Journal of Hospital Infection (2008) 69, 33e38

Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

Successful control of an outbreak of invasiveaspergillosis in a regional haematologyunit during hospital construction works*

C.C. Chang a, A.C. Cheng a, B. Devitt b, A.J. Hughes a,P. Campbell b, K. Styles c, J. Low c, E. Athan a,*

a Department of Infectious Diseases, Geelong Hospital, Geelong, Victoria, Australiab Clinical Haematology Unit, Geelong Hospital, Geelong, Victoria, Australiac Infection Prevention Unit, Geelong Hospital, Geelong, Victoria, Australia

Received 19 November 2007; accepted 7 February 2008Available online 3 April 2008

KEYWORDSAspergillosis;Outbreak;Construction;Haematology;Voriconazole

* Presented in part at the Australasi* Corresponding author. Address: D

Australia. Tel.: þ61 3 5260 3205; fax:E-mail address: eugene@barwonhe

0195-6701/$ - see front matter CrowAll rights reserved.doi:10.1016/j.jhin.2008.02.010

Summary We report the successful control of an outbreak of six cases ofnosocomial invasive aspergillosis (IA) in our haematology unit coincidingwith major hospital construction works. Infection control changes includedunit relocation, impermeable barriers at construction site, face-maskingand voriconazole prophylaxis of 18 further high-risk patients, none ofwhich developed breakthrough IA. A multi-faceted pre-emptive approachinvolving clinicians, hospital management, engineering and buildingdepartments is essential in preventing nosocomial IA outbreaks.Crown Copyright ª 2008 Published by Elsevier Ltd on behalf of The HospitalInfection Society. All rights reserved.

Introduction

Invasive aspergillosis (IA) remains the leadinginfective cause of infectious mortality in mosthaematology and bone marrow transplant units.1

Hospital building works is an established risk factorfor invasive fungal infections (IFI).2,3 Risk stratifi-cation for IFI by host-related factors is well-

an Society of Infectious Diseasepartment of Infectious Diseþ61 3 5260 3040.alth.org.aun Copyright ª 2008 Publishe

established.4 Less attention has been given toinstitutional risk factors such as the presence ofconstruction works. We report an outbreak of sixcases of IA in our haematology unit and describethe successful implementation of environmentalmeasures and voriconazole prophylaxis.

es (ASID) meeting, Hobart, Tasmania, 23 March 2007.ases, Geelong Hospital, Ryrie Street, Geelong, Victoria 3220,

d by Elsevier Ltd on behalf of The Hospital Infection Society.

34 C.C. Chang et al.

Methods

Setting

Geelong Hospital is a 406-bed regional referralcentre in Victoria, Australia. The haematologyeoncology service has been in existence for >20years and manages w40 patients annually withacute leukaemia, high-grade lymphoma andpatients undergoing autologous stem cell trans-plantation. A small number of allogeneic trans-plantation recipients return to our centrefollowing their transplantation elsewhere. Priorto this outbreak, fluconazole 200 mg daily wasadministered to patients with acute leukaemiaundergoing induction or consolidation treatmentand to patients receiving aggressive therapy forhigh-grade lymphoma. A search and chart reviewof medical records for haematologyeoncologypatients revealed no cases of IA (ICD-10 codeB44.0) in the preceding eight years.

Ambulatory care occurs within a comprehensiveday-oncology facility linked to the main hospitalbuilding by a 200 m open breezeway (Figure 1).The inpatient haematology ward is located onthe sixth floor of the main building with five four-bed rooms and four single rooms. The inpatientward is serviced by standard air filtration withpre-filters to remove large particulate matterusing standard deep-bed filters, which are washedmonthly and replaced annually. There is no high-efficiency particulate air (HEPA) filtration in theinpatient or day-oncology ward.

As part of the service expansion, hospital con-struction works in and around the vicinity of the day-oncology building commenced in early 2006. Majorexcavation and demolition works were planned forJune 2006. A transfer of the day-oncology service tothe main hospital building was to precede this;however, construction works unexpectedly ranahead of schedule before this could occur.

