Physician Opinions About an Anatomy Core Curriculum:A Case for Medical Imaging and Vertical Integration

Courtney P. Orsbon,1,2* Rebecca S. Kaiser,1 Callum F. Ross2

1The University of Chicago Pritzker School of Medicine, Division of Biological Sciences, Chicago, llinois2Department of Organismal Biology and Anatomy, the University of Chicago, Chicago, Illinois

Pre-clinical anatomy curricula must provide medical students with the knowledge needed ina variety of medical and surgical specialties. But do physicians within specialties agreeabout what anatomical knowledge is most important in their practices? And, what is thecommon core of anatomical knowledge deemed essential by physicians in different special-ties? Answers to these questions would be useful in designing pre-clinical anatomy courses.The primary aim of this study was to assess the importance of a human gross anatomycourse by soliciting the opinions of physicians from a range of specialties. We surveyed 93physicians to determine the importance of specific anatomical topics in their own practices.Their responses were analyzed to assess variation in intra- and inter-departmental attitudestoward the importance of anatomy. Nearly all of the topics taught in the course weredeemed important by the clinicians as a group, but respondents showed little agreement onthe rank order of importance of anatomical topics. Overall, only medical imaging receivedhigh importance by nearly all respondents, and lower importance was attached to embryol-ogy and lymphatic anatomy. Our survey data, however, also suggested distinct hierarchiesin the importance assigned to anatomical topics within specialties. Given that physiciansview the importance of anatomy differently, we suggest that students revisit anatomythrough a vertically integrated curriculum tailored to provide specialty-specific anatomicaltraining to advanced students based on their areas of clinical interest. Integration of medicalimaging into pre-clinical anatomy courses, already underway in many medical schools, is ofhigh clinical relevance. Anat Sci Educ 7: 251–261. VC 2013 American Association of Anatomists.

Key words: gross anatomy education; medical education; anatomy curriculum; core cur-riculum; clinical anatomy; embryology; medical imaging; radiology; undergraduate medi-cal education; vertical integration

INTRODUCTION

Some medical educators argue that, in response to decreasedcontact hours with students, pre-clinical anatomy curricula

should focus more heavily, though not exclusively, on a corecurriculum of the most clinically relevant topics (Heylings,2002; Older, 2004; Drake et al., 2009; Fincher et al., 2009;Craig et al., 2010). Pre-clinical anatomy curricular guidelineshave been developed by surveying physicians in a single spe-cialty or through consensus opinions of groups of expert anat-omists (Leonard et al., 1996, 1999, 2000, 2001; Fasel, 1998;Drake, 2007; McHanwell et al., 2007; Bergman et al., 2008,2011). Such expert-generated guidelines rest on the assumptionthat intuition and experience are sufficient to construct anintroductory anatomy curriculum that adequately emphasizesthose topics most relevant to a variety of specialties. However,non-medically trained anatomists may not have a physician’sperspective of what a medical gross anatomy curriculumshould emphasize. Furthermore, they may expect students toknow more detail than a clinical educator would, potentiallyoverwhelming students with seemingly irrelevant detail thatresults in poor learning and knowledge retention (Koens et al.,2006; Ward and Walker, 2008; Bull and Mattick, 2010; Ward,

Additional Supporting Information may be found in the online versionof this article*Correspondence to: Ms. Courtney P. Orsbon, Department of Organ-ismal Biology and Anatomy, the University of Chicago, 1027 East57th St, Chicago, Illinois 60637 USA. E-mail:courtneyorsbon@uchicago.edu

Grant sponsor: Pritzker School of Medicine Summer ResearchProgram.

Received 17 June 2013; Revised 3 August 2013; Accepted 6 August2013.

Published online 10 September 2013 in Wiley Online Library(wileyonlinelibrary.com). DOI 10.1002/ase.1401

VC 2013 American Association of Anatomists

Anatomical Sciences Education JULY/AUGUST 2014 Anat Sci Educ 7:251–261 (2014)

RESEARCH REPORT

2011). Even a physician course director faces the challenge ofdesigning a curriculum that is not unduly influenced by theclinical importance of specific anatomical concepts in his orher own field of medicine. For example, studies have shownthat, although many physicians value anatomy as the mostimportant basic science, they value and apply anatomicalknowledge differently depending on specialty (Pabst, 2009;Arr�aez-Aybar et al., 2010; Lazarus et al., 2012).

Certain anatomical topics, such as basic heart anatomy,intuitively seem important across many specialties, whereasothers seem more important in specific specialties and lessimportant in others. The few studies that identify those spe-cific anatomical topics most important to a single specialtydo not compare whether or how these topics significantly dif-fer between specialties (Fasel, 1998; Fasel et al., 1999; Kilroyand Driscoll, 2006; Wadey et al., 2007; DeFriez et al., 2011).As such, the current literature largely neglects the specifics ofwhat anatomical knowledge is most important in differentclinical settings. Given that a pre-clinical anatomy curriculummust prepare medical students to enter a variety of special-ties, we wanted to explore what anatomical knowledge ismost essential to different kinds of physicians, and whetherthere exists a “core” of anatomical knowledge that a major-ity of physicians agree is essential to medical and surgicalpractice, as some medical educators have suggested (Louwet al., 2009; Abu-Hijleh, 2010; Chapuis et al., 2010).

The primary aim of this study was to evaluate the contentof the human gross anatomy course at the University of Chi-cago relative to the importance attached to anatomical con-cepts by physicians in different medical specialties at theUniversity of Chicago Hospitals. We wanted to knowwhether this pre-clinical anatomy course prepares studentsfor a variety of specialties, and we hypothesized that 75% oftopics taught in the course would receive a median rating of“moderately important.” We also wanted to assess whetherand how the importance of anatomical knowledge differsamong specialties, and we hypothesized that physicianswould agree on the importance of some areas of anatomicalknowledge but not on others. Furthermore, we hypothesizedthat there would be higher agreement among physicians inthe same specialty regarding the rank order of importance ofbody regions and anatomical themes.

