 11.0 statistical soft-ware.

3. results In research results suitable set of variables, wich were cro-

ssed and whose tables and figures are shown in further text, were used.

Structure and number of employees on B&H UniversityAccording to engagement of those employed in process of

teaching, and outside, the structure of employe engagement on

•

•

•

observed universities is as follows:tablE 1. Structure and number of employees

rb UNiVErSitY

total the number of employed personal

in teaching

Percentage (%)

outside teaching (administration)

Percentage (%)

1UNiVErSitY oF SaraJEVo

2533 1640 64.7 893 35.3

2UNiVErSitY oF baNJa lUKa

726 491 67.6 235 32.4

3UNiVErSitY oF tUZla

655 430 65.6 210 32.1

4UNiVErSitY oF MoStar

134 111 82,8 23 17,2

5UNiVErSitY (SV.) oF MoStar

800 700 87.5 100 12.5

From this table, it is clear, that the number of employsed personal on these universities in B&H is 3372 or in average 65 persent of total number of employees. (7, 8).

When, matching the number of employees on universities in BiH the following graphic ilustrated:

The organization

tablE 2 Grading marking the organization on B&H universities Grade Percent Valid Percent

1 - insufficient 6,7 7,1

2 - sufficient 20,0 21,4

3 - good 20,0 21,4

4- very good 26,7 28,6

5 - excellent 20,0 21,4

total 93,3 100,0

Unexpressed opinions 6,7

From this tabel is clerly visiable that the most of testers has evaluated theirs organizational system with grade very good

information technologies and Quality of Menagement decision in University EducationMensura Kudumovic1, Stevica Krsmanovic2, dzafer Kudumovic3

Faculty of Medicine University of Sarajevo, b&H1, Faculty of Economy University of belgrade, SCG2, Faculty of Mechanical University of tuzla b&H3

NENASTAVNASTAVNIUPOSLENI

Mea

n

700

600

500

400

300

200

100

FiGUrE 1. The outlook of number and structure of employsyeyes

AIM 2005, 13(4): 206-207



26,7 percent, while 20 percent of testers has evaluated the orga-nizational system with grade very good and excellent.

Evaluation of IStablE 3. Evaluation of IS on B&H universities Grade Percent Valid Percent



3 - good 46,7 50,0

4- very good 6,7 7,1

5 - excellent 13,3 14,3

total 93,3 100,0

Unexpressed opinions 6,7

46,7 percent has evluated its IS with grade good, while 20 percent of tested theirs IS evaluated with grade insufficient- 1, and 13,3 percent with grade excellent-5 .

IT usage tablE 4. The number of computers

title Mean Std. deviation

average number of computers on b&H universitiues

274,6000 339,31048

tablE 5. Grading of IT usage Grade Percent Valid Percent



3 - good 37,5 37,5

4- very good 12,5 12,5

total 100,0 100,0

From the standing of usage –by usage of information tech-nologies the results of grading have been set up from the lo-mest level insufficient -1 to the higest excelent -5. Grades su-fficient -2, good -3 were accomplished by the largest number of testers 37,5 percent while the highest grade was excellent -5 and was not accomplished by any of tetsters. The results of grading IT usage in deciding are: 25 percent – grade insuffici-ent, 37,5 percent grade sufficient, 25 percent grade good and 12,5 percent grade excellent. Grading of informations quality necessary in making decision is: 12,5 percent grade insuffici-ent, and sufficient, 37,5 percent grade good, 25 percent grade very good, 12,5 percent grade excellent.

Correlatoins tablE 6. Correlation between the usage of IT in teaching and the number of employees

Usage of it Employees

it usage Pearson Correlation 1 -0,770

Sig. (2-tailed) , 0,043

N 8 7

Employees Pearson Correlation -0770 1

Sig. (2-tailed) 0,043 ,

* Correlation is significant Within the 5 percent validity level, border line t with 65 per-

cent of freedom equals 2. Our t is mark ably higher and so that correlation is statistically significant.tablE 7. Correlation between the usage of IT and and quality of information

Usage of it. Q.iNForMUsage of it Pearson Correlation 1 0,681 Sig. (2-tailed) , 0,063 N 8 8Q.iNForM Pearson Correlation 0,681 1 Sig. (2-tailed) 0,063 ,

** Correlation is significant Statistical processing of data and analysing show that the-

re is mesaurable statistical rellation on IT influence on grading of quality and quality of information needed in process of ma-king manager decisions as well as other evaluations. Results of testing show that averagely, universities that have more compu-ters along with computer network , show better evaluation of IT usage as well as quality of information in those universities which don’t have network or have less computers.

r = + 0.66 0< r < 1Correlation, between IT usage and quality of collected in-

formation needed in process of making manager decisions is higher within universities – with higher degree of usage of in-formational technologies than with those with lower degree. Calculated t equals:

t=16.64Within significant level of 5 % border line value t with 65

degrees of liberty equals 2.00. The t we got is higher, and there-fore we conclude that correlation we get is statisticaly signifi-cant.

4. Conclusion Statistical processing and analysis show that there is mea-

surable statistical connection on IT influence on degree of evaluation of quality and quality of information needed in pro-cess of making manager decisions and other evaluations.

Research results show that in average universities that have more computers and computer network express higher degree of IT usage as well as quality of information than universities with less computers and without computer network.Universi-ties and other educational institutions are shedding toward us-age of electronic advancement and the goal is to efficientlyy and effectivelyy improve education and bussines.

Organizations are changing their bussines processes in or-der to increase quality which is unavoidable for highly educa-tional institutions and organizations. Usage of TQM or other concept (six sigma) provides increasing of responsibility of all members of organization employees.There are all expected on their income for total improvement of qualityy (from profes-sor, who efficientlyy and correctlyy teach to secretaryy or oth-er administrative official). to avoid mistakes.

Many studies show that in circular process of eliminating of errors and disadventages, expences are also decreased, which means that quality changes are decreasing expences apart from improving quality of education. (1,6)

REfEREnCEs1. Laudon K, Laudon J. Management Information Systems, Upper

Sadle River, NJ: Prentice Hall, 2000: 350-51.2. Mašić I, Ramić A, Bešlagić Z. Obrazovni proces putem modernih

tehnologija. AIM, 1997; 5(3-4): 59-61.3. Kalterborn KF. Virtuelna realnost – primjene, mogućnosti i gran-

ice nove kompjuterske tehnologije. AIM, 1997; 5(3-4): 87-94.4. Masic I, Catak-Ramic A, Kudumovic M, Pasic E. Distance learn-

ing in the medical education in B&H: E-Health & E-Education Proceedings, Zagreb, 2002: 17.

5. Kudumović M, Masic I, Novo A, “Menadžment informacioni sistemi”, MAT SOC MED, Sarajevo, 2002; 14 (1-2): 1-64.

6. http://users.efpu.hr; 05/07/2005.7. http://unsa.ba/e-lerning ; 17/08/2005.8. http://www.unsa.ba; 27/08/2005.

Correspondence to: Mr Sci Mensura Kudumovic, Medical faculty of University of Sarajevo, phone: 00 387 33 444 714,

e-mail: [email protected]


Papers

Professional paperSummaryin this essay are presented various deformities and developmental malformations in the locomotor sys-tem, which are noted in one year research study of Clinic for ortopedics-traumatology of the Clinical Center of the University of Sarajevo, pediatric ambu-lance (01.01.2004-31.12.2004.) the main goal of the research was to establish incidence of the most fre-quent malformations, as well as to show importance of early detection of each classified malformation and its impact on the final treatment results. during the year 2004 (01.01.-31.12.2004.), on the pediatric am-bulatory Clinic for ortopedics & traumatology, KCUS (Kot), was examined total of 5820 patients, of which 347 (5.9%) patients were for the first time diagnosed with already established congenital malformation in the locomotor system. ratio between boys and girls was 215:132 (347). developmental dislocations of the hip (ddH) were dominating and were marked as fre-quency i. Second dominating were scoliosis with 25.8 %, marked as frequency ii. there were also congeni-tal deformities of the foot 10.9%, marked as frequen-cy iii; congenital torticolisis with 9.7%, marked as frequency iV; genua crura vara 4.0%, marked as fre-quency V; and coxa valga 0.8% marked as frequency Vi. indications for immediate hospital treatment were presented in 24% patients with developmental dislo-cation of the hip, 14.4% scoliosis, 44.7% congenital malformation of the foot, and 14.7% torticolisis.Key words: classifications, developmental malfor-mations, locomotor system.

1. introductionThe most often occurring deformities and developmental

malformations in the locomotor system are:Developmental dislocation of the hip (DDH), also called congenital hip, in all phases of development is marked as decentralization of this joint. Etiopathogenesis was unex-plained, but everyone is in agreement that this is multifac-torial inheritance disease and that there exist many theo-ries about this subject (1, 2, 3):

Antropological theory by Le Damany: acetabulum is shallower at birth than during fetal time, as well as that the head of femur is only 1/3 covered. Full coverage is possible only in the third year.Theory of inheritance by Ruth Wayne-Davies shows us

•

•

•

that displacement of acetabulum is primary genetic dis-ease with polygenetic heritage, and it is responsible for late diagnosed DDHTheory of developmental abnormality states that dislo-cations of the hip develop in later pe-riod, in the peri-od during shaping of the hip. It is of importance to men-tion postnatal peri-od of the develop-mental dislocation of the hip since that is the best time for prevention and treatment. If it is detected early, there exists the possibility towards the normal development.Dualistic theory explains influence of mechanical and embryonic factors (pubic disposition of the fetus with flexed hid and hyperextension of the knees).Hormones therapy shows relationship between abnor-malities in the hormonal metabolism of the mother, in-creased secretion of estrogene and estradiol, which re-sults in looseness of the hip.

2. risk factorsThere exist family anamneses. Predispositions to dislo-cation of the hip have 12% children if one of the parents had dislocation while in case when one of the parents and

•

•

•

•

Classifications of deformities and developmental Malformations in the locomotor SystemZoran HadziahmetovicCenter for Emergency medicine, Clinical Center of the University of Sarajevo, bosnia and Herzegovina

FiGUrE 1. Radiographs bilateral congenital dislocation of the hip

FiGUrE 2. Congenital convex foot

AIM 2005, 13(4): 208-210



a child had dislocation there is a 36% of chance that child born after will also have dislocation.Dislocations are 4-6 times more frequent in girls than in boys.Dislocation more often occurs on the left side, and in chil-dren with a risk factorDifference of developing DDH are greater and shorter than 0.41-168.6 (5).

There also exist congenital spinal deformities with frequency of 1%. The most represented is scoliosis making it the most studied.

Scoliosis represent complex deformations characterized with lateral curvature and rotation of the vertebra. Since it af-fect vertebra, spinous process and pedicle rotate toward con-cave curvature. Ribs are also closing on concave curvature

while they are expending on the convex side (Picture 3.). Since many organs develop concurrently with scoliosis. Scoliosis can be unified with the urinary (20-30%), cardiac (15%), pulmo-nary and other abnormalities.

Acquired scoliosis changes in structure and look of the ver-tebras are happening throughout life and over 90% are of un-known etiology. Ratio in idiopathic scoliosis between men and women is 3:1. In the thoracic segment, right convex curvatures

•

•

•

are four time more likely to occur than left. Congenital curvature can be vertebral or

extravertebral. Vertebral type can be open or closed. When open, strong “mielomeningo-celae” is often in companied with partial or complete neurological defect and paraple-gia.

Closed vertebra type of the congenital scoliosis can be etiologically shown in the following forms:partially unilateral absence of the vertebra,complete unilateral absence of the vertebra (hemivertebra),unilateral segmentation, bilateral segmentation (block vertebra)

Congenital and acquired deformations of the foot are very frequent; clinically, the most frequent are (7):pes equinovarus (inversio pedis), with the frequency of 1/100. In half of the patient, deformation is both sided,pes metetarsus varus (pes adductus), with deformation like pes talus verticalis,pes excavatus,pes planovalgus, etc.

There is also torticollis congenita (muscle and bones form), Klippel- Feil´s syndrome (congenital cervical synostosis), kyphosis congenita, congenital elevation of the scapu-la, manus vara, coxa antetorta, genu recurva-

tum, crus varum congenitum, osteogenesis imperfecta, etc.