Definitions

We assessed cases of IA according to EuropeanOrganisation for Research and Treatment of Cancer(EORTC) definitions.5 Proven IFI was defined ashistopathologic or cytopathologic examinationshowing hyphae from needle aspiration or biopsyspecimen with evidence of associated tissue dam-age, or a positive culture result for a sampleobtained by sterile procedure from normally sterileand clinically or radiologically abnormal site consis-tent with infection. Probable IFI required at leastone each of host factor and microbiological

criterion and one major (or two minor) clinicalcriteria from abnormal site consistent with infec-tion. Possible IFI was defined as at least one hostfactor criterion and one microbiological or onemajor (or two minor) clinical criteria.

Galactomannan assays are not widely availablein Australia outside of research studies. Polymer-ase chain reaction (PCR) for Aspergillus wasperformed at Westmead Hospital, Australia, usinga nested qualitative real-time Aspergillus-specificPCR assay with a LightCycler (LC) system (RocheDiagnostics, Sydney, Australia).6

Following identification of the outbreak, caseswere identified prospectively by active surveillanceand supplemented by a review of hospital records ofhaematology patients treated prior to the outbreak.We considered patients as at high risk for IFI if theyhad acute leukaemia, were undergoing transplan-tation and those predicted to develop prolongedneutropenia. Air-sampling was performed usinga MAS-100 air-filter (EMD Chemicals Inc., Merck PtyLtd, Kilsyth, Australia) whereby 500 L of air wassampled for 5 min each, in the day-oncology unit, in-patient ward and pharmacy department, collectedon horse blood agar and incubated at 37�C for 48 h.

Outbreak description and management

An outbreak was declared on 23 June 2006 afterthe identification of a cluster of two proven, twoprobable and one possible diagnosis of IA over twomonths (Table I). Urgent steps were taken includ-ing immediate relocation of the day-oncologyservices for high-risk patients to a vacant ward inthe main hospital building on 4 July 2006 (Figure 2).Regular building site inspections were held toensure that sites were sealed with impermeablebarriers, doors and windows remained shut, andcleaning was adequate (e.g. wet mop surfaces,vacuum carpeted areas, debris removal). Patientsin the haematologyeoncology services were givenwritten notification, advised to avoid buildingconstruction sites and to use N95 face maskswhen they were out of the inpatient ward whileon hospital premises. The patient car park wasrelocated away from the construction site.

Weekly meetings were held between infectioncontrol, cancer services, hospital management,building and engineering departments to facilitateupdates and communication. Air sampling in theinpatient facility, day-oncology and pharmacy didnot reveal any fungal growth.

Construction works continued from June tillNovember 2006. In view of the unprecedentedhigh incidence, morbidity and mortality rate of

Open walkway(6th floor: Oncology inpatient

services)

Main hospital

Dayoncology

Psychiatryunit New

wingof

mainhospital

1

Engineering and allied healthdepartments

4

3

Helipad

2

Oncologypatient

car park

Figure 1 Schematic outline of hospital layout. (1) Endoscopy suite; (2) finance department; (3) library; (4) stafflounge.

Aspergillosis outbreak during building works 35

IFI, during this period, voriconazole prophylaxis(loading 400 mg for two doses followed by 200 mgtwice daily orally) was administered to patientsstratified as ‘high risk’ by their treating clinicians.Patients receiving vinca alkaloids and cyclophos-phamide were administered thrice-weekly prophy-lactic liposomal amphotericin at 3 mg/kg in placeof voriconazole due to potential pharmacokineticinteractions. Those intolerant to amphotericinB were administered fluconazole and vigilantlymonitored for development of IFI. Three patientsreceived amphotericin B prophylaxis.

None of the 18 patients placed on voriconazole oramphotericin B prophylaxis developed IFI. How-ever, one man aged 75 years with chronic lympho-cytic leukaemia treated with methylprednisoloneand rituximab, not stratified as high-risk, wasdiagnosed with disseminated IA at post mortemafter presenting with confusion and fevers on 9July 2006.