METHODS

We designed a cross-sectional study to assess physicians’opinions about the importance of anatomical knowledge intheir own clinical practice. The study consisted of a web-based survey of a single-institution, stratified, random sampleof faculty from all departments at the University of ChicagoHospitals (UCH). The study was granted exempt status bythe University of Chicago Biological Sciences Division Institu-tional Review Board on the basis of anonymity of theresponses, voluntary participation, and minimal risk toparticipants.

Sample Selection

Physicians were selected from a publically available, onlinelist of faculty at UCH. Only those holding medical doctor(M.D.), doctor of osteopathic medicine (D.O.), or doctor ofoptometry (O.D.) degrees were included. Candidates werestratified according to their primary department of affiliation,

as determined by what was listed first in the faculty listing,and cross-referenced with the faculty member’s publicallyavailable UCH biography.

Sample size was determined utilizing the sampsi functionof Stata (Stata 12, StataCorp, College Station, TX), assuminga significance level of 0.05 and a power of 0.80. Utilizingthis function, it was determined that a minimum of six physi-cians from each department was necessary. Twenty facultymembers were randomly selected from each department orsection; in the event that the department or section wassmaller than 20 faculty members, all members of that depart-ment or section were included in the study. After the facultymembers were selected, the online University of Chicagodirectory was used to find each individual’s email address. Ifa faculty member could not be contacted, that individual wasremoved from the sample and another faculty member wasrandomly selected to take his or her place. If less than sixphysicians from a department completed at least 80% thesurvey, then that department was excluded from departmen-tal comparisons, although the individual faculty wereincluded in analyses involving all physicians.

Survey Design

A comprehensive list of 251 anatomy course topics was gen-erated by the authors from lecture slides used in the 2010and 2011 first-year combined gross anatomy and embryologycourse at the University of Chicago Pritzker School of Medi-cine. The course is 10 weeks long, consisting of 84 lecturesof 68.5 contact hours and 56 laboratories of 137.5 hours, fora total of 206 contact hours (Supporting Information TableS1). The course consists of seven modules: one embryology/body plan module and six regional gross anatomy moduleswith medical imaging and embryology integrated into them.The course director (C.F.R.) developed the medical imagingcurriculum in collaboration with a radiology faculty memberat the University of Chicago. Topics presented in lecturesgiven by one of the four primary instructors were included inthe survey. One author (R.K.) independently compiled the listfrom the lecture slides of the 2010 course, and another(C.O.) compared this list to the lecture slides from the 2011course and made adjustments to the list as necessary.

A web-based survey (SurveyMonkey, Palo Alto, CA) wascreated to distribute to faculty physicians at UCH. Prior todistribution to the sample population, the survey was distrib-uted to a focus group, composed of UCH faculty and a medi-cal education specialist, who provided qualitative feedbackon the goals of the survey, and a pilot group of UCH faculty,including two general internists, a radiologist, and two pedia-tricians, who provided preliminary results for mock dataanalysis and qualitative comments on the functionality andclarity of the survey.

The list of anatomy topics was imported into the web-based survey, and each topic was assigned to both a“module,” defined as the body region module during whichthe topic was taught, and a “theme” within each module,defined as one of seven basic elements of anatomy that theauthors agreed were consistent throughout the modules, asexemplified in Supporting Information Figure S1 and TableS2. The themes were: embryology, function, lymph, medicalimaging, innervation, structure, and vasculature. The struc-ture theme included those topics featuring general morphol-ogy, such as musculoskeletal anatomy, general organ

252 Orsbon et al.

anatomy, spatial relationships, and pathways of nerves andvasculature. The function theme included those topics featur-ing biomechanical and clinical applications of anatomy,including cardiopulmonary blood flow, testing cranial nerves,muscle function, and gait analysis. Questions regarding medi-cal imaging were included as a separate page in the survey.For a complete listing of survey topics and their assignedmodules and themes, see Supporting Information Table S2. ALikert-type scale was used to rate the importance of eachanatomical topic. Options included “essential,” “veryimportant,” “moderately important,” “less important,” “notimportant,” and “no opinion/not sure.” A no opinion/notsure option was included so that a physician did not need toguess the value of the topic in the event that he or she foundthe question unclear, and these responses were coded as miss-ing data in data analysis. We defined essential as that knowl-edge which physicians deemed fundamental in providingclinical care to patients, communicating with other healthcare professionals, and/or keeping up with relevant literature.The survey was specifically designed and worded to empha-size that the physicians should respond based on their opin-ions of the importance of anatomy in their own practicesrather than their opinions of what anatomy is most impor-tant for students and physicians to know in general. Self-identified demographic data included: gender, degrees held,year of graduation from medical school, elective anatomycourses taken or taught, involvement in medical education,department of primary affiliation, self-identified medical spe-cialty, and opinions of the knowledge of anatomy by medicalstudents at the University of Chicago.

Data Analysis

Data were directly imported into and analyzed with the sta-tistical software package SPSS 19, (IBM Inc., Armonk, NY).Each Likert response was imported as an integer on an ordi-nal scale. Summative scales were used to create a total scoreof the ratings of all topics within each module and theme;modules and themes are referred to as “domains.” For statis-tical tests that required comparisons of different domains, anindividual’s score for each domain was divided by the maxi-mum possible score for that domain in order to generate astandardized summative score. This was necessary becauseeach domain had a different number of topics assigned to it,and standardization allowed for comparison of domains ofdifferent sizes. For example, if a physician responded withessential to all topics in a domain, the standardized summa-tive score would be 1.00. Likewise, if a physician respondedwith moderately important to all topics in a domain, thesummative score would be 0.50. If a physician respondedwith not important to all topics in a domain, the summativescore would be 0.00. In the event that a value was missing, itwas replaced with the average response within the domain.Individuals missing more than 20% of values within a givendomain where excluded from analysis. Chi-squared analyseswere used to compare responses from physicians with differ-ent demographic variables and screen for confoundingvariables.