For many congenital deformations of the upper and lower limbs, there exist effective prevention. Early diagnosis, dur-ing the first weeks after birth, gives a bigger chance of com-plete recovery without consequences in later period of life. Be-sides clinical examinations other kinds of tests are necessary, such as: radiography, CT, EM, laboratory tests, etc. Concur-rently with start of mild diagnoses and classifications of spe-

••

••

•

•

••

tYPEosseal moudl. acetabul.

osseal protuberance

Chondral roof a β Control

i a Good rectangleNarrow cover femoral head

> 60 < 55Some- times

i b Good roundnessWide cover femoral head

> 60 > 55 Yes

ii a - ossification to 6 weeks

Satisfactory round Wide cover femoral head

50 – 59 > 55 Yes

ii a – ossification to 3. months

insufficient round Wide cover femoral head

50 – 59 > 55 therapy

ii b - ossification from 3. months (dysplasia)

insufficient round Wide cover femoral head

50 - 59 > 55 therapy

ii c - ctitical hip insufficient round Wide still cover head

43 - 49 70 – 77 therapy

ii d - hip toward decentration

Poorly round / Steep Press 43 – 49 > 77 therapy

iii a - decentration hip

bad Steep Press without struxctural change

< 43 > 77 therapy

iii b - decentration hip

bad Steep Press with structural change

< 43 > 77 therapy

iV dislocation bad Steep Press < 43 > 77 therapy

tablE 1. Ultrasonography classification developmental dysplasia of the hip (DDH) - the Graf method

development malformations

No (M) No (F) total, 166 %, 47,7 Frequency, i

dysplasia 52 74 126 36,2

Subluxatio 3 2 5 1,4

dislocation 6 28 34 9,7

tablE 2. DDH, Orthopaedic clinic of Sarajevo, 01.01.2004. – 31.12.2004 (N 0 = 347)

No (M) No (F) total % Frequency ii

Scoliosis 25 65 90 25,8

tablE 3. Scoliosis, Orthopaedic clinic of Sarajevo, 01.01.2004. – 31.12.2004. (N 0 = 347)

development malformations of foot

No (M) No (F) total % Frequency iii

Pes equinovarus 16 17 33 9,4

Pes talus verticalis 0 1 1 0,2

Pes calcaneovalgus 0 1 1 0,2

arthrogriposis + equinovarus

2 0 2 o,5

aplasio pedis 1 0 1 0,2

total 19 19 38 10,9

tablE 4. Congenittal deformity of foot, Orthopaedic clinic of Sarajevo), 01.01.2004. – 31.12.2004. (N 0 = 347)

torticolis congenita No (M) No (F) total %FrequencyiV

Genua crura vara 9 5 14 4,0 V

Coxa valga 1 2 3 0.8 Vi

Manus vara 1 1 0.2 Vii

Scapula alta (Sprengel)

1 1 0,2 Vii

Multiplices vert.cervicalis

1 1 0,2 Vii

total 16 18 34 9,7

tablE 5. Rest deformity, Orthopaedic clinic of Sarajevo,, 01.01.2004. – 31.12.2004. (N 0 = 347)


Papers

cific anomalities, surgical procedure can also be recommend-ed. For example, ultrasound classification and interpretation of DDH, according to Graph, determines further preventive and therapeutic measures. (Table 1) (8).

Mild examination and ultrasound’s dynamic screening used in all developed countries of Europe, helps with early diagno-sis and treatments with a result of complete functional recov-ery from DDH in more that 90% of cases (9).

There exist many classifications of scoliosis:etiology (functional, structural, congenital, acquired)starting time (infant/toddler from 0-4 years, toddler/early adolescent 4-10 years, late adolescent after 12 years).locations (cervical vertebra, cervicothoracic vertebra, tho-racic vertebra, lumbar vertebra)

3. discussionReasons of high percents developmentally abnormality in

the locomotors system during the research DDH, and scolio-sis, we can investigate in lately hospital treatment and healing. There exists:

different attitude of family,

••

•

•

problems about earlier discovering disease in different places,long time for analyzing problem,lately and bad treatment, and bad classifications deformi-ty.

4. ConclusionBest treatment for all developmental malformations, is us-

ing treatments in screening school specially aimed for babies. So there exists possibility for successful healing. If deformity and malformation was not controlled, and disease is in pro-gressing phase, only solutions in this case is surgery method, but there is always risk and insecure.

REfEREnCEs1. Tachdjian MO. Pediatric orthopedics, 2 ed. Philadelphia,

London, Toronto, Montreal, Sydey, Tokio, Saunders co.: 1990: 297-549.

2. Vrdoljak J. Prirođeno iščašenje kuka. Pediatar Croat, 1999; 43: 15-8.

3. Tonnis D, Storch K, Ulbrich H. Results of newborn screen-ing for CDH with and without sonography and correla-tion of ris factors. J Pediatr Orthop, 1990; 10: 145-52.

4. Catteral A. The early diagnosis of congenital dislocation of the hip. JBJS, 1994; 76-b: 515-16.

5. Amico MD, Merolli A, Santmbrogio GC Three dimen-sional analysis of spinal deformities. Suudies in health technology and informatics, Amsterdam, IOS Press, 1995: 5: 15.

6. Epeldegui T. Flatfoot and forefoot deformities. Vicente, Madrid, 1999.

7. Szepsi K. Modern trends in the treatment of conegeni-tal dislocation of the hip, European Instructional Course Lecturers, 1997: 119-32.

8. Bialik V, Bialik GM, Blazer S. et all. Developmental of the hip: Anew approach to incidence. Pediatrics, 1999; 1: 93-9.

9. Bjerkreim I, Hagen OH, Ikonomou N, Kase., Kristiansen T, Arseth PH. Late diagnosis of developmenral dislocation of the hip in Norway during the years 1980-1989. J pede-iatr Orthop B, 1993; 2: 112-4.

10. Graf R. Sonographie der Sauglinshufte, Ein Kompedium, 2 auf Stuttgart, Ferdinand Enke Verlag, 1986.

Correspondence to: Prof Zoran Hadziahmetovic, Clinical Centre Sarajevo, 00 387 33 297 000.

•

••

FiGUrE 3. Thoracal/Thoracolumbal scoliosis



rEViE

WS

reviewSUMMarYinternet is one of information technologies marking the transition from the second to third millennium. the present role and expansion of internet in medi-cine and healthcare is reviewed together with the per-spective of further development. the beginning and initial expansion of the use of internet in medicine is described. the World Wide Web (WWW or Web) is rec-ognized as a major reason for this expansion, reach-ing a state described as a Web-pandemic. the rapid increase of the number of papers dealing with internet in medical literature is presented as well as the ap-pearance of several journals dedicated to internet in medicine. Specialized symposia, among them MEd-NEt world conferences, are noted. First uses of in-ternet in medicine comprised databases, discussion groups, electronic newsletters, software archives and online public access catalogues. the appearance of the Web led to a significant improvement of the inter-net use in medicine, which is reflected in an exponen-tial increase of the number of publications. it is noted that internet allows “to do old things in new ways”, but also “to do new things”. it became clear that the information revolution evoked by the internet shall leave a deep trace in medicine, as health information became accessible to the public and ceased to be in exclusive control of health professionals. New medical fields – telemedicine and cybermedicine - appeared as a result of the development and global expansion of information and communication technologies, with cybermedicine dealing more specifically with the use of internet. the advantages and disadvantages of cy-bermedicine are discussed, and major problems arise in connection with the quality of health information. Several systems for quality criteria of health relat-ed Web sites are described, and the Web sites have to conform with quality criteria such as transparen-cy and honesty, accountability, privacy and data pro-tection, currency, accountability and accessibility. a review of the development of the use of internet in medicine in Croatia is given. Worldwide it is more and more recognized that the internet is capable to create the conditions for a partnership between patients and clinicians. in the concluding remarks the reasons are given for a sooner access to the internet and its every-day use in healthcare activity.Keywords: Cybermedicine, internet, telemedicine, Web.

1. introduction The world we are living in, on the start of third century,

is marked with wide use of new information and communi-cation technologies (ICT), like mobile phones, TV teletext, telefax and e-mail. It is often said, therefore, that we are living in the so-called information society. The basis for developing and propagation of ICT is the technology of computers, pow-erfully marking the second half of the twentieth century and the start of the third millennium. In the information society, which succeeds the industrial society, people are using these new technologies in all aspect of their everyday life, at work, at home, while resting or having fun, making it clear that these new technologies are changing our lifestyles, ways of learning and free time.

The important ICT is Internet, a computer network spread all around the world, connecting millions of computers by us-ing a common protocol (1). That ICT is securing use of differ-ent group net resources named “World Wide Web” or shorter Web, of which the most important is access to computer serv-ers with the help of HTTP (HyperText Transfer Protocol) which enables use of text, graphics, sound and other kinds of records. It is appropriate to draw attention to the often quoted state-ment that the spread of Internet is comparable with the im-portance of the Gutenberg´s invention of printing technology, accelerating in the past for the order of magnitude commu-nication of information saved until then in the form of hand-written texts (2). Internet has again accelerated communica-tion of information for the order of magnitude by using new electronic technologies.

The aim of this text is to present the beginning of Internet as a new ICT in medicine, its use today as well as its use in medi-cine and healthcare worldwide, with evaluation of main direc-tions and prospects of its development.

2. Start of internet use in medicine and its spreading

The information society we are living and working in, and in which ICT enables unimaginably comprehensive access to information from around the world produced in numerous fields of human activity, is highly influencing medical science and everyday healthcare work. During past decades of the last century, the use of Internet has been spreading rapidly ow-ing to the appearance of commercial Internet service provid-

internet and Medicine: Yesterday – today – tomorrow1

Gjuro dezelic“andrija Stampar” School of Public Health, Medical School, University of Zagreb, Zagreb, Croatia

AIM 2005, 13(4): 211-217

1 This is a revised and updated English version of the paper: Deželić Gj. Internet in Medicine – Development and Perspectives. Acta Med Croatica 2002; 56, 131-143 (original in Croatian).


Papers

ers (ISP). Due to this, Internet was no longer available only for the academic and military use, but was available to business and households as well (3). This, for sure, enabled wide usage of Internet in medicine. The main reason for spreading of the Internet was the use of Web in such enormous volume, that it gave place to the statement that the “Web fever has become a pandemic” (4).

Worldwide spreading of Internet is one of the most spectac-ular events in development of the information society, shown by a very fast growth of Internet hosts (host – a computer sys-tem with registered ISP address) from the beginning of Inter-net until now, in time intervals of few years (Table 1). The “Web pandemic” is presented by the growth of number of Web sites from the appearance of Web on Internet until now (Table 2).

tablE 2. “Web pandemic” shown by number of Web sites throughout years*

Year** Number of Web sites

1993 130

1994 2.738

1995 23.500

1996 252.000

1997 1.117.259

1998 2.410.067

1999 6.177.453

2000 17.119.262

2001 29.302.656

2002 38.807.788

2003 40.936.076

2004 51.635.284

*Information from Hobbes Internet Timeline (5)**Data is taken from the month of June each year

In medical sphere, the “Web pandemic” can be easily fol-lowed through the number of published articles. If we take the bibliographic database Medline as the main source of medical literature, by searching the database we get numbers shown in Table 3. The increase of the Internet importance in medicine is reflected by the change of descriptors (MeSH terms) used in Medline for finding articles linked with Internet. In 1991, just one article could be found, and the MeSH term “Internet” in the “Cumulated Index Medicus” was a part of the term “Com-puter Communication Networks” introduced in the same year. In 1999, “Internet” became an independent term with sub-terms “Economics”, “Instrumentation”, “Legislation & Jurispru-dence”, “Organization & Administration”, “Standards”, “Statis-tics & Numerical Data”, “Trends” and “Utilization”. Today, new sub-terms, “Classification”, “Ethics”, “History” and “Supply &

Distribution” are introduced.From the stabilization of published articles in Medline in

last years, and from their decrease in some of the years, it can be concluded that the maximum of pandemic was at the transi-tion of centuries. This leads to the assumption that publishing of articles on the phenomenon called Internet has reached its saturation point, as Internet became an everyday notion and an indispensable tool in scientific and professional medical work. This effect was already predicted previously (3).

3. internet in professional medical circles and in medical literature

Medical aspects of Internet belong to medical informatics as one of the youngest medical disciplines. This follows, after all, from the definition of medical informatics (6,7) regard-ed as a scientific discipline dealing with theory and practice of information processes in medicine, comprising data com-munication by ICT. Also in Medline, Internet is regarded as a part of medical informatics, and that to its computing part. At medical informatics scientific conferences, e.g. at internation-al MEDINFO and MIE congresses, from the middle of eighties and into the nineties of the last century the number of papers on computer networks and Internet increased. In the USA, the entire annual conference of the American Medical Informatics Association (AMIA) in 1996 was devoted to Internet in medi-cine with the motto “Beyond the Superhighway – Exploiting the Internet with Medical Informatics”. AMIA also devoted their next conference (1997) to Internet, with the motto “The Emer-gence of ‘Internetable’ Health Care – Systems that Really Work”. Due to high interest in the Internet use in medicine, the Soci-ety of the Internet in Medicine (SIM) was established in Europe. Every November, SIM is organizing congresses on Internet in medicine named MEDNET. The success of the first Europe-an MEDNET congress in Brighton, Great Britain, in 1996 pro-voked the next MEDNET, organized next year also in Brighton, to be marked as a “world congress”. A series MEDNET con-gresses followed: 3rd in London, Great Britain (1998), 4th in Heidelberg, Germany (1999 – Motto “Toward the Millennium in Cybermedicine”), 5th in Bruxelles, Belgium (2000 – Motto: “Real world medical applications”), 6th in Udine, Italy (2001 – Motto: “Medicine beyond the Boundaries”), 7th in Amsterdam, The Netherlands (2002 – Motto: “Qualit-e-Health”), 8th in Ge-

Year** Number of computers

1970 9

1974 62

1981 213

1985 1961

1990 313.000

1995 4.852.000

2000 93.047.785

2002 162.128.493

2004 285.139.107

tablE 1. The growth of Internet shown by number of hosts depending on time*

*Information from Hobbes Internet Timeline (5)**Data is taken closer to the middle of the year

Year Number of published articles

1991 1

1992 8

1993 32

1994 95

1995 225

1996 463

1997 706

1998 1.324

1999 2.408

2000 3.138

2001 3.052

2002 2.898

2003 2.921

2004 3.203

tablE 3. Number of published articles on Internet in bibliographic database Medline*

**Search in Medline/Pub Med (http://www.ncbi.nlm.nih.gov/en-trez/query.fcgi); MeSH term “Internet”



neva, Switzerland (2003 – Motto: “Internet in Health for All”), and 9th in Buenos Aires, Argentina (2004). The 10th MED-NET 2005 will be held from December 4-7, 2005 in Prague, Czech Republic, and is going to have four Mottos:

Internet for informaticians and clinicians: “Towards the new web technologies”,Internet for teachers: “Towards effective use of new e-learn-ing tools”,Internet for librarians: “Towards the new roles of medical libraries”,Internet for Web service providers: “Toward the new ways in Web presentation”.