Day-oncology was relocated to the new facili-ties on 11 December 2006 after completion ofconstruction works and five negative on-site air-samples were obtained. Fluconazole 200 mg daily

Table I Summary of invasive aspergillosis cases

Case EORTCdefinition

Underlyingdisease

Chemotherapy Neutropenicdays (<0.5)

Respiratorysymptoms

CT chest Mi biology/mole lar results

Outcome

1 Probable De-novoAML

Cycle 1consolidationFLAG-Ida

13 Moderate Nodule andinterstitialchange

Sputum: A. migatusDelayed br hoscopynegative.

Alive and in remission

2 Proven TransformedAML

InductionFLAG-Ida

20 None Smallperipheralnodule

CT-guided psy:fungal hyphresemblingAspergillus

Completed one cycleof reduced doseconsolidation.Alive and in remission

3 Proven De-novoAML

InductionBig-ICE

12 Mild Smallperipheralnodule

CT-guided psy:fungal hyphresembling pergillus,and Asperg s-positivePCR

Completed two cyclesof consolidation. Aliveand in remission

4 Possible De-novoAML

InductionBig-ICE

23 Severe Multiplebilateralnodules

Refused br hoscopyand biopsy til 1 and 4weeks post piricaltreatment(both nega )

Deceased withGram-negativesepsis in secondconsolidationcycle

5 Probable Advancedmultiplemyeloma

Thalidomide 65 Severerespiratoryfailure

Extensivebilateralinterstitialchange

Sputum:A. fumigat

Palliated

6 Proven CLL Methylprednisoloneand thalidomide

2 Mild (butsevereneurologicalchange)

CXR: RLLpatchychange

Post-mortedisseminataspergillosi

Deceased. Day 6admissiondespite antifungaltreatment

7 Proven DLBCL R-CHOP andintrathecalmethotrexate

5 None Soft tissuemass inmediastinum

Hyphae see nmicroscopy(not sent f ulture).Aspergillus gativePCR

Successfullyunderwentallotransplant

EORTC: European Organisation for Research and Treatment of Cancer; CT: computed tomography; FLAG-Ida: fludarabine; Ara-C arubicin, G-CSF; Big-ICE: idarubicin, high-dose Ara-C,etoposide; R-CHOP: rituximab, cyclophosphamide, doxorubicin, vincristine, prednisolone; PCR, polymerase chain reaction; PET n: positive emission tomography; AML: acute myeloidleukaemia; CLL: chronic lymphocytic leukaemia; DLBCL: diffuse large B-cell lymphoma; CXR: chest radiograph; RLL: right low lobe.

36C

.C.

Chang

et

al.

crocu

fuonc

bioae

bioaeAsillu

oncun-em

tive

us

m:eds

n o

or c-ne

: idscaer

24/0

4/20

0622

/05/

2006

19/0

6/20

0617

/07/

2006

14/0

8/20

0611

/09/

2006

9/10

/200

66/

11/2

006

4/12

/200

61/

01/2

007

29/0

1/20

0726

/02/

2007

26/0

3/20

0723

/04/

2007

21/0

5/20

0718

/06/

2007

16/0

7/20

07

0

1

Wee

kly

case

s

2

3Outbreak declared.Meetings held;environmental measuresand voriconazoleprophylaxis implemented.

(Case 7: not onvoriconazoleprophylaxis)

Day-oncology servicerelocated to newfacilities. Reverted tofluconazole asantifungal prophylaxis.

Relocation ofday-oncologyservice to mainbuilding

Proven

Probable

Possible

Figure 2 Timeline of invasive aspergillosis outbreak.

Aspergillosis outbreak during building works 37

was re-established as routine antifungal prophy-laxis. Three of the five patients involved in theoutbreak remain alive at end of the currentfollow-up period, 31 July 2007. In March 2007,a further patient with non-Hodgkin’s lymphomawas diagnosed with IA from a mediastinal biopsy enotably, part of his earlier chemotherapy coin-cided with the IA outbreak period.