Friedman tests were used to assess differences in ratingsassigned to each domain by all physicians as a single groupas well as by physicians stratified by department. Addition-ally, Kendall’s W was used to determine the amount of agree-ment between physicians regarding the rank order of the

domains, where 0 5 no agreement and 1 5 complete agree-ment. For significant results of the Friedman test, we con-ducted a post-hoc analysis using a Wilcoxon signed-rank test.Kruskal-Wallis tests were used to assess differences amongdepartments rating of the same domains, and significant testswere analyzed post-hoc with a Mann-Whitney U test. Signifi-cance of the Friedman, Wilcoxon, Kruskal-Wallis, and Mann-Whitney U tests was defined as P < 0.05, and Kendall’s Wwas interpreted as follows: <0.11, very weak agreement;0.11–0.30, weak agreement; 0.31–0.50, moderate agreement;0.51–0.70, strong agreement; and >0.71, very strong agree-ment (Schmidt, 1997). For tests involving multiple pair-wisecomparisons, a Bonferroni correction was employed.

Reliability and Validity

Because summative score values could be the result of manypossible combinations of responses, Kendall’s tau B andCronbach’s alpha were used to assess the validity and reliabil-ity (respectively) of the criteria used to assign topics to eachdomain (Bland and Altman, 1997; Cluzeau et al., 2003).According to Bland and Altman (1997), Cronbach’s alphavalues of 0.7 to 0.8 are sufficient for comparing groups,whereas values of 0.9 are acceptable for clinical application.In this study, values greater than 0.8 were considered accept-able to justify grouping the topics on a summative scale.

RESULTS

Demographics and Response Rate

Response rates by department and specialty are shown inTables 1 and 2. A total of 93 of 205 selected faculty physi-cians completed the survey for a response rate of 45%(Tables 1 and 2, Fig. 1). Chi-squared tests of response ratesshowed that female physicians (P 5 0.920) and MD/PhDs (P5 0.791) responded in comparable numbers to the samplepopulation. There was a non-significant trend for physiciansinvolved in undergraduate medical education (UME) to ratethe importance of the summative scale of all topics lowerthan physicians not involved in UME (P 5 0.077). Therewere no significant differences in the summative scale of alltopics based on year of graduation (P 5 0.700), gender (P 5

0.225), status as a graduate medical educator (P 5 0.408),possession of a PhD (P 5 0.172), or previous completion ofelective courses in anatomy (P 5 0.409). Departments withfewer than five respondents (Ophthalmology and RadiationOncology) were excluded from interdepartmentalcomparisons.

Validity and Reliability

To assess the validity of the criteria used to determine thecontents of each domain, we calculated Kendall’s tau B rankcorrelation coefficients between physicians’ domain scoresand their overall scores. All Kendall’s tau B rank correlationcoefficients among physicians’ domain ratings and their over-all ratings were significant (P < 0.001). Cronbach’s alphacoefficient was used to measure the reliability of the internalconsistency of the summative Likert scale of all survey topics,of each module, and of each theme. The Cronbach’s alphacoefficient for the summative scale was 0.994 for all topics,0.994 for gross anatomy topics, 0.968 for embryology topics,

Anatomical Sciences Education JULY/AUGUST 2014 253

0.875 or greater for each module, and 0.940 or greater foreach theme. The results of the Kendall’s tau B and Cron-bach’s alpha tests support the validity and reliability of thesummative scale values of the importance of the overall sur-vey as well as of each domain.

Overall Results

A total of 98% of topics had a median rating greater than orequal to moderately important. The median rating of allresponses was very important with an interquartile range ofessential to less important. The frequencies of median ratingsof individual topics by each department and among all physi-cians are shown in Table 3. Median ratings of topics areshown in Supporting Information Table S2. Although 13(5%) topics had a median rating of essential among all physi-cians, a total of 163 (65%) topics had a median rating ofessential in at least one department. These topics are indi-cated with an asterisk in Supporting Information Table S2.Similarly, although 129 (51%) topics had a median rating ofessential or very important among all physicians, a total of242 (96%) unique topics had a median rating of either essen-tial or very important in at least one department.

Although there was a tendency for physicians in differentdepartments to rate the overall importance of all gross anat-omy and embryology topics differently, these differences werenot significant (Kruskal-Wallis test, P 5 0.087, N 5 83, Fig.2). Moreover, when psychiatrists were excluded from analy-sis, this tendency disappeared (Kruskal-Wallis, P 5 0.583, N5 76). When physicians’ ratings of gross anatomy topicswere considered separate from embryology topics, a signifi-

cant departmental effect was detected (Kruskal-Wallis test, P5 0.026, N 5 83). Again, when psychiatrists were excludedfrom analysis, the differences between departments were nolonger significant (Krukal-Wallis, P 5 0.267, N 5 76). Post-hoc comparisons using Mann-Whitney U tests found the psy-chiatrists to rate gross anatomy topics significantly lowerthan the anesthesiologists (P 5 0.005), those in Medicine,(P 5 0.005), the neurologists (P 5 0.001), and the radiolog-ists (P 5 0.007). Differences between departments on allembryology topics were not significant (Kruskal-Wallis test,P 5 0.165, N 5 77).

Note that not all tests included data from all 93 respond-ing physicians. If a physician did not complete 20% of thetopics within each summative scale grouping, then that physi-cian’s responses were excluded from analysis.

Module Analysis

When all physicians’ responses were considered together, theimportance of each of the modules was not significantly dif-ferent (Friedman test, P 5 0.554, N 5 84), and agreementon the rank order of the modules was very weak (W 5

0.009). Physicians within the same department showed more

Table 1.

Response Rate by Department

DepartmentResponse

rateTotal eligible

faculty

Anesthesia and

Critical Care

65% (13/20) 52

Neurology 60% (12/20) 27

Radiology 60% (12/20) 41

Obstetrics and

Gynecology

55% (11/20) 29

Surgery 50% (10/20) 92

Psychiatry 50% (9/18) 18

Pathology 40% (8/20) 29

Pediatrics 35% (7/20) 110

Radiation Oncology 31% (4/13) 13

Medicine 30% (6/20) 269

Ophthalmology 7% (1/14) 14

Total 45% (93/205) 694

Table 2.