Several years ago journals devoted to Internet in medi-cine, registered in Medline database (http://www.ncbi.nlm.gov/entrez/query.fcgi?CMD=search&DB=journals) started being published. The first journal, on the borderline to med-icine, was “Cybertechnology & behavior – the impact of the In-ternet, multimedia, and virtual reality on behavior and society” (8), starting in 1998. In the field of medicine and healthcare since 1999 publication of journals “Internet Healthcare Strate-gies” and “Journal of Medical Internet Research” (JMIR) started (10). To the journals containing the word “Internet” in their ti-tle belongs also “Medical Informatics and the Internet in Medi-cine” (11), a known English medical informatics journal, from 1999 an official journal of SIM.

Different articles on Internet are traditionally published in journals specialized in the field of the medical informat-ics. Due to the great interest in many medical fields, a series of other medical journals, some of them world-known, are also sometimes publishing papers on Internet. Some of them, as e.g. “Journal of the American Medical Association” (JAMA), have sections devoted to Internet. The Italian journal “Haematologi-ca” has sections “e-cases”, “e-letters” and “e-images” which bring information on Internet. Some other journals, as e.g. “British Medical Journal” (BMJ) since May 27, 1995, weekly publish all papers in full text on Internet. BMJ’s Web site contains text of all papers published since January 1994) (12). Medline’s data-base also contains many full text papers from various journals. There are also builletins/newspapers dedicated to doctors and people working in fields allied to medicine, which are publish-ing shorts news and articles, and specialized for issues on In-ternet in medicine (13). One of them is “Internet Medicine”, initially published from 1999 until 2003 by Lippincot Williams & Wilkins, and at present available on Internet as a monthly from Health Care Publications.

4. Short review of medical use of internet until the end of 20th century

From the very beginning, Internet was interesting to med-ical experts, especially due to possibilities of sending e-mails and accessing remote computers (remote login) on which it was possible to search databases, with the biggest interest in catalogs of medical libraries and bibliographic databases. In that period it was also possible to transfer files from remote computers on a local computer (FTP - File Transfer Protocol). Those files were documents in form of text, picture as well as computer programs ready for use. Since the number of person-al computer (PC) users was increasing as well as spreading, ac-cess to Internet became wider as well.

The appearance of Web induced thoughts about the possi-bility of significant promotion of world public health by using ICT. In 1994, a group of medical experts from several Amer-ican and Japanese universities, the World Health Organiza-tion, state government bodies for information infrastructure

•

•

•

•

and development, and U.S. corporations, lead by Ron LaPorte, formed their views regarding the future of information man-agement in public health (14). They had the opinion that the present technology can be used for creating global infrastruc-ture for health information, capable to improve health all over the world. They proposed, as a first step, the linking of pub-lic health institutions, health ministries and departments, the WHO, and other organizations of the UN, and called profes-sional medical and public health societies to recommend their members electronic linking. Some of other proposed possibil-ities were global disease telemonitoring, environmental mon-itoring, distance education and publishing of the electronic journals.

Previously mentioned AMIA 1996 Conference provoked an unexpectedly big interest of medical informaticians, as well as other medical experts already using Internet, which result-ed in more than 330 published articles and summaries pub-lished in the conference proceedings (15). The fast growth of cited works found in Medline was noticed, and as a positive sign first literature reviews, aimed to get medical informati-cians and also a broader medical community acquainted with Internet, were also noticed. That enabled James J. Cimino, by analyzing conference papers, to recognize three ways of Inter-net use in medicine: 1) “Using the Internet to do Old Things in New Ways”, 2) “Using the Internet to do New Things”, and 3) “Contributions Beyond Medical Informatics”.

In the group under (1) it was recognized that many med-ical informatics applications, demonstrating interesting tech-nologies and applications, although available on-line, could not reach potential users due to “simplistic, awkward, charac-ter-based user interfaces, due to technical difficulties with pro-viding distributed graphical user interfaces, due to technical difficulties with providing distributed graphical user interfaces and the ‘lowest common denominator’ of hardware among us-ers” (15). Only the appearance of Web technologies and start of usage of the Web search engines based on the HTML (Hy-pertext Markup Language) language, with multimedia possibil-ities (color pictures, static or moving video, sound effects and later music), provoked the interest of wider circles of potential users. It became possible for everyone to access big numbers databases with bibliographic and other medical information. That led to the nice statement that the Internet environment and Web-paradigm can improve already exiting resources and, for example, improve Medline (“make a better-Medline-than-Medline”) (16). Systems for evaluation of the health status, for medical decision-making with a series of expert systems, and systems for clinical support became available in direct access. Applications in education, improving communication between institutions, teachers and students as well as development of advanced teaching methods and first attempts of distant learn-ing were reported. First papers on approaching clinical infor-mation, with warnings on problems about security and confi-dence, started to appear.

As new possibilities under (2) Cimino mentioned electron-ic publishing, development of new graphic methods for data presentation and direct follow-up in real-time, what was rec-ognized as especially important for clinical practice. Inter-net enabled development of new methods for collaboration in clinical consultations and discussions, as well as for electronic communication between patients and physicians.

And finally, medical informatics research reported in pa-pers under (3) led to solutions applicable in non-medical fields as well.

By analyzing the papers presented at the AMIA´96 Con-ference Cimino remembered that former U.S. President Harry


Papers

Truman once described nuclear energy as “a new force too rev-olutionary to consider in the framework of old ideas” and add-ed that “a half century later we are confronted with the Inter-net – another revolutionary force. Although it can be used to support conventional tasks, it also offers new ways to carry out entirely new tasks for those with sufficient vision.” (15) Cimi-no compared past “energetic explosion” at the beginning of nu-clear era with the “information explosion” half century later. He also said that Internet has done for computer communica-tion the same what the printing machine has done for the writ-ten word. He also predicted that the Internet is going to help in bringing the patient and physician closer together and that it is going to improve the care of patients by widening patient information and making possible integration with medical de-cision-making systems, but issued a warning concerning prob-lems with the intellectual property.

Europe also tried to follow the turbulent development of application of Internet in medicine in the USA, which result-ed in the second MEDNET being attributed as “world confer-ence”, because it succeed in gathering experts from all around the world. The best 11 papers presented at the conference were of such a high quality, that could qualify for publication in the journal Medical Informatics in a special issue (17), but also in-spired the publisher to enlarge at the beginning of 1999 the journal title to “Medical Informatics and the Internet in Medi-cine”, demonstrating the importance of this new filed of medi-cal informatics.

At the end of 20th century it become clear that the informa-tion revolution, provoked by the Internet, is going to make a profound trace in medicine, like in other fields. By the help of Internet health information, opened to the public, ceased to be in the exclusive scope of health professionals. The growth has been rapid as regarding the quantity of information, so also in the increase of the number of people who were using Inter-net, with proportionally more patients then physicians having access to the Internet. Quite early it became evident that that medical information on the Internet is very useful to the pa-tients (18). New questions started to open, e.g. about the qual-ity of information, privacy, rights of use, etc. In that way the traditional patient-doctor relation started to change, patients acquired possibility to access in decisions-making about their treatments as well to participate in active partnership with many groups of decision-makers, as physicians, politicians and researchers (19).

Although to that time only a small number of physicians in-troduced Internet to their practices, it became clear that this will change quickly. Starting from the statement that the Inter-net has been the most fundamental transformation of business and commerce since the industrial revolution, Jeffery Gruen formulated in 1999 a (somewhat exaggerated) assertion that “within five years every business will be an Internet business – or it will not be in business”, with a clear message to the physi-cians to start using Internet (20).

5. New fields of medicine developed in the era of information society

Development and global expansion of ICT caused the ap-pearance of new medical fields, for which two names started to be used: telemedicine and cybermedicine.

Telemedicine is the older term, and already from the sev-enties of the last century most of the authors understood it as “medical treatment at distance” (21). As such definition does not specify the technological basis of telemedicine, since com-munication methods can be conventional (non-computer),

from the very beginning of telemedicine development it was understood that telemedical procedures should be ICT based. From several more recent definitions of telemedicine let us quote the definition of the European Commission for Telemat-ic Programs in Healthcare, considering telemedicine as “rap-id access to shared and remote medical expertise by means of telecommunications and information technologies, no mat-ter where the patient or relevant information is located.” Of course, in telemedical procedures Internet is also applied, but in such cases its application is intended for the medical treat-ment of the patient. Conceiving the importance of Internet for telemedicine, a conference has been organized 1995 in Gene-va on the use of Internet and Web in telemedicine. There were 60 participants from 11 countries, among which were many prominent physicians, professors and researchers, as well as professionals representing WHO, the International Telecom-munication Union, the European Laboratory for Particle Phys-ics (CERN, the institution that created Web), the European Commission, etc. The main result of the conference was the formation of a new organization, Health On the Net Founda-tion (HON) (in 1996), which had a goal to “promote the effec-tive and reliable use of the new technologies for telemedicine in healthcare around the world.” (22).

Cybermedicine is a newer name, appearing in the nineties of the past century, meaning “medicine in cyberspace”.1* There-fore, cybermedicine would be medicine practiced on the Inter-net by cyberdoctors communicating with their users (patients) predominantly via e-mail. A different definition of cybermed-icine can be found in the paper of Gunther Eysenbach et al. (23), denoting it as “the science of applying Internet and global networking technologies to the area of medicine and public health, of studying the impact and implications of the Internet and of evaluating opportunities and the challeng-es for health care”. Eysenbach further states that cybermedi-cine is a new academic discipline appearing at the crossroads of medical informatics and public health. Cybermedicine is working also on “the exploration and exploitation of the Inter-net for consumer health education, patient self-support, pro-fessional medical education and research, on the evaluation of the quality of medical information on the Internet, the impact of the Internet on the patient-physician relationship and quali-ty of health care and the use of global networking for evidence-based medicine.”2** Cybermedicine is different from telemed-icine, even though telemedicine, as already mentioned, can also use Internet. Some of the authors state that telemedicine is mostly used in diagnostic procedures and curative medicine, while cybermedicine is mostly applied in preventive medicine and public health. Also, while telemedicine is often progress-ing as a result of technology development, cybermedicine is to a great extent characterized by consumer influence (23). More-

1 The “Merriam-Webster Online Dictionary” (http://www.m-w.com/cgi-bin/dictionary?book=Dictionary&va=cyber&x=11&y=8) assigns the attribute “cyber” as a combining form to the term “cybernetics” and “cyberspace”. The etymology of the term cybernetics, “the science of communication and control theory that is concerned especially with the comparative study of automatic control systems (as the nervous system and brain and mechanical-electrical communication systems)”, is from Greek κυβερνήτης pilot, governor + English –ics. The term “cyberspace” is “coined by William Gibson in his 1984 (science-fiction) novel ‘Neuromancer’; it is a futuristic computer network that people use by plugging their minds into it! The term now refers to the Internet or to the online or digital world in general.” (http://www.eweek.com/encyclopedia_term/0,2542,t=cyberspace&i=40647,00.asp)

2 URL address: http://yi.com/home/EysenbachGunther/cybermedicine.htm



over, according to Eysenbach, telemedicine is represented as an overlap between medical informatics and clinical medicine (working on diagnostics and therapy), while cybermedicine represents, as mentioned, an overlap of medical informatics and public health (24).

While studying cybermedicine problems can be noticed. Most of them are those related to the quality of information, but also to the standards in their structuring and communi-cation, as well as to their availability. Problems are mostly ex-pressed in the doctor-patient relationship. This relationship is necessarily changing (25), because patients and other health-care consumers are more and more interested in gaining de-tailed information, which can give answers to their questions about diagnostics and treatments, as well as possible alterna-tives, including preventive measures which are recommend-ed by their doctor or they wish to use themselves. One of im-portant monographs on cybermedicine is the work by Slack (26), in which he explains how cybermedicine helps patients become more informed, and in some cases assists in care of their health, but also warns that the control over patient´s health should be left to the human-doctor, while the machine-computer is a valuable collaborator, but not a competitor. Re-gardless how much cybermedicine looks fascinating to scien-tists, it also provokes resistance. In that respect an interesting example is the recent publication of Murray, Burns, See, Lai and Nasareth on interactive health communication applica-tions for people with chronic disease (27), which immediate-ly after its appearance on Internet provoked a number of news in public media all around the world, some of them being sen-sationalistic in their character, claiming with exaggeration that using information obtained from Internet my be dangerous. Although the publication has been retracted 13 days after ap-pearance because of errors in data analysis, the public media did not publish the announcement that the paper was retract-ed. So the public has not been informed that the publication, which wrongly accused cybermedicine and Internet as danger-ous sources of information, contained wrong conclusions and was even denoted as a “methodological catastrophe”, bringing uninformed readers to misunderstanding and led to malicious critics. Later Eysenbach and Kummervold (28) carefully anal-ysed all details of this affair and discussed all errors of critics and media. This event, however, shows that in the development of cybermedicine one has to watch carefully for objectivity in communicating health data and their reliability. And finally, it is necessary to call attention to the juridical aspects in the development of cybermedicine and its methods. According to Terry, professor of law, teaching healthcare and cyberspace law, Internet and cybermedicine “is undoubtedly helping to make health consumers smarter about treatments and drugs, but also there is a downside. The growing number of advice sites may cause particular concern because the giver of the advice may not be known -- and it may, in fact, be a pharmaceutical man-ufacturer pushing a person toward a particular drug” (29). At times, information is not complete or is even wrong. Since in real life medicine can bring doctors, pharmacists and produc-ers of drugs to court, Terry states that cybermedicine may re-sult in cyber-malpractice. He also warns that courts will face difficulties in sorting out cyber-malpractice cases. Noting such a significant shift in medicine and the relationship between pa-tient, doctor and pharmacist, Terry concludes that “old rules do not make sense any more and the new roles have not been defined yet” (29).