Discussion

Hospital building works are an ever-present phenom-enon in modern-day healthcare.7,8 Current guide-lines call for management plans to be implementedduring times of construction, and our experiencehighlights the need for communication betweenclinical, management and engineering services atthe hospital.2,9e13 Effective measures include relo-cation of high-risk patients, masking, wet cleaning,reducing unnecessary traffic through affected areas,environmental sealing of construction areas, adviceto patients and air filtration.2,9,13,14

There are conflicting data on the efficacy ofHEPA filters in outbreak control.15e17 We did notelect to install HEPA filters during this outbreak.HEPA filters are complex to install and maintain,and suboptimal maintenance can lead to outbreaksof aspergillosis.7,11 A meta-analysis of 16 con-trolled trials in high-risk haematology patients

demonstrated ambiguous findings, with a reductionin IFI but no difference in mortality, in studies withconsiderable clinical heterogeneity and evidenceof publication bias.18

The role of air sampling for fungal spores alsoremains controversial. Outbreaks of IFI with neg-ative air-sampling data have been described pre-viously.8 The incubation period for IFI remainsunknown and single-point air-sampling may misspeaks in fungal spores. Further, the lack of stan-dardized methods to perform air sampling makesinterpretation difficult.

We used voriconazole as antifungal prophylaxisfor high-risk patients during the outbreak in viewof its activity against Aspergillus spp., its avail-ability in oral and intravenous formulation andease of dosing. Voriconazole as primary prophy-laxis is yet to be supported by clinical trials, butit has clear efficacy against IA and in secondaryprophylaxis.19 This outbreak occurred prior tothe publication of the posaconazole primary pro-phylaxis papers.20,21

Criteria for assessing patients at high risk of IFIare generally based on patient characteristics,such as underlying disease, degree of immunosup-pression, chemotherapy utilised and duration ofneutropenia.4,22 It is important also to recognise en-vironmental factors, reflected in the baseline insti-tutional rate of IFI and modifying factors such asbuilding works. We believe that the threshold for

38 C.C. Chang et al.

antifungal prophylaxis should be lowered in the set-ting of hospital construction works. The relativeconvenience and safety of newer oral antifungalagents make this possible. Although we did not per-form a formal cost analysis, the expense of thesenewer antifungal agents remains an issue (the costof voriconazole alone for prophylaxis and treatmentduring this outbreak approached A$160,000), butshould be weighed against the high cost and morbid-ity associated with IA.

In conclusion, we report the first publishedexperience of the use of voriconazole as antifungalprophylaxis in the setting of building works and anoutbreak of IA. Hospital construction work is a clearrisk factor for development of IFI. Clinicians, hos-pital management, building and engineering staffshould have a strategic plan to reduce this risk.Ongoing vigilance and interdepartmental commu-nication is important. The threshold for antifungalprophylaxis should be lowered. There is a pressingneed for further research and management guide-lines for prevention of IA in hospital building works.

Acknowledgements

We thank O. Harris, M. Chan and D. Alfredson(Microbiology Department, St John of God) andMicrobiology Department, Westmead Hospital fortheir invaluable assistance.

Conflict of interest statementNone declared.

Funding sourcesNone.

References

1. Marr KA, Boeckh M, Carter RA, Kim HW, Corey L. Combina-tion antifungal therapy for invasive aspergillosis. Clin InfectDis 2004;39:797e802.

2. Guidelines for preventing opportunistic infections amonghematopoietic stem cell transplant recipients. Recommen-dations of CDC, the Infectious Disease Society of America,and the American Society of Blood and Marrow Transplanta-tion. Morb Mortal Wkly Rep 2000;49:1e128.

3. McCann S, Byrne JL, Rovira M, et al. Outbreaks of infectiousdiseases in stem cell transplant units: a silent cause ofdeath for patients and transplant programmes. Bone Mar-row Transplant 2004;33:519e529.