Number of Responses per Specialty

Specialty Number of responses

Anesthesia and Critical Care 13

Emergency Medicine 2

General Surgery 5

Internal Medicine 5

Neurology 12

Neurosurgery 1

Obstetrics and Gynecology 11

Ophthalmology 1

Orthopedic Surgery 1

Otolaryngology 1

Pathology 8

Pediatrics 8

Plastic Surgery 1

Psychiatry 8

Pulmonary and Critical Care 1

Radiation Oncology 4

Radiology 10

Urology 1

254 Orsbon et al.

agreement on the rank order of the importance of modulesand some departments differed in the order of importance ofmodules (Table 4). In particular, physicians in Obstetrics and

Gynecology (Ob/Gyn), Neurology, and Psychiatry showedstrong or very strong agreement. Within these departments,pairwise comparisons of modules were not significantly dif-ferent, although several trends toward significance werenoted. Each module exhibited significant differences inimportance between departments (Kruskal-Wallis test, P 5

0.015 to P < 0.001) (Fig. 3).

Theme Analysis

When all physicians’ responses were considered together,themes received significantly different ratings (Friedman testP < 0.001, N 5 80). Agreement of the rank order of eachtheme was moderate (W 5 0.385). Medical imaging topicswere ranked significantly higher than all of the other themes(P < 0.001) and lymph and embryology were ranked signifi-cantly lower than most of the other themes (P 5 0.002 to P<0.001) (Table 5).

Similar to the module analyses, physicians within the samedepartment showed more agreement about ranking themesand, in all departments, significant differences in theresponses to themes (Table 6). Radiation oncology wasincluded post-hoc due to the exceptionally high agreementbetween the four physicians, particularly regarding the highamount of importance given to lymph-themed topics (W 5

0.915, Table 6). However, these physicians were excludedfrom departmental comparisons because their numbers didnot meet the minimum of six required for these comparisons.Most departments showed at least moderate agreement aboutthe order of importance of themes, and those in radiationoncology, psychiatry, surgery, and medicine all showed verystrong agreement. Themes exhibited varied significance in dif-ferences of ratings between departments (Kruskal-Wallis test,P 5 0.234 to P < 0.001) (Fig. 4). Within departments,

Figure 1.

Flowchart of sample selection and response rate. MD, medical doctor; DO,doctor of osteopathic medicine; OD, doctor of optometry.

Table 3.

Frequencies of Median Ratings of Topics by Department

Department EssentialVery

importantModeratelyimportant

Lessimportant

Notimportant

Anesthesia and Critical Care 55 118 60 18 0

Medicine 76 120 27 27 1

Neurology 58 93 57 41 2

Obstetrics and Gynecology 31 82 81 57 0

Pathology 12 58 67 114 0

Pediatrics 45 81 100 24 1

Psychiatry 6 29 64 152 0

Radiation Oncology 40 81 37 83 10

Radiology 59 110 65 17 0

Surgery 42 125 67 17 0

All departments 13 116 118 4 0

Anatomical Sciences Education JULY/AUGUST 2014 255

pairwise comparisons of themes were not significant,although again several trends toward significance were noted.

DISCUSSION

Limitations

Validity and reliability of the faculty responses must be con-sidered in relation to a variety of factors. Faculty may nothave accurately assessed the importance of a topic withintheir practice or misinterpreted the prompt. The length of thesurvey at over 250 questions may have caused fatigue andhabituation—that is, physicians may have selected the sameresponse for multiple questions without thoughtful considera-tion of the topic. Response bias may have influenced theoverall results; for example, responders may have dispropor-tionately consisted of those who feel strongly about theimportance of anatomy, or lack thereof. Unequal representa-tion of departments, such as Anesthesia and Critical Care ascompared to Medicine, may have skewed overall analysis infavor of those departments that are better represented. Fur-thermore, subspecialization within departments may haveinfluenced responses. For example, a relatively high represen-tation of general surgeons could easily have overwhelmed alow representation of orthopedic surgeons, thus increasingthe ratings of abdominal anatomy and decreasing ratings oflimb anatomy.

While the categorization of topics into modules was facili-tated by the regional organization of the lecture material,choice of categorization of topics into themes was at the dis-cretion of the authors. Although the survey did not enumer-ate all of the topics taught in the course, differences between

Figure 2.

Physician opinions of the importance of all anatomy topics. Box and whiskerplot of standardized summative scales, presented as Likert values, of theimportance of all anatomical topics across departments. Dashes within theboxes indicate the median. Circles indicate outliers greater than 1.5 times theinterquartile range. Ob/Gyn, Obstetrics and Gynecology.

Table 4.

Mean Rankings of Perceived Importance of Course Modules Within Departments

Department NFriedmanP-value

Agreement(Kendall’s W) Thorax Abdomen Pelvis

Headand neck

Upperlimb

Lowerlimb

Obstetrics andGynecology

9 <0.001 Very strong (0.778) 3.56 2.00 1.11 4.89 5.33 4.11

Neurology 10 <0.001 Strong (0.679) 3.90 5.30 5.00 1.60 2.20 3.00

Psychiatry 7 0.003 Strong (0.514) 3.50 4.36 3.36 1.00 4.50 4.29

Anesthesiaand Critical Care

12 <0.001 Moderate (0.448) 2.92 4.83 5.08 3.67 2.08 2.42

Surgery 11 0.042 Weak (0.256) 3.78a 1.94 3.17 4.78 3.78a 3.56

Pathology 8 0.111 Weak (0.224) 2.87 2.50 3.62 3.00 4.19 4.81

Pediatrics 7 0.281 Weak (0.179) 2.57 3.00 3.86 2.86 4.29 4.43

Medicine 6 0.410 Weak (0.168) 3.17 3.67 4.67 3.50 2.50 2.83

Radiology 11 0.725 Very weak (0.052) 3.73 3.27 3.95 3.32 2.86 3.86

All responders 84 0.554 Very weak (0.009) 3.41 3.51 3.70a 3.23 3.45 3.70a

Modules are ordered as they are presented in the course; lower values indicate higher importance.aModules were tied.