The quality of medical information and action in cyber-space became a subject of discussions early. Coiera (18) warned that the quality of information varies in wide span from most

recent instructions of leading clinical bodies up to outdat-ed and wrong recommendations. Among organizations, that early started to fight for making medical and health informa-tion reliable and trustworthy, was also HON, which formulat-ed the HON Code of Conduct for medical and health Web sites (30). These rules are self applied by Web sites, while HON only now and then checks whether a Web sites satisfies HON crite-ria. Therefore, even when a Web site points out the HON seal («HON-logo»), it is using it on its own responsibility, so the presence of this logo on the Web site does not guarantee in-formation quality, and for sure does not have to be taken as an awarded «stamp of quality». Since it was noticed that many in-vestigations showed lack in information quality, British Med-ical Journal, in 1998, warned users by publishing an editorial by Coiera (31) and an article by Eysenbach and Diepgen (32). Both authors published also results from a research about eval-uation of “cyberdocs” (33). Among the systems for rating in-formation quality on Internet one has to mention DISCERN (34), which is used to estimate the quality and reliability of in-formation in publications aiming to choose best methods of solving health problems. The system NETSCORING (35) can be used to estimate the quality of health information on the In-ternet by list of criteria divided in eight categories (credibility, content, hyperlinks, design, interactivity, quantitative aspects, ethics and accessibility).

Connected with questions of quality of health informa-tion on the Internet it is often warned about the possible harm caused by information of bad quality. Crocco, Villasis-Keev-er and Jadad (36) were researching bibliographic databases (MEDLINE, CINAHL, HealthStar, PsycINFO and EMBASE) with the goal to find articles about harmfulness on health caused by information from the Internet. They estimated that there is a low number of reports about harmful cases, which led them to the opinion that this is a consequence of low risk connected with the use of information from Internet, but also that such cases are rarely published, as by doctors themselves, so also by journal editors who refuse to publish them. Dis-cussion about (non)harmfulness of health information gath-ered from the Internet was started elsewhere (37, 38). To bet-ter study that problem, projects for studying harms caused by wrong information found on Internet were started (Database of Adverse Events Related to the Internet-DAERI) (39), as a part of the project financed by the European Union MedCER-TAIN (MedPICS Certification and Rating of Trustworthy and Assessed Health Information on the Net) (40). Later, that proj-ect was expanded towards development of a system for access to trustworthy health information from the Internet MedCIR-CLE (Collaboration for Internet Rating, Certification, Labeling and Evaluation of Health Information), which uses a separately developed language HIDDEL (Health Information Disclosure, Description and Evaluation Language). Its purpose is helping people, patients and medical experts in identifying trustwor-thy and useful information, making a network of information providers and users, taking into consideration the power of In-ternet (41).

Today, Internet is an everyday source for getting informa-tion about health for many people. Absolute numbers are still increasing. At the end of 2002 there were 606 millions peo-ple connected to the Internet, about 191 millions just in Eu-rope. Results of research in the USA in middle of 2002 showed that about 110 millions adults searched monthly for health in-formation on the Web, while just a year ago that number was about 97 millions. In the year 2005 that number was raised to 117 millions. Company Harris Interactive named them cyber-chondriacs, a word with its root in «hypochondriacs» (43). Re-


Papers

search, which included 4531 persons in Europe and the USA, showed that in 2002 the proportion of those searching for health information amounted to 32% in Europe and 43% the USA. They searched more often for information over websites sponsored by BBC, Yahoo, etc., instead of using Web sites of pharmaceutical companies, health institutions or government (44). Most of those people had some kind of chronic health problem, while some were just trying to find out how to stay healthy. Anderson (45) noted a case about a patient with ovar-ian cancer who, while searching on the Internet, found over 200 experimental therapies. Due to the information found this patient started asking for stronger therapies even though her doctors disagreed, because she was not strong enough to han-dle them. Connected with similar cases, the author agreed with a previously published opinion (46) that patients need to be taught how to care about their health, but also how to find and interpret authoritative medical and health information, and how to develop methods to protect themselves from the risk of inappropriate information and making own diagnoses. By considering the motivation of people for searching informa-tion on Internet, Powell and Clarke concluded that in the era of personal participation of users, consumer empowerment and broad dissemination of health information over Internet it is important to find out who are the consumers of health infor-mation on Internet, what are their information needs, and to understand why and how they search information, because in that way it is possible to improve the quality of information and the questions of privacy and access to information (47).

Problems of quality of information on the Internet showed need of educating users and providers of health services and in-formation. Because of that in 1997 a non-profit organization In-ternet Healthcare Coalition has been established, starting a dis-cussion on ethic at Internet, the eHealth Ethics Initiative (48). Also the European action plan “eEurope 2002”, launched sev-eral years ago, could reaffirm as especially important that all Web sites dealing with health problems satisfy the established quality criteria (49): transparency and honesty, authority, priva-cy and data protection, updating of information, accountability, responsible partnering, editorial policy, and accessibility.

6. Concluding remarksBased on everything presented in this review it has been

shown that the Internet is one of information technologies sig-nifying the transition from second into the third millennium. Today, Internet is used in all segments of life in a modern so-ciety labelled as “information society”. More and more users are using Internet as a resource for satisfying their information needs in everyday life – at home, in the school and during rest time. It is clear that the Internet era, despite all problems men-tioned, is going to affect significantly medical practice and de-velopment of the healthcare system. Although many people are justly cautious in using some of new methods developed by ap-plying modern ICT, one should not forget those (fortunately not numerous) who are too much depending on conventional methods and oppose to the broader implementation of ICT in medicine and healthcare. Nowadays attitudes prevail that the Internet has potentials to empower patients in a positive way and to develop partnership with physicians (50). It is clear that in medicine and healthcare, as in other fields of human activity, the necessity for a new way of information thinking and work-ing is increasing. Now it is advisable that everybody secures as soon as possible his access to Internet and its everyday use. In today’s information society it is obviously necessary to accept Gruen’s message (20): “Get online – or get left behind”.

REfEREnCEs1. The protocol is generally known as TCP/IP (TCP - Transmission

Control Protocol, IP – Internet Protocol); it should be mentioned that internet (written with lower case i) is established any time you connect two or more networks together (you have an inter-net - as in inter-national or inter-state); Internet (upper case I) is the vast collection of inter-connected networks that are con-nected using the TCP/IP protocols. The Internet connects tens of thousands of independent networks into a vast global inter-net and is probably the largest Wide Area Network in the world. In: Matisse´s Glossary of Internet Terms. Copyright 1994-2002. URL address: http://www.matisse.net/files/glossary.html. Date of approach to information September 19, 2005.

2. See e.g. Engelbrecht R, Surján G, McNair P. Preface. Proceed-ings XVIIth International Conference of Medical Informatics MIE2002, Budapest.

3. Briggs JS, Early GH. Internet developments and their significance for healthcare. Med Inform, 1999; 24: 149-64.

4. Kilbridge PM. E-Healthcare – Urging Providers to Embrace the Web. MD Computing, January/February, 2000; 36-9.

5. Zakon RH. Hobbes’ Internet Timeline (Copyright ©1993-2005). URL address: http://www.zakon.org/robert/Internet/timeline/

6. Deželić Gj. Zdravstvena informatika. In: Popović B, Letica S, Škrbić M (Editors). Medicinske struke. Zagreb: Jugoslavenska medicinska naklada, 1981: 226-8.

7. Deželić Đ. Medicinska informatika. Zagreb: Hrvatsko društvo za medicinsku informatiku, 1997: 5-6.

8. Medline abbreviation: Cyberpsychol Behav, Publisher: Mary Ann Liebert, SAD.

9. Medline abbreviation: Internet Healthcare Strategies, Publisher: COR Health LLC, SAD.

10. Medline abbreviation: J Med Internet Res, izlazi u SAD. URL a dress: http://www.jmir.org

11. Medline abbreviation: Med Inform Internet Med, Publisher: Tay-lor & Francis, England.

12. British Medical Journal (BMJ), URL address: http://bmj.com13. Internet Medicine, it can be ordered at Health Care Publications,

MCMC llc., Bethesda, SAD, URL adresa: http://www.mcman.com/784.htm.

14. LaPorte RE. Akazawa S. Hellmonds P. I sur. Global public health and the information superhighway – Tomorrow, the world. BMJ 1994;308:1651-2.

15. Cimino JJ. Beyond the Superhighway – Exploiting the Internet with Medical Informatics. J Am Med Inform Assoc 1997;4:279-84.

16. Detmer WM, Shortliffe EH. WebMedline: transforming Medline into a hypertext environment with links to full-text documnets. J Am Med Inform Assoc, 1996; 3: 933 (suppl.).

17. Newell J. Editorial. Med Inform, 1998; 23: 177.18. Coiera E. The Internet’s challenge to health care provision. BMJ,

1996; 312: 3-4.19. Jadad AR. Promoting partnerships: challenges for the Internet

age. BMJ, 1999; 319: 761-4.20. Gruen JM. The Physician and the Internet – Observer or Partici-

pant? MD Computing, 1999; Nov/Dec: 46-8.21. ibid. 7, p. 130.22. Health On the Net Foundation. Our beginnings. URL address:

http://www.hon.ch/Global/23. Eysenbach G. Sa ER. Diepgen TL. Shopping around the Internet

today and tomorrow: towards the millenium of cybermedicine. BMJ, 1999; 319: 1294.

24. Eysenbach G. Cybermedizin und globale Kommunikation. Pub-lic Health Forum; 2000 (1): 2-3. URL address: http://www.yi.com/home/EysenbachGunther/internet_public_health.htm

25. Anderson JG. CyberHealthcare: Reshaping the Physician-Patient



Relationship. MD Computing, 2001; Jan/Feb: 21-2.26. Slack VW. Cybermedicine: How Computing Empowers Doctors

and Patients for Better Care (Revised and Updated). Wiley Pub-lishers, 2001.

27. Murray E, Burns J, See TS, Lai R, Nazareth I. Interactive Health Communication Applications for people with chronic disease. Cochrane Database Syst Rev, 2004 Oct 18; (4): CD004274.

28. Eysenbach i Kummervold. JMIR, Vol. 7, Issue 2, Article e21, 2005, Editorial, “Is Cybermedicine Killing You?” - The Story of a Cochrane Disaster.

29. Saint Louis University, St. Louis, Missouri. Nicolas Terry - Pro-fessor at School of Law Issues Warnings About ‘Cybermedicine’. URL address: http://www.slu.edu/publications/gc/v6-3/news_11.shtml

30. Health On the Net Foundation. HON Code of Conduct (HON-code) for medical and health Web sites. URL address: http://www.hon.ch/HONcode/Conduct.html

31. Coiera E. Information epidemics, economics, and immunity on the Internet. BMJ, 1998; 317: 1469-70.

32. Eysenbach G, Diepgen TL. Towards quality management of med-ical information on the internet: evaluation, labelling, and filter-ing of information. BMJ, 1998; 317: 1496-502.

33. Eysenbach G, Diepgen TL. Evaluation of cyberdocs. Lancet, 1998; 352: 1526.

34. The DISCERN Project, funded from 1996-7 by The British Li-brary and the NHS Executive Research & Development Pro-gramme. URL address: http://discern.org.uk/.

35. Centre Hospitalier Universitaire de Rouen. NETSCORING. URL address: http://www.chu-rouen.fr/netscoring/netscoringeng.html i http://www.netscoring.com

36. Crocco AG. Villasis-Keever M. Jadad AR. Analysis of Cases of Harm Associated With Use of Health Information on the Inter-net. JAMA, 2002; 287: 2869-71.

37. Smith R. Almost no evidence exists that the internet harms health. BMJ, 2001; 323: 651.

38. Kiley R. Does the internet harm health? BMJ, 2002; 324: 238.39. Eysenbach G, Köhler C. Database of adverse events related to the

internet has been set up. BMJ, 2002; 324: 238.

40. MedCERTAIN - Database of Adverse Events Related to the Inter-net (DAERI). URL address: http://www.medcertain.org/daeri/

41. Kohler C, Darmoni SD, Mayer MA, Roth-Berghofer T, Fiene M, Eysenbach G. MedCIRCLE - The Collaboration for Internet Rat-ing, Certification, Labelling, and Evaluation of Health Informa-tion. MEDNET 2002 – 7th Annual World Conference on the In-ternet and Medicine. URL address: http://www.medcircle.org

42. Nua Internet Surveys. How many online? 2002 September. URL address: http://www.nua.com/surveys/how_many_online/index.html. Date of approach to information September 19, 2005.

43. Number of “Cyberchondriacs” / U.S. Adults Who Go Online for Health Information – Increases to Estimated 117 Million. URL address: http://www.harrisinteractive.com/news/ newsletters/healthnews/HI_HealthCareNews2005Vol5_Iss08.pdf. Date of approach to information September 19, 2005.

44. Eaton L. A third Europeans and almost half of Americans use In-ternet for health information. BMJ, 2002; 325: 989.

45. Anderson JG. Goodman KW. CyberHealthcare: case Studies in Social and Ethical Issues. New York: Springer-Verlag (in press). Quoted according to 40.