4. Prentice HG, Kibbler CC, Prentice AG. Towards a targeted,risk-based, antifungal strategy in neutropenic patients. Br JHaematol 2000;110:273e284.

5. Ascioglu S, Rex JH, de Pauw B, et al. Defining oppor-tunistic invasive fungal infections in immunocompromisedpatients with cancer and haematopoietic stem cell

transplants: an international consensus. Clin Infect Dis2002;34:7e14.

6. Halliday C, Wu QX, James G, Sorrell T. Development ofa nested qualitative real-time PCR assay to detect Aspergil-lus species DNA in clinical specimens. J Clin Microbiol 2005;43:5366e5368.

7. Streifel AJ. In with the good air. Infect Control Hosp Epi-demiol 2002;23:488e490.

8. Vonberg RP, Gastmeier P. Nosocomial aspergillosis in out-break settings. J Hosp Infect 2006;63:246e254.

9. Centers for Disease Control and Prevention. Guidelines forenvironmental infection control in health-care facilities:recommendations of CDC and the Healthcare InfectionControl Practices Advisory Committee (HICPAC). Morb MortWkly Rep 2003;52:1e48.

10. Policy Number 10.4. Construction Renovation and Main-tenance. Infection Prevention and Control Manual. 2ndedn. Victoria: Rural Infection Control Practice Group; 2005.

11. Anderson K, Morris G, Kennedy H, et al. Aspergillosis inimmunocompromised paediatric patients: associations withbuilding hygiene, design, and indoor air. Thorax 1996;51:256e261.

12. HCAMC in association with UNSW. Part D: infection Controland Prevention. Australasian Health Facility Guidelines[Revision v.1.0]; 2007. p. 637e660.

13. Sehulster LM, Chinn RYW, Arduino MJ, et al. Guidelines forenvironmental infection control in health-care facilities.Recommendations from CDC and the Healthcare InfectionControl Practices Advisory Committee (HICPAC). Chicago:American Society for Healthcare Engineering/AmericanHospital Association; 2004.

14. Fitzpatrick F, Prout S, Gileece A, Fenelon LE, Murphy OM.Nosocomial aspergillosis during building work e a multidisci-plinary approach. J Hosp Infect 1999;42:170e171.

15. Hahn T, Cummings M, Michalek A, Lipman B, Segal B,McCarthy PL. Efficacy of high-efficiency particulate airfiltration in preventing aspergillosis in immunocompromisedpatients with haematologic malignancies. Infect ControlHosp Epidemiol 2002;23:525e531.

16. Loo VG, Bertrand C, Dixon C, et al. Control of construction-associated nosocomial Aspergillosis in an antiquated hae-matological unit. Infect Control Hosp Epidemiol 1996;17:360e364.

17. Opal SM, Asp AA, Cannady Jr PB, Morse PL, Burton LJ,Hammer II PG. Efficacy of infection control measures duringa nosocomial outbreak of disseminated aspergillosis associ-ated with hospital construction. J Infect Dis 1986;153:634e637.

18. Eckmanns T, Ruden H, Gastmeier P. The influence of high-efficiency particulate air filtration on mortality and fungalinfection among highly immunocompromised patients: a sys-temic review. J Infect Dis 2006;193:1408e1418.

19. Cordonnier C, Maury S, Pautas C, et al. Secondary antifun-gal prophylaxis with voriconazole to adhere to scheduledtreatment in leukemic patients and stem cell transplantrecipients. Bone Marrow Transpl 2004;33:943e948.

20. Ullmann AJ, Lipton JH, Vesole DH, et al. Posaconazole orfluconazole for prophylaxis in severe graft-versus-hostdisease. N Engl J Med 2007;356:335e347.

21. Cornely OA, Maertens J, Winston DJ, et al. Posaconazole vs.fluconazole or itraconazole prophylaxis in patients withneutropenia. N Engl J Med 2007;356:348e359.

22. McLintock LA, Jornaides NE, Allan EK, et al. The use of a riskgroup stratification in the management of invasive fungalinfection: a prospective validation. Br J Haematol 2004;124:403e404.