256 Orsbon et al.

the importance of those topics that were included in the sur-vey nonetheless exist. Notably, the course does not explicitlyteach anatomical knowledge as applied in the physical exam.Lastly, the generalizability of the study is limited, as the studywas conducted at a single institution and the survey wasbased off of a single anatomy curriculum.

Overall Analysis

Our results are consistent with the notion that a majority ofphysicians in a variety of specialties identify anatomy as oneof the most important basic sciences in medicine (Cottam,1999; Arr�aez-Aybar et al., 2010). As it stands, the anatomycourse at the University of Chicago teaches students topics ofpractical value to a variety of physicians. The vast majorityof topics had a median rating greater than or above moder-ately important, and the median rating of all responses wasvery important. Although only 5% of the topics had amedian rating of essential when all physicians’ responseswere pooled together, 65% of the topics taught in the coursewere given a median rating of essential by physicians in atleast one department, and 96% of the material is similarlyvery important to physicians in at least one department.Physicians must have had non-overlapping topics which theyrated as essential in order for such a discrepancy between themedian ratings of each department and the median ratingsfrom all physicians to occur.

With the exception of the psychiatrists, physicians in dif-ferent departments did not significantly differ in their opin-ions about the overall importance of anatomy in theirpractice, a noteworthy result given that other studies haveargued that differences between different types of physiciansdo exist (Cottam, 1999; Arr�aez-Aybar et al., 2010). However,these other studies did not ask about the importance of indi-vidual anatomical topics, but rather the overall importance ofanatomy. Notably, the radiologists and surgeons, who regu-larly utilize anatomical knowledge and whose perspectivesare commonly included in gross anatomy education (Cottam,1999), did not find anatomy to be significantly more impor-tant than other non-psychiatrist physicians.

Module Analysis

Across all physicians, no body region was considered moreimportant than others. However, the results support thecommon-sense notion that physicians’ opinions of the clinicalimportance of anatomy are more similar to those of membersof their own departments than other departments. Indeed,the departments that intuitively seem to have the mostobvious anatomical specialization (Obstetrics and Gynecol-ogy, Neurology, and Psychiatry) showed the strongest agree-ment about the rank order of the importance of bodyregions. While the highest ranking modules for these depart-ments are not surprising, the other highly rated modules offer

Figure 3.

Physician opinions of the importance of anatomy topics grouped by course modules. Box and whisker plot of standardized summative scales, presented as a Likertscale, regarding the importance of anatomy topics within modules, sorted by departments, with results of Kruskal-Wallis tests. Dashes within the boxes indicate themedian. Circles indicate outliers greater than 1.5 times the interquartile range, and stars indicate extreme outliers greater than three times the interquartile range.Ob/Gyn, Obstetrics and Gynecology.

Anatomical Sciences Education JULY/AUGUST 2014 257

further insight into the practical importance of anatomy. Pre-dictably, the obstetricians and gynecologists highly valuedpelvis topics, but they also highly valued abdomen topics.The neurologists found head and neck topics to be the mostimportant, but they also placed high importance on upperand lower limb. The anesthesiologists were one of the fewgroups to rate upper and lower limb topics more highly thanother body regions, although their intra-departmental agree-ment of the rank order of modules was moderate. The sur-

geons exhibited weak but nonetheless significant agreement,giving the most importance to abdomen topics. This is possi-bly due to a relative abundance of responses from generalsurgeons compared to other surgical subspecialties (Table 2)and the clinical emphasis of the Department of Surgery at theUniversity of Chicago. The physicians in the Departments ofPathology, Pediatrics, Medicine, and Radiology all exhibitednon-significant agreement about the order of importance ofmodules. It is possible this lack of agreement is due to the

Table 5.

Post Hoc Wilcoxon Signed Rank Comparisons if Course Themes

Module themes Ordinal rank 1 2 3 4 5 6 7

Medical imaging 1 1.000 <0.001a <0.001a <0.001a <0.001a <0.001a <0.001a

Function 2 1.000 0.078 0.068 0.003 < 0.001a < 0.001a

structure 3 1.000 0.851 0.075 < 0.001a < 0.001a

Vasculature 4 1.000 0.174 < 0.001a < 0.001a

Innervation 5 1.000 0.002a 0.051

Embryology 6 1.000 0.820

Lymph 7 1.000

aSignificant with Bonferroni correction.

Table 6.

Mean Rankings of Perceived Importance of Course Themes by Department

Department NFriedmanP-value

Agreement(Kendall’s W) Embryo Function Innervation Lymph

Medicalimaging Structure Vasculature

RadiationOncology

4 0.001 Very strong (0.915) 7.00 4.50 4.75 2.00 1.00 3.25 5.50

Psychiatry 7 <0.001 Very strong (0.872) 3.21 2.21 4.36 6.64 1.00 4.79 5.79

Surgery 8 <0.001 Very strong (0.608) 6.00 4.75 5.63 4.06 1.06 3.13 3.38

Medicine 6 <0.001 Very strong (0.746) 6.67 2.25 5.00 5.25 1.17 3.83a 3.83a

Anesthesiaand Critical Care

11 <0.001 Strong (0.642) 6.09 3.36 3.23 6.18 1.23 3.91 4.00

Neurology 9 <0.001 Strong (0.655) 5.94 3.11 2.56 6.06 1.61 5.11 3.61

Radiology 10 <0.001 Strong (0.575) 5.20 4.30 5.40 5.60 1.15 3.70 2.65

Pediatrics 7 0.003 Moderate (0.474) 4.71 3.00 4.21 6.64 1.86 3.71 3.86

Obstetrics and

Gynecology

9 0.004 Moderate (0.356) 4.22a 4.22a 5.44 4.39 1.28 4.00 4.44

Pathology 8 0.045 Weak (0.268) 3.81 4.38 5.69 4.69 2.25 3.13 4.06

All responders 80 <0.001 Moderate (0.385) 5.23 3.64 4.54 5.34 1.37 3.91 3.98

Note: Lower values indicate higher importance, with a minimum of 1.00 and a maximum of 7.00.aThemes were tied.