46. Lindberg JD. Providing reliable medical information to the public – caveat lector. JAMA, 1989; 262: 945-6.

47. Powell J, Clarke A. The WWW of the World Wide Web: Who, What, and Why? Journal of Medical Internet Research 2002;4(1):e4. URL address: http://www.jmir.org/2002/1/e4/

48. Internet Healthcare Coalition. URL address: http://www.ihealth-coalition.org/

49. Commission of the European Communities: eEurope 2002 - Quality Criteria for Health related Websites. Journal of Medical Internet Research 2002;4(3):e15. URL address: http://www.jmir.org/2002/3/e15/

50. Bauer KA. Using the Internet to empower patients and to develop partnerships with clinicians. World Hosp Health Serv, 2002; 38: 2-10. (abstract available in PubMed/Medline database).

Correspondence to: Prof Gjuro dezelic, ilica 56, 41000 Zagreb, Croatia, E-mail: [email protected]


Papers

Professional paperSUMMarYFindings during the surgical procedures confirmed the preoperative ones which indicated excellent sen-sitivity and value of the ultrasound diagnostic meth-od.based on our experience and obtained results it can be concluded that the ultrasound is irreplaceable diagnostic method, and the method of choice in dif-ferential diagnosis of scrotal trauma.Not only that it can determine the further treatment (surgical or oth-er) but also cal help to avoid unnecessary surgical in-terventions and support final plan of surgery.Key words: scrotum traumas, ultrasound examina-tion.

1. introductionAnatomic connection and external position of scrotum and

scrotal organs represents main reasons for the rarely isolated injury of testicles which is usually combined with the lesion of other structures in the scrotum (1,2,3,4).

According to the causal mechanism injuries can be divid-ed into:

blunt scrotal injuries – which usually occurs in case of kick by foot, in sport, horsemanship, driving a bike, fall. They are of contusion type and manifested most often by hae-matomas.open scrotal injuries – which are rare and occurs due to different causes: use of knife, fire and explosive arms, bru-tal force traction, bites by domestic and wild animals. It is often combined with the injury of urethra, testicles and epididymis.

The injuries can be inflicted by accident or intentionallly and they are often conjoined with polytrauma.Extent of the in-jury depends on cause mechanism and intensity of the force

•

•

that is causing them. The frequency if less than 2 % of all trau-mas. More frequently the right testicle is injured due to certain anatomical relations.

2. GoalIn this presentation we want to present our experiences as

well as the possibilities of ultrasound in diagnosis and differ-ential diagnosis of scrotal trauma. Adequate diagnostic proce-dures can provide significant decrease in number of surgical procedures and shortened the length of treatment.

3. Matherial and methodsPatients that was after clinical assessments are referred from

the Center for Urgent Medicine and Clinic for Urology to the ultrasound examination were included in the study. A one year

Value of Ultrasound Examination in Case of Scrotum traumasEdin Herceglija, Fahrudin Smajlovic, Faruk dalagijainstitute of radiology, Clinical Center of the University of Sarajevo, bosnia and Herzegovina

iMaGE 1. Ultrasonograph tescital edema – peritesticular discrete zone of free fluid longitudinal

iMaGE 2. Ultrasonograph contusion of testical tissue longitudinal

iMaGE 3. Ultrasonograph rupture of testicle free fluid CaSE r

EPortS

AIM 2005, 13(4): 218-219



retrospective study is done on 76 patients with scrotal trauma in the time period from January 1st 2003. untill January 1st

2004., from 10 up to 50 years of age. During the examinations we have used conventional ultrasound as well as color Dop-pler analyses with following equipment Toshiba – linear probe 8 MHz, ESSAOTE AU550 – linear probe 10 MHz and Philips HDL4000 – linear probe 7,5 MHz. In individual cases due to large extent of injuries we also used a sector probes.

4. resultsAmong 76 patients referred as scrotal trauma, 54 cases had

blunt trauma which was visualized by ultrasound, and 12 cases

iMaGE 4. Ultrasonograph acute scrotom (changed echo structure of the testicles, minimal flow – Doppler analysis) torsion of morgany hydatide hydrocele, epidydimis, funiculitis

iMaGE 5. Torsion of the testicle - Doppler analysis

of open trauma.In 10 cases there was no visualization of struc-tural changes.In 40 patients we have identify edema, among which in 18 cases also intratesticular haematoma is visualized (sub capsular haematoma and contusion of tissue; in 10 case haematoma of scrotal sack, in 7 cases rupture of testicle with haemorrhage in surrounding structures, and in 8 cases ruptu-re of tunica albuginea.

In case of open scrotal trauma 11 patients had rupture of te-sticle and 5 epididymis.18 patients underwent surgical treat-ment in case of rupture of testicle, tunica albuginea and hae-matocoele.By ultrasound examination as well as Doppler analysis the following posttraumatic changes in scrotum cah be visualized.

5. ConclusionUltrasound is the method of choice in diagnostics and dif-

ferential diagnostics of scrotal trauma. Main adventage of this method is high sensitivity in detection of pathomorphologic and haemodinamic changes in the scrotum. Ultrasound is a relatively inexpensive noninvasive method whic allows real - time imaging. Modern technology 3D, 4D, US and power Dop-pler offers a view of entire tissue volume thus making picture interpretation and anatomical orientation easier, and enhances objectivity and accuracy of a diagnosis.

REfEREnCEs1. Lewis AG, Bukowski TP, Jarvis PD, Wacksman J, Sheldon

CA. Evaluation of acute scrotum in the emergency depart-ment. J Pediatr Surg, 1995; 30: 277-82.

2. Al Mufti Ra, Ogedegbe AK, Lafferty K. The use of Dop-pler ultrasound in the clinical management of acute testic-ular pain. Br J Urol, 1995; 76: 625-7.

3. Gordon LM, Stein SM, Ralls PW. Traumatic epididymitis: evaluation with color Doppler sonography. A Am J Roent-genol, 1996; 166: 1323-5.

4. Fournier GR, Jr, Laing FC, McAninch JW. Scrotal ultraso-nography and the management of testicular trauma. Urol Clin North Am, 1989 May; 16(2): 377-85.

5. Munter DW, Faleski EJ. Blunt scrotal trauma: emergen-cy department evaluation and management. Am J Emerg Med, 1989 Mar; 7(2): 227-34.

6. Dewire DM, Begun FP, Lawson RK, Fitzgerald S, Foley WD. Color Doppler ultrasonography in the evaluation of the acute scrotum. J Urol, 1992; 147: 89-91.

Correspondence to: dr Edin Herceglija. institute of radiology, Clinical Center of Sarajevo, bolnicka 25, phone: 00 387 33 297 000.


Papers

Case reportSUMMarYPleural effusion - liquidothorax1 (fluid collection in pleural cavity) after cardio surgery is not rare. it is connected to serouse secretion from traumatized sur-face of parietal pleura, endothoracic fascia, rib peri-ost, after mammary artery harvesting, imperfect haemostasis and immobility of patient particular-ly breathe immobility. Suppuration of the pleural ef-fusion is extremely rare complication and is related to patient’s co-morbidity. We are going to present a case of 57 years old male patient, insulin dependant diabetic, with polivascular diabetic complication who is operated on at our clinic because of aortic steno-sis, and double vessel coronary disease. Patient un-derwent biological aortic valve replacement and dou-ble coronary artery bypasses. after hospital discharge in a good general and local condition his statement at home has been impaired progressively, particularly respiratory function. after undertaken diagnostic pro-cedures, conventional and digital we noticed a huge liquid collection in the left hemithorax and complete sternal instability. Performing surgery we evacuated 1litre of pus from empyema saccatum and mediasti-num, refixate sternum with continuous irrigation and drainage of mediastinal and pleural cavity with con-tinuous postoperative irrigation throughout 4 drains. Preoperative digital diagnostic tools, good timing of surgery and supportive medication have had good fi-nal results.Key words: Cardiac surgery, digital techniques, aseptic pleural empyema.

1. introductionPostoperative Liquidothorax (fluid collection in pleural

cavity) is not rare complication in cardiac surgery (1). It hap-pens because of parietal pleura end endothoracic fascia trauma as well the periosteum after mammary artery dissection, im-perfect haemostasis, patient incompliant for respiratory phys-iotherapy postoperatively, avoiding early getting out of bed and in patients with different systemic comorbidities (2,3,4,5). Se-rous and sero-hemorrhagic liquidothorax take place usually in range of 40 postoperative days (6). Heart failure may contribute to amount of pleural effusion in left pleural cavity but not to its primary development on this pleural side (7). Active therapeu-

tic approach, pleurocentesis and pleural drainage are rare re-quested, because of positive therapeutic response to appropri-ate doses of diuretics and nonsteroid anti-inflammatory drugs. Huge pleural effusions with protracted coughing, jeopardized respiratory function may impact to sternal non union and in-stability (3,4,5). Suppuration of the pleural effusion and devel-oping empyema that progress to the mediastinum is extremely rare complication and is possible in imunocompromised pa-tients with advanced systemic disease. In the case of patients with implanted valve, clinical condition can go to the disaster because of development valve endocarditis and complication of it: heart failure, septic embolisation, valve dehiscence, ab-scess and fistula forming or aortal rupture, etc. In forgoing text we are going to present the patient with coronary artery disease (CAD), aortic valve disease (AVD), preoperatively implanted permanent pacemaker, insulin dependant diabetic, with di-abetic complications. The patient had been operated on get-ting double coronary artery bypass grafting and aortic valve replacement. Post operatively after discharge, saccate pleural effusion has been developed of the left pleural space progress-ing to mediastinitis and sternal dehiscence.

2. Case report Male patient, 57 years, admitted with aortal valve stenosis

and insufficiency, unstable angina pectoris, Left Anterior De-scending artery proximal subocclusion and Circumflex artery medial segment stenosis, and planed for surgery correction of aortic vitium and double bypasses.

Heart ejection fraction was 35%, 16 years ago implanted VVI pacemaker because of complete AV block , he was insulin dependent diabetic patient with intermittent claudication with 100 meter of walking distance, he has microangiopata and tro-fic skin.

We implanted biological valve BIOCOR-A-23 as aortic re-placement, and create two aortocoronary bypasses, one with Saphenouse vein graft on first marginal branch from Circum-flex artery, second with left Mamarial artery graft on Left An-terior Descending artery. We found out severe calcified aortic valve, multisegmental atherosclerotic coronary artery disease, dilated heart with impaired contractility.

Postoperative period was without serious complication with a short period of serous wound secretion that healed primar-ily. After mobilization and switching to per-oral anticoagulant therapy we discharge him in good local and general condition.

Left costodiaphragmatic sinus was shadowed by small

digital techniques in diagnostic of aseptic Pleural Empyema in Patient after Cardiac SurgerySanko Pandur1, Mehmed Kulic1, Haris Vila1, omer Perva1, Safet Guska2, Nermin Granov1, Michele Musci3, bedrudin banjanovic1

Cardiac Surgery Clinic, University Hospital of Sarajevo, bosnia and Herzegovina1

thoracic Surgery Clinic, University Hospital of Sarajevo, bosnia and Herzegovina2

deutches Herz Zentrum berlin3

AIM 2005, 13(4): 220-222



amount of effusion without need for aggressive treatment. Ac-cording to general and local clinical condition we administered antibiotic protection with Ciprofloxacin tbl. 500mg TID with other needed therapy for period up to the first clinical check up in 10 days.

Patient came to first check up in a very bad clinical condi-tion, dispnoic, subfebrile, liver was prominent below ribs for 4

centimeter, pretibial swelling, absent breathing sound in two third of the left thoracic space, signs of complete sternal in-stability, maintained continuity of post operative scar without wound secretion.

Lab test showed: blood glucose 25 mmol/l, INR 9,5, BUN 16,2 µmol/l, WBC 15,0, anemia, total protein 52g/l albumino-globulin inversion. On chest X-ray (postero-anterior and profil projection) we found huge pleural effusion that reach the third intercostal space jeopardizing respiratory function.

After performed chest X-ray we decided to make pleuro-centesis. During decision making process we consult cardiol-ogist in term of ultra sound lead pleurocentesis. We noticed huge effusion in which fibrin network floats and make septa-tion of effusion. Digital ultrasound technique helped us to un-derstand why pleurocentesis have failed.

By Pleurocentesis we get a small amount of blurred sero-haemorrhagic content. Patient was readmitted, we adminis-tered diuretics, antibiotics, ACE inhibitors, K vitamin and in-terrupted anticoagulant therapy (INR 9,5), glicemia has been

controlled. Respi-ratory function was improved a little, but temperature raised as well as(White blood count) WBC count. Preoperative CT scan of thorax confirmed all men-tioned diagnoses, having that in mind we hurried with an operative interven-tion.

Because of ster-num instability ac-ceptable opera-tive approach was resternotimic one.

We chose this approach because of two goals: exploration of the left pleural cavity and sternal refixation. Surgery has per-formed in general anes-thesia with Carlen’s endo-tracheal tube. After skin incision we have gray-radish pus going out un-der huge pressure, swab for microbiology was tak-en at three levels: skin, sternum, pleura. Sternum dehiscence noticed with multifragmentar facture of the left half of sternum. Anterior surface of the heart was covered by old granulations, left pleu-ral cavity was full filled by septated purulent con-tent. Manual lyses were performed with irrigation of pleura and mediastinum.

We put two drains into left pleural cavity ant two drains into the mediastinum. With another four drains in pleura and me-diastinum irrigation fluid has been instilled. Operation team

PiCtUrE 1. Huge left pleural effusion up to third intercostals space

PiCtUrE 2. The same condition presented in the profile projection.

PiCtUrE 3. Huge pleural effusion septated by fibrin network

PiCtUrE 4. CT scan, Sternum dehiscence and septated effusion in left pleural space

PiCtUrE 5. Double drainage of mediastinum and double drainage of left pleural cavity with four Redon drains for instillation and irrigation fluid. Visible Robiseck way of sternal refixation


Papers

gave up from omentoplastic because of preexisting expanded median laparotomy after complicated cholecystectomy. Ster-num is refixated according to Robiseck technique, subfascia-ly two drains were applied. Fascia was sutured continuous-ly and other layers with retention suture. Continues irrigation has been done with betadine solution in normal saline for 15

days. Postoperatively antibiotic protocol with Ampicilin, Gen-

tamicin and Metronidazol was applied. Antibiogram resulted in a “sterile culture”. Repeated culture of irrigation fluid (nor-mal saline) was contaminated with Staphilococus epidermi-dis without etiological meaning. We took of successively one by one drain, and then stopped antibiotics. General and local conditions were improved. Blood glucose level has been easily controlled, WBC count was going down below 10.000, anemia was corrected and last drain was removed the fifteenth post operative day.

Improvement of cardial condition was, with ejection frac-tion of 50%. Good valvular hemodynamic parameters.

3. discussionOperative finding was a surprise and explanation for septic

aspect of patients (2,4,5). Empyema, by definition is contami-nated purulent collection by bacteria. We did not isolate caus-ative agent even the content was the pus according to organo-

leptic and histological characteristic. We consider that using of antibiotic in continuity made the content sterilized.

Our approach to antibiotic therapy in patients with im-planted valve, who have bad living conditions, possible incom-pliance, and co morbidity is strict in term of using antibiotics one month post operatively regardless of absent sings of gen-eral and local inflammation. Up to now we did not experience endocarditis of implanted valve in our patients. We are sur-prised by fact that in literature last 15 years there was not pub-lished description of post cardiosurgical empyema. Additional problem in this particular case was penetrating empyema into mediastinum and sternum dehiscence. Thanks to preoperative evaluation, superiority of digital diagnostic techniques clinical stabilization of patients, surgery in time, huge irrigation and drainage of pleural cavity and mediastinum for 15 days, nega-tive microbiological culture, cardial function improvement BG controlling, post operative course has past uneventfully. WBC fall was obvious; patient was not febrile with well controlled BG level. Successive drain removal and wound treatment got good result, and patient was discharged 19th day after the op-eration.

4. Conclusion Empyema development of pleural effusion after cardiac sur-

gery in diabetic patient is possible. Antibiotic usage ex-iuvan-tibus got good result in empyema sterilization. The operation decision made in time, after good diagnostic evaluation with sensitive diagnostic digital techniques like ultrasound and CT scanning got good final results.

REfEREnCEs1. Kollef MH. «Symptomatic pleural effusion after coronary

artery revascularisation: unsuspected pleural injury from internal mammary artery dissection»: South. Med J, 1993 May; 86(5): 585–8.

2. Vargas FS, Uezuni KK, Janete FB. «Acute pleuropulmo-nary complications detected by computed tomography following myocardial revascularization». Rev. Hosp. Clin. Fac. Med. Sao Paulo, 2002 Jul-Aug; 57 (4): 135-42.

3. Jain U, Rao TL. Radiographic pulmonary abnormalities after different types of cardiac surgery. J Cardiothorac Vasc Anesth, 1991 Dec; 5(6): 592-5.

4. Aarnoi P, Kettunen S. Pleural and pulmonary compli-cation after bilateral internal mammary artery grafting. Scand J Thoracic Cardiovascular Surg, 1991; 25(3): 175-8.

5. Bonacchi M, Prifti E. Respiratory dysfunction after cor-onary artery bypass grafting employing bilateral internal mammary arteries: the influence of intact pleura. Eur J Cardiothorac Surg, 2002 May; 21(5): 952.

6. Kollef MH. Chronic pleural effusion following coronary artery revascularization with the internal mammary ar-tery. Chest, 1990 Mar; 97(3): 750-1.

7. Vargas FS, Cukier A. Relation between plural effusion and pericardial involment after myocardial revascularization. Chest, 1994 Jun; 105(6): 1748-52.

Correspondence to: dr Sanko Pandur, Clinical Centre Sarajevo, bolnićka 25, 71000 Sarajevo, bosna i Hercegovina, Phone +387 61

21 80 08, FaX: +387 33 44 55 77

PiCtUrE 6. Satisfactory pulmonal reexpansion without pneumothorax and fluid effusion

PiCtUrE 7. Ultra sound 2D and M view of aortal biological valve



Case reportSUMMarYCoronary artery aneurysm (Caas) are rare and their management is controversial. their incidence varies from 1,5% to 5% of the coronary angiography, with predilection of the right coronary artery. Unruptured coronary aneurysm are often silent and may remain undiagnosed. the etiology can be congenital or ac-quired. We describe a case of a left anterior descend-ing artery (lad) aneurysm treated with an off-pump surgical revascularization with a liMa to lad without exclusion or ligature of the aneurysm.Keywords: Coronary aneurysm, off-pump CabG.

1. introductionCAAs are defined by the Coronary Artery Surgery Study

as a coronary dilatation that exceeds more than 1.5 times the diameter of normal adjacent segments, or the diameter of the patient’s largest coronary vessel (1). The first case of coronary artery aneurysm was reported by Morgagni in 1961 (2). In our-days incidence of CAAs varies from 1,5% to 5% of the coronary angiographies (1), with male dominance. The most common etiology is atherosclerosis (50%) (3). CAAs of atherosclerotic or inflammatory origin, are usually multiple and involve more than one coronary vessel, compared to the congenital, trau-matic or dissecting ones (3,4). CAAs most frequently devel-ops on the right coronary artery, while a LAD involve-ment is less common. The natural history of CAAs is largely unknown: previous reports (3-5) have report-ed thrombosis and distal embolization, rupture and spasm.

2. Case reportWe report a 60 year-old man, obese, medicated

with anti-hypertensive therapy and β-blockers, com-plaining on effort dyspnea. After a positive stress test at 100W on the anterior wall, without angina, he was admitted to our institution. The coronary angiogra-phy showed a calcified left main stem. The LAD was calcified in the proximal third, with a significant ste-nosis (20 mm long) at the middle third, followed by an aneurysm of 6 mm of diameter involving the origin of a second diagonal branch (Figure 1).

Distally to the lesion, the LAD was normal as oth-er coronaries as well. Septal branches emerging from the LAD proximally an distally to the aneurysm were visualised well. We decided that it was too risky to stent the LAD (6), because the lesion seemed sever-

ly calcified at the angiogram, so we perform an elective surgi-cal revascularization with an off pump LIMA to LAD, using an Octopus® stabilizer (Medtronic Inc, USA). Intra-operatively, the aneurysm confirmed to be calcified, so we preferred to leave it, concerning that having a mammary graft distally, would de-crease the flow stress on the aneurysmatic wall, avoiding com-plications that other procedures like ligature, plicature, end to end anastomosis or patch of the aneurysm have in the litera-ture (7-10).

Postoperative coronary angiography revealed a well-estab-lished blood flow throught the LIMA to distal LAD, and an in-tact coronary aneurysm (Figure 1, 2).

Patient was extubated on 6th postoperative hour, discharged from Intensive care unit on 1st postoperative day. He was dis-charged from hospital on 7th postoperative day, symptom free and with normal electrocardiographic pattern, on medical ther-apy with calcium antagonist, transdermic nitrates and aspirin. Ten months post-operatively, he was symptom free with a nega-tive stress test, on aspirin and anti-hypertensive treatment.

3. discussionNatural history of CAAs goes to calcification, intraaneu-

rismatic trombosis uderwise may leed to spontanious rupture and bleeding.

Preoperative evaluation of digitaly derivated datas, alowed us to confirm angiogram finding and bring proper descission

off-pump bypass for left anterior descending aneurysm with Support of digital techniquesMirsad Kacila, Marco l.S. Matteucci, Stefano bevilacqua, Nermir Granov, Mattia GlauberCardiac Surgery department, iFC-CNr-Pasquinucci Hospital, Massa, italy

FiGUrE 1. Aneurysm of RIVA 6 mm of diameter involving the origin of a second diagonal branch

AIM 2005, 13(4): 223-224


Papers

intraoperativly. As we found calcified aneurism whithout thret of rupture we decided to apply terminolateral anastomosis dis-tally avoiding aneurysm touch.

Postoperative course was uneventfull, patient was simptom free, chronicaly medicated due to other things.

4. Conclusion Surgical treatment of CAAs is a big challenge. It dependes

on preoperative good diagnostic tool and surgical experience and skill. Creating the Bypass and not touching the aneurysm was optimal choice for the patient in this particulare case.

REfEREnCEs1. Roberson T, Fischer L. Prognostic significance of cor-onary artery aneurysm and ectasia in the coronary artery surgery study (CASS) registry. In Shulman ST eds.2. Kawasaki disease: Proceedings of the Second Interna-tional Kawasaki Symposium. New York, NY, A R Liss, 1987: 325-39.3. Morgagni JB. De sedibus et causis morborum. Venectus Tom I, 1961; Epis 27, Art 28.4. Syed M, Lesch M. Coronary artery aneurysm: a review. Progress in Cardiovascular Diseases, 1997; 40: 77-84. 5. Virmani R, Robinowitz M, Atkinson JB. et al. Acquired coronary arterial aneurism: an autopsy study of 52 patients. Human Pathology, 1986; 17: 575-83. 6. Von Rotz F, Niederhauser U, Straumann E, Kurz D, Brtel O, Turina MI. Myocardial infarction caused by a large coro-nary artery aneurysm. Ann Thorac Surg, 2000; 69: 1568-9.7. Leung AW, Wong P, Wu CW, Tsui PT, Mok NS, Lau ST. Left main coronary artery aneurysm: sealing by stent graft and long term follow-up. Catheter Cardiovac Interv, 2000; 51: 205-9. 8. Dralle JG, Turner C, Hsu J, Replogle RL. Coronary ar-tery aneurysm after angioplasty and atherectomy. Ann Thorac Surg, 1995; 59: 1030-5.9. Harandi S, Johnston SB, Wood RE, Roberts WC. Oper-ative therapy of coronary arterial aneurysm. The American

Journal of Cardiology, 1999; 83: 1290-93.10. Westaby S, Vaccari G, Katsumata T. Direct repair of a giant

right coronary aneurysm. Ann Thorac Surg, 1999; 68:1401-3.

11. Kuwaki KK, Morishita KK, Abe TT. Saphenous vein patch an-gioplasty for a discrete saccular aneurysm of left anterior de-scending coronary artery. Ann Thorac Surg, 2000; 25: 342-4.

Correspondence to: dr Mirsad Kacila, Clinical Centre Sarajevo, bolnićka 25, 71000 Sarajevo, bosna i Hercegovina, Phone: +387 61

13 32 88, Fax: +387 33 44 55 22

FiGUrE 2. Postoperative coronary angiography revealed a well-established blood flow through the LIMA to distal LAD, and an intact coronary aneurysm



Case reportsSUMMarYEndometriosis is a disease at which cells endometri-oses of uterus are found beyond its anatomic border. Hotbeds of endometriosis are acting as an autoch-thonous uterus, which reacts adequately on hormon-al cycle. Endometriosis has been described in the late 1860.g. rokitasky. this is the disease of fertile wom-en’s, and with clime passes by, though in the litera-ture two exempts are described. Endometriosis is di-vided on the external and internal endometriosis. internal endometriosis is spreading stage endome-triosis in uterus and fallopian tube. External endo-metriosis means all other localizations. also on the external endometriosis it is required to make differ-ence with intra peritoneal and extra peritoneal local-ization. Predicting diagnosis for the external and in-ternal endometriosis intrinsically differs. after detail anamneses, inspection of the whole body may dis-cover increases on different localization. Gynecolog-ic examination, where palpate increases tumefaction hysterosalpingography, rectal examination, cistos-copy are some of the few examination, which will al-low us to closely predict diagnosis. Ultrasound is the one that can provide us with the valuable information about increase of certain parts. Scintigraphy with ob-vious boost of lymph C-125 is the one, which shows the existence of endometriosis though the most mod-ern method is diagnostic laparoscopy with biopsy. in the cases which we have presented, obviously we did not in one case considered the external endometrio-sis. in the first case the surgeons have joined the op-eration once they have realized pathological exam on the rectum. if in the preoperative examination, we have consulted a change, it would be possible and pathologically verified and the therapy would be dif-ferent (except surgical and hormonal). another case shows ultrasound from radiologist, which had impart-ed maximum information on substratum. in this case the patient has been additionaly sent to gynecolo-gist with recommendation of control and counseling on behalf of the further contingent therapy. diagnos-tic of every patient must be totalitarian. in the case and with at least doubt on the external endometrio-sis, proceed beside ultrasound with scintigraphy. lap-aroscopy with biopsy presents method of choice. Key words: Endometriosis externa.

1. introductionEndometriosis is a disease at which cells endometrioses of

uterus are found beyond its anatomic border. Hotbeds of en-dometriosis are acting as an autochthonous uterus, which re-acts adequately on hormonal circle. Endometriosis has been described in the late 1860.g. Rokitasky. This is the disease of fertile women’s, and with climate passes by, though in the liter-ature two exempts are described.

Endometriosis is divided on the external and internal en-dometriosis. Internal endometriosis is spreading stage endo-metriosis in uterus and fallopian tube. External endometriosis means all other localizations. Also on the external endometri-osis it is required to make difference with intraperitoneal and extra peritoneal localization.

Intra peritoneal endometriosis is sorted by type in the once which are localized on ovary, top of uterus, Lig. Rotunda, Lig. latum Lig. ovary proprium, tunice serosae of rectum, sigmoid, caecum, appendix, intestinum tenuis and etc.