258 Orsbon et al.

subspecialization within these departments. A small samplesize could also result in a lack of a clear trend in agreement;however, the radiologists (N 5 12), who demonstrated theweakest agreement, had response rates comparable to depart-ments reporting significant agreement.

Theme Analysis

Physicians in all departments displayed significant differencesin the rank order of importance of the themes. Notably, thephysicians thought that medical imaging was significantlymore important than all other themes. The relatively lowerimportance of lymph among the non-radiation oncologistswas unexpected considering the utility of lymphatic knowl-edge in assessing the spread of cancer and assessing the etiol-ogy of edema (Zumwalt et al., 2007). The lower ratingsassigned to embryology topics are also interesting as embry-ology has been de-emphasized, in terms of teaching hours, inthe modern anatomy classrooms while research in embryo-logical/fetal development and fertility has flourished (Drakeet al., 2009).

Physicians showed more agreement about the rank orderof importance of themes across all modules than the rankorder of the importance of the modules themselves. This sug-gests that while some physicians thought specific moduleswere less important overall, they consistently ranked certainanatomical themes higher across all modules. For example,

neurologists, not surprisingly, thought that innervation wasthe most important theme after medical imaging, a trend thatnecessitated high rankings of importance to innervationtopics regardless of anatomical module. Although an insuffi-cient number of radiation oncologists responded to beincluded in departmental analysis, their very strong agree-ment (W 5 0.915) on the rank order of important themeswas noteworthy. This was also the only group to rate lymphtopics nearly as high as medical imaging, reflecting theimportance of lymphatics and imaging in planning a radia-tion therapy field (Zumwalt et al., 2007; Chino et al., 2011).As with the modules, psychiatrists exhibited exceptionalagreement about the rank order of the thematic elements ofanatomy. All psychiatrists rated medical imaging most highlyamong all themes; indeed, medical imaging was one of thefew domains in which psychiatrists did not have one of thelowest median ratings of importance.

Implications

The high level of agreement of the importance of medicalimaging across many specialties supports the continued inte-gration of medical imaging into pre-clinical curricula. How-ever, the lack of interdepartmental agreement in the face ofsignificant intradepartmental agreement warrants a closerlook at the emphasis that a course places on the materialtaught during a single, broad-scope, “one-size-fits-all”

Figure 4.

Physician opinions of the importance of anatomy topics grouped by course themes. Box and whisker plot of standardized summative scales, presented as a Likertscale, regarding the importance of anatomy topics within themes, sorted by departments, with results of Kruskal-Wallis tests. Dashes within the boxes indicate themedian. Circles indicate outliers greater than 1.5 times the interquartile range, and stars indicate extreme outliers greater than three times the interquartile range.Ob/Gyn, Obstetrics and Gynecology.

Anatomical Sciences Education JULY/AUGUST 2014 259

exposure to the material. If students simply do not have thetime to learn anatomy (or other basic sciences) in the depthnecessary for clinical proficiency in a given specialty duringthe pre-clinical years, then perhaps pre-clinical basic sciencecurricula should be designed to prepare students for the com-mon demands of a variety of specialties and develop the life-long learning habits that physicians use to refresh andmaintain their knowledge of anatomy. Students could thenmore deeply engage the more specialty-specific basic materialin more advanced courses in the clinical years in preparationfor safe practice during residency, as recommended by otherauthors (Turney, 2007; Ahmed et al., 2010; Lazarus et al.,2012). We suspect that many introductory anatomy curricula(and certainly ours at the University of Chicago) are designedto achieve this first goal, and that the concern over anatomi-cal knowledge expressed by some authors and residencydirectors reflects the paucity of mandatory and specialty-specific anatomy coursework for undergraduate medical stu-dents in their clinical years outside of non-surgical fields. Foran example of a required anatomy course within an optionalradiation oncology rotation and its efficacy, see Zumwaltet al. (2007) and Chino et al. (2011).

When developing such advanced coursework, the audiencemust be considered in its design. We have shown that theanatomical knowledge required for practice in various spe-cialties not only differs between specialties, but also thatthere exists a hierarchy of the importance of anatomicalknowledge within specialties. That is, the importance of dif-ferent areas of anatomical knowledge depends on the typesof physicians who wield them. Therefore, the design ofadvanced anatomy courses that prepare students for thedemands of their residency training and careers should con-sider the practical needs of each specialty rather than followa generalized format for all clinical-year students.

Many have debated whether anatomy should be taughtfrom regional or systemic perspective, and our findings showthat the benefits of either approach in preparing students formedical practice could be dependent on the student’s spe-cialty of choice. For example, based on the very strongdepartmental agreement about the rank order of modules butonly moderate agreement of themes within those modules,students entering Obstetrics and Gynecology may benefitmore from a principally regional approach to anatomy, givingparticular attention to the pelvis and the abdomen. In con-trast to a regional approach, students entering Neurologymay benefit from a combined regional and theme- orsystems-based approach, highlighting the head, neck, andlimbs as well as innervation of the whole body.

Furthermore, when considering the pre-clinical course, evi-dence of the few common threads of importance—such asthe pervasive importance of medical imaging as shownhere—is useful in the design of a curriculum that preparesstudents for the demands of different clinical perspectivesduring their medical school training, particularly as time allo-cated to first-year gross anatomy continues to decrease(Drake et al., 2009). Furthermore, the data show that anat-omy is very important to physicians in specialties other thanradiology and surgery, and perhaps their perspectives shouldbe included more frequently in first year anatomy courses.

The high importance assigned to medical imaging in oursurvey suggests that one of the most important ways in whichphysicians use anatomical knowledge is through medicalimaging. Our findings support the use of medical imaging invertically integrated curricula that revisit anatomy in the

depth and scope required for clinical applications, as our col-leagues have suggested elsewhere (Phillips et al., 2013). Spe-cific instruction in medical imaging improves performance inexaminations assessing knowledge of both medical imagingand anatomy (Erkonen et al., 1990, 1992, 2000; Lufler et al.,2010; Phillips et al., 2012), making it a promising mediumfor revisiting anatomy in the clinical years.

Future directions of this project include determining whatindividual topics are most important for students to masterduring the first year and what topics are most important tomaster during advanced coursework. Given the small samplesize of each department, replication of the study with a largersample from a variety of institutions is warranted.

CONCLUSIONS

To our knowledge, this study is the first to assess the valuethat physicians in different medical specialties assign to dis-tinct anatomical concepts. Our findings suggest that anatomi-cal knowledge is hierarchical within departments, but less soamong all physicians, making it difficult to determine a coreof essential anatomical knowledge that is practically impor-tant to a variety of physicians. Our results support the grow-ing call for vertical integration of anatomy across the pre-clinical and clinical years, thus allowing students to revisitthe anatomy that will be most important in their chosen spe-cialty. Anatomy presented through medical imaging is anexception in that all departments highly value the ability tointerpret major structures on radiographs, computed tomog-raphy scans, and magnetic resonance imaging. Specialty-specific courses should incorporate and emphasize medicalimaging of the anatomy most important to a given specialty.

ACKNOWLEDGMENTS

The authors thank Drs. Vineet Arora, Dana Edelson, andValerie Press for providing feedback on the project and par-ticipating in the pilot study; James O’Reilly, Ph.D., for hissupport in the development of the survey; to ChristopherStraus, M.D., for his development of the integrated medicalimaging material for the Pritzker School of Medicine grossanatomy course; to the many members of the focus groupand pilot study for their time and feedback: Drs. JeanneFarnan, Barrett Fromme, Shannon Martin, Monica Vela, andWendy Yanow, Ed.D.; to Andrew Phillips, M.D., M.Ed., forhis medical education research advice; and finally but by nomeans least, to all of the physicians who responded to thesurvey—the authors are deeply grateful for their time andconsideration.

NOTES ON CONTRIBUTORS

COURTNEY P. ORSBON, B.S., is a M.D./Ph.D. student atThe University of Chicago Pritzker School of Medicine andDepartment of Organismal Biology and Anatomy, The Uni-versity of Chicago, Chicago, Illinois. Her research interestsare in human anatomy, including its evolutionary history andits role in medical education and practice.

REBECCA S. KAISER, B.A. is a fourth-year medical stu-dent at The University of Chicago Pritzker School of Medi-cine, Chicago, Illinois. She is planning her residency indermatology, with particular interest in medically complexdermatology.

260 Orsbon et al.

CALLUM F. ROSS, Ph.D. is a Professor in the Depart-ment of Organismal Biology and Anatomy, The University ofChicago, Chicago, Illinois. He is the course director of thegross anatomy and embryology course at The University ofChicago Pritzker School of Medicine.

LITERATURE CITEDAbu-Hijleh MF. 2010. The place of anatomy in medical education: Guide sup-plement 41.1—Viewpoint. Med Teach 32:601–603.

Ahmed K, Rowland S, Patel V, Khan RS, Ashrafian H, Davies DC, Darzi A,Athanasiou T, Paraskeva PA. 2010. Is the structure of anatomy curriculumadequate for safe medical practice? Surgeon 8:318–324.

Arr�aez-Aybar LA, S�anchez-Montesinos I, Mirapeix RM, Mompeo-CorrederaB, Sa~nudo-Tejero JR. 2010. Relevance of human anatomy in daily clinicalpractice. Ann Anat 192:341–348.

Bergman EM, Prince KJ, Drukker J, van der Vleuten CP, Scherpbier AJ. 2008.How much anatomy is enough? Anat Sci Educ 1:184–188.

Bergman EM, van der Vleuten CP, Scherpbier AJ. 2011. Why don’t they knowenough about anatomy? A narrative review. Med Teach 33:403–409.

Bhangu A, Boutefnouchet T, Yong X, Abrahams P, Joplin R. 2010. A three-yearprospective longitudinal cohort study of medical students’ attitudes toward anat-omy teaching and their career aspirations. Anat Sci Educ 3:184–190.

Bland JM, Altman DG. 1997. Cronbach’s alpha. BMJ 314:572.

Bull S, Mattick K. 2010. What biomedical science should be included in under-graduate medical courses and how is this decided? Med Teach 32:360–367.

Chapuis P, Fahrer M, Eizenberg N, Fahrer C, Bokey L. 2010. Should there bea national core curriculum for anatomy? ANZ J Surg 80:475–477.

Chino JP, Lee WR, Madden R, Sims EL, Kivell TL, Doyle SK, Mitchell TL,Hoppenworth EJ, Marks LB. 2011. Teaching the anatomy of oncology: Evalu-ating the impact of a dedicated oncoanatomy course. Int J Radiat Biol Phys79:853–859.

Cluzeau F, Burgers J, Brouwers M, Grol R, M€akel€a M, Littlejohns P,Grimshaw J, Hunt C. 2003. Development and validation of an internationalappraisal instrument for assessing the quality of clinical practice guidelines:The AGREE project. Qual Saf Health Care 12:18–23.

Cottam WW. 1999. Adequacy of medical school gross anatomy education asperceived by certain postgraduate residency programs and anatomy coursedirectors. Clin Anat 12:55–65.

Craig S, Tait N, Boers D, McAndrew D. 2010. Review of anatomy educa-tion in Australian and New Zealand medical schools. ANZ J Surg 80:212–216.

DeFriez CB, Morton DA, Horwitz DS, Eckel CM, Foreman KB, Albertine KH.2011. Orthopedic resident anatomy review course: A collaboration betweenanatomists and orthopedic surgeons. Anat Sci Educ 4:285–293.