From extra peritoneal localization it is very important to mention vaginal, perineum, hernia inguinalis and femoralis, lymphatic bow, on extremities, in muscles and bones, in CNS, and not infrequently in lungs. Considering measurement pro-cess and consecutively capturing the organs next to it, endo-metriosis is divided on minimal, modern, temper and strong endometriosis. American society for fertility proposed a fifth degree, which would consider inoperable cases (1).

Appearance of disease is understood in two ways. First one is the differential of cells peritoneum in the cell of endometri-osis, and second, appearance means exit of menstrual blood across tube of uterine, when it comes to plantation of cells en-dometriosis on the other places. Disease is not uncommon and it is present in between 4-10 gynecologic patients. These facts are referred mainly on the internal endometriosis (2).

Sequence of disease is not typical and it does not have de-termined sequence. There are cases, which have been followed many years, unlike from examples that are flushing very fast. Close to localization and sequence of disease the symptom are different. The most common characteristics of endometriosis are pain and bleeding. Providing that localization is on geni-tal system, bleeding will be noticeable. However, at localization on purge or urinal funnel, we may have symptom of bleeding from digestive purge or in urinal system (3,4,5). Bleeding is ap-pearing in the sequence of menstruation and pains that are ap-pearing are atypical.

Diagnosis of disease is difficult. After detail anamneses, in-spection of the whole body may discover increases on differ-ent localization. Gynecologic exam, where palpate increases

Endometriosis Externa treated by diagnostic laparoscopyHaris tanovic, ratko Juricic, Samir MuhovicClinic for abdominal surgery, Clinical center Sarajevo, Sarajevo, bosnia and Herzegovina

AIM 2005, 13(4): 225-227


Papers

tumefaction, hysterosalpingography, rectal exam, cistoscopy are some of the examination, which will allow us to be clos-er to diagnosis. Ultrasound is the one that can provide us with the valuable information about increases of certain parts of tis-sue (Figure 1), (6).

Scintigraphy with the obvious boost of lymph C-125 is the one, which demonstrates the existence of endometriosis (Pic-ture 2), (7,8). Although modern method is diagnostic laparot-omia (Figure 3) respectively today laparoscopic (Figure 4), (9).Then it is possible to take biopsy and with pathological exam consult what type of change is about (10).

Treatment of disease depends on the increase of symptom, stage and localization of disease, as well as sequence of disease. Conservative treatment means hormonal therapy. Operational treatment means total excision and extirpation, but also a pal-liative reduction of tumor masseur, in the cases of non opera-tion it comes into consideration. Operative treatment can be combined with hormonal therapy.

2. resultsPatient R.Z. age 43 years is accepted on the Gynecologic

clinic due to operative treatment of verified tumor uterus. Pa-tient beside child disease has been always healthy. Two chil-dren. Complete laboratory exam is good. Rtg-of lungs with exam of pulmology in normal borders. The ultrasound con-firmed verified tumor of uterus, measuring 12x 8 cm. Executed

preoperative preparation and the patient was operated under general anesthesia 23rd. of March 2000.During the operation the distinct tumor of rectum was found and urgently called ab-dominal surgeant. Before joining the operation RT: at the top of one forefinger on 6-7 cm palpable tumefaction, which covers half of circumference, confirming thick structure. Operation-ally consulted tumefaction on rectum, measurement of man-darin. Decided for lower resection of rectum with T- T anas-tomosis with circularly stapler diameter 33 mm, and transitory

colostomy. Postopera-tive sequence is pass-ing through well, pas-sage of purge appeared third postoperative day. Drain is taken on the fourth postopera-tive day. Wound fine, surgical tread ex tenth postoperative day.

After two weeks PH exam 1138/2000: Dg: Endometriosis recti.

After seven weeks closed colostomy. Patient P.A. old 34 years, accepted on gynecologic clinic due

to childbirth. 11th. of November 2003, the indication set for section Caesarea, with child birth 13th. of November.2003, op-eration went well. Child healthy. Postoperative se-quence also passes well, and the patient is released home seventh postopera-tive day. Tread taken out ambulatory. The patient returned back to the am-bulance clinic for abdom-inal surgery on 5th. of February 2004, where it has been consulted tu-mefaction in the region of operation of wound. Ultrasound is done and it’s showing that in me-dium line of tegmentum, low suprapubical, isolat-ed focal lesion which is in under skin structure, hipodenzal aspect, mea-sure 2, 5 cm, and there is no communication with abdominal whole. It is deemed that it could be about inflammation of gland, which perhaps should be required to op-erationally be removed.

After that the patient has been accepted at the abdominal clinic, where the preoperative prepa-ration has been done. All laboratory findings are normal. EKG, as well as Rtg. of lungs is fine. Exam of internist is fine with proposal for operation-al treatment. Patient op-erated 13th. of February 2004. Operational diagnosis: Tu Schloferi. Operations Exci-sio Tu Schloferi in toto. In operational finding is described tu-mefaction in the size of walnut, in operational wound, which as a whole is removed. Postoperative sequence is passing fine, fit of purge is established spontaneously. Wound is healing per primam intentionem and the patient is released home 17th. of February 2004, with the recommendation of extracting thread

PiCtUrE 1. Ultraonography of endometriosis ovarii

PiCtUrE 2. Scintigraphy endometriosis with hight seum C-125

PiCtUrE 3. Endometrisis of abdominal wall

PiCtUrE 4. Laparoskopic wiew of endometriosis abdominal wall

PiCtUrE 5. Laparoskopic wiew of endometriosis abdominal wall

PiCtUrE 6. Laparoskopic wiew of endometriosis inguinal hernia



in ambulance. The 20th.of February 2004, PH finding Br: 818/2004: Dg: Endometriosis extern.

3. discussion Endometriosis as a disease is relatively often (11). Setting

diagnosis for external and internal endometriosis is intrinsi-cally different. While internal endometriosis is a routine diag-nostic as well as therapy, external endometriosis is presenting in real meaning diagnostic challenge. The most interesting are comparative results about preoperative exams of patients from gynecologists and surgical.

With the same number of tested patient, and this is 36, sur-geon’s did not had any search, nor questions directed at an en-dometriosis. Gynecologist found out in 70% patient genitals endometriosis. Suggested hormonal therapy has been present at 30% patient, from the aspect neither of gynecologist, while surgeons have not suggested nor in one case. Patients have not been instructed by surgeons to gynecologic exam in 19% cases, while gynecologists haven’t allowed doing it.

Operating treatment is also different in both tested group. Partial excision has been done by gynecologist in 30% cases and by surgeons in 8% cases. In postoperative treatment gy-necologist have suggested hormonal therapy in 47% cases and surgeon’s only 8% cases (12).

In the cases which we have presented, it is evident that in none of the cases we have thought on external endometrio-sis. In a first case the surgeon’s have joined the operation when they have noticed pathology finding on rectum. If in the preop-erative exam, we had constitution change, it would have been pathological verified and the therapy could have been different (except surgical and hormonal). Second case is showing ultra-sound exam from radiology, which is giving maximal informa-tion about case. In this case patient has been additionally sent to gynecologist with the recommendation of control and coun-seling for the future therapy.

4. ConclusionDiagnosis for each patient has to be consistent. In the case

even of slim assumption on the external endometriosis, we

need to do beside ultrasound a scintigraphy. Laparoscopy with biopsy presents methods of choice.

REfEREnCEs1. Revised American Fertility Society classification of endo-

metriosis. Fertil Steril, 1985; 43: 351-2.2. Lansac J, Pierre F, Letessier E. Digestive endometriosis:

Results of multicentre investigation. Contrib Gynecol Ob-stet, 1987; 16: 192-204.

3. Charles S. Pelvis and umbilical endometriosis presenting with hemorragic pleural effusion. Int Surg 1981; 66: 341

4. Fubertta A, et al, Endometriosi interessante l apparao uri-nario. Min Urol Nefr, 1987; 39: 1.

5. Macafee C, Greer H. Intestinal endometriosis. A raport of 29 case and a survey of literatue. J Obstet Gynecol Br Emp, 1960; 67: 549-55.

6. Kurijak A, Fučkar Ž, Charbi HA. Atlas of abdominal and small parts sonography. Naprijed, Zagreb, 1990; 293.

7. Kennedy SH, Soper NDW, Mojiminiyi OA, Shepston BJ, Barlow DH. Immunoscintigraphy of ovarian endometioa-sis. Preliminary study. Br J Obstet Gynecol, 1988; 95: 693 -7.

8. Gordon R, Evers K, Krassel Laufer I, Herlinger H. Duble contrast enema in pelvic endometriosis. Am J Rontgenol, 1982; 52: 906-7.

9. Tanović H. Osnovi laparoskopijske hiruške tehnike. Šahinpašić, Sarajevo, 2000: 16.

10. Muephi AA, Green WR, Bobbie D, dels Cruz, ZC, Rock JA. Unsuspect endometriosis documented by scanning electron microscopy in visually normal peritoneum. Fertil Steril, 1986; 46: 522-4.

11. Di Giacomo G. et al. Aspetti istiopatologici,clinici e prog-nostici dei tumori a muscolatura liscia dell utero (miomi cellulati, leomiosarcomi, sarcomi stronali endometriali). Min Gin1988; 40: 75.

12. Moris JM, Malat AR. Oxford textbook of surgery. Oxford university press, Oxford, 1994; 1432-7.

Correspondence to: Prof Haris tanovic, Clinic for abdominal surgery, Clinical Center Sarajevo, phone: 00 387 33 297 000.


Papers

1. introductionIn the period 28 to 31 August in Geneva, Switzerland was

held XIX congress of Medical Informatics Europe 2005 under the title: Connecting medical Informatics and Bioinformatics. The congress was organized by the European Federation for Medical Informatics, the Swiss Society for Medical Informatics and the University of Geneva. Main goal is the convergence of scientific leaders during three days for the exchange of ideas, opinions, solutions and other original experiences.

The Medical Informatics Europe (MIE) congress is the main scientific event in medical informatics to be held ev-

ery two years out of three in Europe. Its attendance amounts up to 1000 participants from 26 European member countries of the European Federation for Medical Informatics (EFMI). Several scientists from other parts of the world regularly attend this event. Expected scientific contributions to the con-gress are welcome from any countries all over the world. Inter-national experts reviewed the contributions. The proceedings of the conference are indexed in MEDLINE.

2. aims and scope of the conferenceMIE2005 addresses all topics of bio-informatics, this in-

cluding both medical and bio-medical informatics. The main theme of 2005 was the development of connections between bio-informatics and medical informatics. Tools and concepts from both disciplines can potentate each other. Information derived at the molecular level influences patient care process-es. New bridges must be built. For this purpose, a panel of high level scientists was invited to participate and to give keynote speeches, tutorials and workshop discussions. A couple of se-lected topics were presented in the form of plenary sessions. The final programme of MIE2005 was organized under the form of parallel tracks.

Thematic tracks with selected papers and posters:sharing resources between bio-informatics and medical informatics: sharing knowledge, common data and tools, interactions and potentialisation for the benefit of pa-tients;

•

image processing and feature extraction; electronic health records and information systems; bio-signal analysis, human interface, quality assessment, patients as citizens, statistical methods, knowledge engi-neering, telematics, etc.

3. Scientific sessionsScientific sessions on Monday, 29 August were organized

by the following schedule: Terminologies, ontology, standards and knowledge engineering, Organization change, informa-tion needs, Natural language processing, Bioinformatics and

•••

MiE 2005 – Geneva, Switzerland ahmed Novo, izet Masic, Selim toromanovic, Zlatan MasicFaculty of medicine, University of Sarajevo, bosnia and Herzegovina

NEW

S

AIM 2005, 13(4): 227-229



medical genomics, Implemen-tation and evaluation of clinical systems, Public health informat-ics and clinical trials, Healthcare networks and Imaging informat-ics. Conference key note was held by Amos Bairoch on SwissProt in the context of medical infor-mation. On Tuesday, 30 August on scientific sessions provided were: Computerized patient re-cord, Implementation and evalu-ation of clinical systems, Natural language processing, text mining and information retrieval, Educa-tional Technologies and Method-ologies and Imaging informatics, Handheld and wireless comput-ing, Online health information & patient empowerment and Deci-sion support and clinical guide-lines. Conference key note was held by Ioannis Xenarios on Bio-informatics for Drug Discovery and Development.

Poster sessions were held in main hall of UNI-MAIL centre on Wednesday, 31 August. Over 1000 participant had opportuni-ty to see over 100 posters on vari-ous subjects. Conference keynote was presented by Ilias Iakovi-dis on Biomedical Informatics – one discipline in the future? As-sessment and future prospects of EU efforts in facilitating synergy between medical and bioinfor-matics. B&H Society was present with two presentation: oral pre-sentation on Distance learning at Biomedical faculties in B&H by Izet Masic, Ahmed Novo, Zlatan Masic, Mensura Kudumovic, Se-lim Toromanovic, Admir Rama, Almir Dzananovic, Ilda Bander, Mirza Basic, Emir Guso and Eldar Balta and poster presentation on Medical Informat-ics Education in B&H by Ahmed Novo, Izet Masic, Mensura Kudumovic and Zlatan Masic.

4. EFMi board meeting reportBefore opening ceremony in UNI MAIL building in Gene-

va, during Saturday and Sunday (26-27 August 2005) were held number of EFMI and IMIA meetings: IMIA Board meeting, EFMI Council meeting, IMIA general assembly, IMIA strategic

Planning Committee and IMIA year book meeting. On EFMI board and council meeting was accepted report from John Bryden, MD, member of EFMI board who visited Bosnia and Herzegovina and Croatia in May this year, in order to check capability of both societies and organization capacities to have XXI congress of Medical Informatics Europe 2009 in Saraje-vo. Also, prof Masic, MD, PhD, had his presentation to repre-sent efforts done by members of Local organizing committee who have made extensive progress for the congress prepara-tion matters. EFMI board decided that MIE 2009 will be or-


Papers

ganized only by Bosnia and Herzegovina Society for Medical Informatics instead of joint organization of B&H and Croa-tian societies for Medical Informatics as planned before. In un-official conversation with EFMI board members we found out that they were very impressed with prog-ress made by BHSMI. Four years in advance we have presented posters, badges (name tags), B&H medical informatics journals, bud-get and other staff for MIE 2009 which is something what have not been ever made before of any con-gress local organizing com-mittee.

5. Congress proceedings

A wide variety of topics of bio-informatics, includ-ing both medical and bio-medical informatics are ad-dressed by MIE. The main theme in this publication is the development of connec-tions between bio-informatics and medical informatics. Tools and concepts from both disciplines can complement each oth-er. Information derived at the molecular level influences pa-tient care processes. New bridges must be built. For this pur-pose, a panel of high level scientists was invited to participate and to give keynote speeches, tutorials and workshop discus-sions. Topics include: sharing resources between bio-informat-

ics and medical informatics; sharing knowledge, common data and tools, interactions and potentialization for the benefit of patients; image processing and feature extraction; electronic health records and information systems; bio-signal analysis, human interface, quality assessment, patients as citizens, sta-tistical methods, knowledge engineering, telematics, etc.

Contents of the proceedings: Bioinformatics and Medical Genomics Computerized Patient Record Decision Support and Clinical Guidelines Educational Technologies and Methodologies Handheld and Wireless Computing Healthcare Networks Imaging Informatics Implementation & Evaluation of Clinical Systems Terminologies, Ontology, Standards aand Knowledge En-gineering Natural Language, Text Mining and Information Retrieval Online Health Information & Patient Empowerment Organization Change, Information Needs Public Health Informatics, Clinical Trials

6. Next EFMi congressNext the 20th International congress of the European Feder-

ation for Medical Informatics MIE 2006 will be held in Maas-tricht, the Netherlands in the period 27-30 August 2006. In

addition to the traditional topics of health and biomedical in-formatics, “Ubiquity: technologies for better health in aging societies”, a promising field for the future of health care, has been chosen as special topic for MIE2006. Besides the scientif-ic programme, the exhibition area will be the place where par-ticipants can get in touch with the different companies who will display their newest systems and products.

•••••••••

••••

PaSt MiE CoNFErENCES• Geneva, Switzerland, 2005

• Saint Malo, France, 2003

• budapest, Hungary, 2002

• Hanover, Germany, 2000

• ljubljana, Slovenia, 1999

• thessalonica, Greece, 1997

• Copenhagen, denmark, 1996

• lisbon, Portugal, 1994

• Jerusalem, israel, 1993

• Vienna, austria, 1991

• Glasgow, Scotland, 1990

• oslo, Norway, 1988

• rome, italy, 1987

• Helsinki, Finland, 1985

• brussels, belgium, 1984

• dublin, ireland, 1982

• toulouse, France, 1981

• berlin, Germany, 1979

• Cambridge, UK, 1978



MIE

SA

RA

JE

VO09

MIE

SA

RA

JE

VO09

CO

NG

RE

SS

PR

ES

EN

TA

TIO

N

AIM

AN

DS

CO

PE

PA

ST

MIE

CO

NF

ER

EN

CE

S

AIM

AN

DS

CO

PE

Gote

nbur

g, S

wede

n, 2

008

Mas

trih

t, Ne

ther

land

, 200

6Ge

neva

, Swi

tzer

land

, 200

5Sa

int M

alo,

Fra

nce,

200

3Bu

dape

st, H

unga

ry, 2

002

Hano

ver,

Germ

any,

2000

Ljub

ljana

, Slo

veni

a, 1

999

Port

o Ca

rras

, Gre

ece,

199

7Co

penh

agen

, Den

mar

k, 1

996

Lisb

on, P

ortu

gal,

1994

Jeru

sale

m, I

srae

l, 19

93Vi

enna

, Aus

tria

, 199

1Gl

asgo

w, S

cotla

nd, 1

990

Oslo

, Nor

way,

1988

Rom

e, It

aly,

1987

Hels

inki

, Fin

land

, 198

5Br

usse

ls, B

elgi

um, 1

984

TH

E2

2N

DIN

TE

RN

AT

ION

AL

CO

NG

RE

SS

OF

TH

EE

UR

OP

EA

N

CO

NF

ED

ER

AT

ION

FO

R

ME

DIC

AL

INF

OR

MA

TIC

S

SO

CIE

TY

FO

RM

ED

ICA

LIN

FO

RM

AT

ICS

OF

BO

SN

IAA

ND

HE

RE

ZE

GO

VIN

A

EU

RO

PE

AN

FE

DE

RA

TIO

NF

OR

ME

DIC

AL

INF

OR

MA

TIC

S

UN

IVE

RS

ITY

OF

SA

RA

JEV

O

The

mai

n m

otto

of M

IE 2

009

is th

e br

idge

ove

rtr

oubl

ed w

ater

- m

edic

al in

form

atic

s in

uni

ted

and

heal

thie

r Eur

ope.

New

brid

ges

in d

evel

opin

g of

med

ical

info

rmat

ics

betw

een

west

ern

and

east

ern

part

of t

he w

orld

mus

t be

built

.

For t

his

purp

ose,

a p

anel

of h

igh

leve

l sci

entis

ts w

illbe

invi

ted

to p

artic

ipat

e an

d to

giv

e ke

ynot

esp

eech

es, t

utor

ials

and

wor

ksho

p di

scus

sion

s.

The

Med

ical

Info

rmat

ics

Euro

pe (M

IE) c

ongr

ess

is th

e m

ain

scie

ntifi

c ev

ent i

n m

edic

al in

form

atic

s to

be

held

eve

ry tw

oye

ars

out o

f thr

ee in

Eur

ope.

Its

atte

ndan

ce a

mou

nts

up to

800

part

icip

ants

from

29

Euro

pean

mem

ber c

ount

ries

of th

eEu

rope

an F

eder

atio

n fo

r Med

ical

Info

rmat

ics

(EFM

I). S

ever

alsc

ient

ists

frm

oth

er p

arts

of t

he w

ord

regu

larly

atte

nd th

isev

ent.

Expe

cted

sci

entif

ic c

ontr

ibut

ions

to th

e co

ngre

ss a

re w

elco

me

from

any

cou

ntrie

s al

l ove

r the

wor

ld. T

he S

cien

tific

Pro

gram

Com

mitt

ee a

nd a

pan

el o

f int

erna

tiona

l exp

erts

will

revi

ew th

eco

ntrib

utio

ns. T

he p

roce

edin

gs o

f the

con

fere

nce

are

inde

xed

in M

EDLI

NE:

Cont

ribut

ions

are

:Fu

ll pa

pers

: orig

inal

sci

entif

ic a

rtic

les,

up

to 6

pag

es, 2

0m

inut

es p

rese

ntat

ions

, pee

r rev

iewe

d by

a c

olle

ge o

fin

tern

atio

nal e

xper

ts, r

elat

ing

rece

nt w

ork

or o

pini

on o

f the

auth

ors,

pub

lishe

d in

the

conf

eren

ce p

aper

pro

ceed

ings

and

the

conf

eren

ce C

D-RO

M.

Shor

t pap

ers:

as

ful p

aper

s, b

y de

cisi

on o

f the

SPC

, for

5m

inut

es p

rese

ntat

ions

, pub

lishe

d on

the

conf

eren

ce C

D-RO

Mon

ly.Po

ster

s: o

rigin

al p

rese

ntat

ions

of p

oste

rs, p

ublis

hed

on th

eco

nfer

ence

CD-

ROM

:Th

emat

ic tr

acks

with

sel

ecte

d fu

l pap

ers

on s

peci

fic to

pics

:bi

oinf

orm

atic

s, o

ntol

ogic

al a

ppro

ache

s, a

nd k

nowl

edge

repr

esen

tatio

n;im

age

proc

essi

ng a

nd fe

atur

e ex

trac

tion;

elec

tron

ic h

ealth

reco

rds

and

info

rmat

ion

syst

ems;

gen

eral

trac

ks o

n ot

her t

opic

s.W

orks

hops

: the

mat

ic d

ebat

es a

nd p

rese

ntat

ions

und

er th

ele

ader

ship

of E

FMI w

orki

ng g

roup

s or

est

ablis

hed

entit

ies.

Dem

onst

ratio

ns: p

rovi

sion

for t

heat

er-s

tyle

dem

onst

ratio

n of

appl

icat

ions

.

ME

DIC

AL

INF

OR

MA

TIC

SIN

HE

AL

TH

IER

AN

DU

NIT

ED

EU

RO

PE

30

/08

-0

2/0

9

THE BRIDGE OVER TROUBLED WATER


Papers

MIE

MIE

SA

RA

JE

VO

SA

RA

JE

VO

0909

THE

LOCA

L OR

GANI

ZING

COM

MIT

TEE

THE

EFM

I BOA

RD

THE

SCIE

NTIF

IC P

ROGR

AM C

OMIT

TEE

(SPC

)

Izet

Mas

ic, p

resi

dent

Ahm

ed N

ovo,

sec

reta

ryAi

da P

ilav,

mem

ber

Arm

in S

krbo

, mem

ber

Zora

n Ha

dzia

hmet

ovic

, mem

ber

Zeke

rijah

Sab

anov

ic, m

embe

rOz

ren

Kord

ic, m

embe

rSe

lim To

rom

anov

ic, m

embe

rAl

mir

Dzan

anov

ic, m

embe

rZl

atan

Mas

ic, m

embe

r

To b

e no

min

ated

by

EFM

I cou

ncil

- mem

bers

of E

FMI c

ounc

il- c

hear

s of

wor

king

gro

ups

Robe

rt B

aud,

Pre

side

ntAs

sa R

eich

ert,

Vice

-Pre

side

nt IM

IARo

lf En

gelb

rech

t, In

form

atio

n Of

ficer

Geor

ge M

ihal

as, V

ice

Pres

iden

tPa

tric

k W

eber

, Tre

asur

erAr

ie H

asm

an, P

ublic

atio

n Of

ficer

Jako

b Ho

fdijk

, Sec

reta

ryJo

hn B

ryde

n, O

ffice

r

ww

w.i

mas

ic.o

rg/b

hsm

i/

HOTE

L AC

COM

MOD

ATIO

N

TRAV

EL

LOCA

L TR

ANSP

ORTA

TION

HOTE

L AC

COM

MOD

ATIO

N

SOCI

AL E

VENT

S

WEA

THER

Holid

ay In

n, S

araj

evo,

Bos

nia

and

Herz

egov

ina

Sara

jevo

is e

asily

reac

hed

by p

lane

, by

trai

n an

d by

road

. Mul

tiple

con

nect

ions

are

ava

ilabl

e.

The

cong

ress

fee

incl

udes

a p

ass

for t

he fu

ll ne

twor

k of

publ

ic tr

ansp

orta

tion.

Solu

tions

exis

t to

fit a

ny b

udge

t and

des

ires,

from

5-

star

s ho

tels

to s

tude

nt ro

oms

as lo

w as

20

Euro

s pe

rni

ght.

For e

xam

ple,

4-s

tars

hot

els

are

quot

ed fr

om 6

0 to

120

Euro

s.Th

e co

ngre

ss o

rgan

izer p

rovi

des

any n

eces

sary

info

rmat

ion.

Num

erou

s op

tions

will

be

prop

osed

in S

araj

evo,

oth

erbe

autif

ul p

arts

of B

osni

a an

d He

rzeg

ovin

a an

d Ad

riatic

seas

ide,

whe

re to

uris

m is

an

all y

ear l

ong

activ

ity.

Usua

lly w

arm

and

dry

at t

he c

ongr

ess

perio

d.

VE

NU

E

REGI

STRA

TION

INST

RUCT

ION

TO A

UTHO

RS

CONG

RESS

REG

ISTR

ATIO

N

CONG

RESS

SEC

RETA

RIAT

INDU

STRI

AL E

XHIB

IT

Regi

stra

tion

is u

niqu

ely p

erfo

rmed

on

the

Inte

rnet

. See

the

conf

eren

ce s

ite a

t www

.mie

2009

.net

Conf

eren

ces

from

: 30t

h Au

gust

200

9 to

3rd

Sep

tem

ber 2

009

Full

pape

r sub

mis

sion

: 16

Janu

ary 2

009

Notif

icat

ion:

18

April

200

9Fi

nal m

anus

crip

t: 16

May

200

9

Avai

labl

e on

web

site

: www

mie

2009

net

PROM

O TO

URS

Tour

istic

age

ncy,

Kem

ala

Kape

tano

vica

b.b

.71

00 S

araj

evo,

Bos

nia

and

Herz

egov

ina

Tel.:

003

87 3

3 65

5 34

6, 7

15 6

91, 7

15 6

92Fa

x: 00

387

33 7

15 6

90E-

mai

l: am

ela@

prom

otou

rs.b

aww

w.pr

omot

ours

.ba

Boot

hs a

re a

vaila

ble.

See

the

webs

ite.

..

* m

embe

rs o

f an

EFM

I affi

liate

d Na

tiona

l Soc

iety

** a

n EF

MI p

olic

y for

stu

dent

s ad

mis

sion

is a

ctiv

ely d

evel

oped

.

CO

NG

RE

SS

RE

GIS

TR

AT

ION

Digital Techniques in Diagnostic of Aseptic Pleural Empyema in Patient After Cardiac Surgery

Documents

Transcript of Digital Techniques in Diagnostic of Aseptic Pleural Empyema in Patient After Cardiac Surgery