Drake RL. 2007. A unique, innovative, and clinically oriented approach toanatomy education. Acad Med 82:475–478.

Drake RL, McBride JM, Lachman N, Pawlina W. 2009. Medical education inthe anatomical sciences: The winds of change continue to blow. Anat Sci Educ2:253–259.

Erkonen WE, Albanese MA, Smith WL, Pantazis NJ. 1990. Gross-anatomyinstruction with diagnostic images. Invest Radiol 25:292–294.

Erkonen WE, Albanese MA, Smith WL, Pantazis NJ. 1992. Effectiveness ofteaching radiologic image interpretation in gross-anatomy: A long-term follow-up. Invest Radiol 27:264–266.

Erkonen WE, Vydareny KH, Sandra A, Ferguson KA, Kreiter CD. 2000. Inter-institutional study to compare the effectiveness of a radiology-anatomy moduleof instruction. Acad Radiol 7:700–704.

Fasel JH. 1998. Teaching of gross anatomy to medical undergraduates: Gen-eral practice as a guideline? A synopsis. J Anat 192:305–306.

Fasel JH, Bader C, Gailloud P. 1999. Anatomy teaching for medical under-graduates: General practice as a guideline? The brain. Clin Anat 12:115–119.

Fincher RM, Wallach PM, Richardson WS. 2009. Basic science right, not basicscience lite: Medical education at a crossroad. J Gen Intern Med 24:1255–1258.

Heylings DJ. 2002. Anatomy 1999-2000: The curriculum, who teaches it andhow? Med Educ 36:702–710.

Kilroy D, Driscoll P. 2006. Determination of required anatomical knowledgefor clinical practice in emergency medicine: National curriculum planningusing a modified Delphi technique. Emerg Med J 23:693–696.

Koens F, Custers EJ, ten Cate OT. 2006. Clinical and basic science teachers’opinions about the required depth of biomedical knowledge for medical stu-dents. Med Teach 28:234–238.

Lazarus MD, Chinchilli VM, Leong SL, Kauffman GL. 2012. Perceptions ofanatomy: Critical components in the clinical setting. Anat Sci Educ 5:187–199.

Leonard RJ, Acland RD, Agur A, Blevins CE, Cahill DR, Collins JD, Dalley AFII, Dolph J, Hagedoorn JP, Hoos PC, Jones DG, Mathers LH, McFee R, MenninSP, Negulesco JA, Nelson ML, Olson TR, Page DW, Pawlina W, Petterborg LJ,Price JM, Spielman JE, Younoszai R. 1996. A clinical anatomy curriculum forthe medical student of the 21st century: Gross anatomy. Clin Anat 9:71–99.

Leonard RJ, Acland RD, Cahill DR, Casey MA, Ger R, Gilroy AM, Guha SC,Pal SK, Rennie WP. 1999. The clinical anatomy of several invasive procedures.Clin Anat 12:43–54.

Leonard RJ, Hoos PC, Agur A, Gilroy AM, Lozanoff S, Nelson ML, NewmanLM, Petterborg, LJ, Rosenheimer J. 2000. A clinical anatomy curriculum forthe medical student of the 21st century: Developmental anatomy. Clin Anat13:17–35.

Leonard RJ, Blevins CE, Cahill DR, Hoos PC, Lozanoff S, Marks SC Jr,Mathers LH, Page DW, Pawlina W, Petterborg LJ. 2001. Clinical anatomy andthe physical examination part I: Thorax, abdomen, perineum, and pelvis. ClinAnat 14:332–348.

Louw G, Eizenberg N, Carmichael SW. 2009. The place of anatomy in medicaleducation: AMEE guide no 41. Med Teach 31:373–386.

Lufler RS, Zumwalt AC, Romney CA, Hoagland TM. 2010. Incorporatingradiology into medical gross anatomy: Does the use of cadaver CT scansimprove students’ academic performance in anatomy? Anat Sci Educ 3:56–63.

McHanwell S, Davies DC, Morris J, Parkin I, Whiten S, Atkinson M, Dyball R,Ockleford C, Standring S, Whiten S, Wilton J. 2007. A core syllabus in anatomy formedical students—Adding common sense to need to know. Eur J Anat 11:S3–S18.

Older J. 2004. Anatomy: A must for teaching the next generation. Surgeon 2:79–90.

Pabst R. 2009. Anatomy curriculum for medical students: What can be learnedfor future curricula from evaluations and questionnaires completed by stu-dents, anatomists and clinicians in different countries? Ann Anat 191:541–546.

Phillips AW, Smith SG, Ross CF, Straus CM. 2012. Improved understanding ofhuman anatomy through self-guided radiological anatomy modules. AcadRadiol 19:902–907.

Phillips AW, Smith SG, Straus CM. 2013. The role of radiology in preclinicalanatomy: A critical review of the past, present, and future. Acad Radiol 20:297–304.

Schmidt RC. 1997. Managing Delphi surveys using nonparametric statisticaltechniques. Decis Sci 28:763–774.

Turney BW. 2007. Anatomy in a modern medical curriculum. Ann R Coll SurgEngl 89:104–107.

Wadey VM, Halpern J, Younger AS, Dev P, Olshen RA, Walker D. 2007.Orthopaedic surgery core curriculum: Foot and ankle reconstruction. FootAnkle Int 28:831–837.

Ward PJ. 2011. First year medical students’ approaches to study and their out-comes in a gross anatomy course. Clin Anat 24:120–127.

Ward PJ, Walker JJ. 2008. The influence of study methods and knowledgeprocessing on academic success and long-term recall of anatomy learning byfirst-year veterinary students. Anat Sci Educ 1:68–74.

Zumwalt AC, Marks L, Halperin EC. 2007. Integrating gross anatomy into aclinical oncology curriculum: The oncoanatomy course at Duke UniversitySchool of Medicine. Acad Med 82:469–474.

Anatomical Sciences Education JULY/AUGUST 2014 261