Health professions and risk of sporadic Creutzfeldt – Jakob disease , 1965 to 2010

E Alcalde-Cabero1, J Almazán-Isla1, J P Brandel2, M Breithaupt3, J Catarino4, S Collins5, J Haybäck6, R Höftberger7, E Kahana8, G G Kovacs7,9, A Ladogana10, E Mitrova11, A Molesworth12, Y Nakamura13, M Pocchiari10, M Popovic14, M Ruiz-Tovar1, A L Taratuto15, C van Duijn16, M Yamada17, R G Will12, I Zerr3, J de Pedro Cuesta (jpedro@isciii.es)1 1. National Centre of Epidemiology Consortium for Biomedical Research in Neurodegenerative Diseases (Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas – CIBERNED), Carlos III Institute of Health, Madrid, Spain 2. Institut National de la Santé et de la Recherche Médicale (INSERM) UMRS 975, National CJD Surveillance Network, Assistance publique Hôpitaux de Paris (APHP), National Reference Centre for CJD, Pitié-Salpêtrière Hospital Group, Paris, France 3. Department of Neurology, National Reference Centre for TSE, Georg-August University, Göttingen, Germany 4. Alameda Epidemiology and Health Statistics Department, Lisbon, Portugal 5. Department of Pathology, University of Melbourne, Melbourne, Australia 6. Institute of Neuropathology, Zurich University Hospital, Zurich, Switzerland 7. Institute of Neurology, Vienna Medical University, Vienna, Austria 8. Department of Neurology, Barzilai Medical Centre, Ashkelon, Israel 9. National Reference Centre for Human Prion Diseases, Semmelweis University, Budapest, Hungary 10. Department of Cell Biology and Neurosciences, Health Institute, Rome, Italy 11. Department of Prion Diseases, Slovak Medical University Research Base, Bratislava, Slovakia 12. National CJD Research and Surveillance Unit, Western General Hospital, Edinburgh, United Kingdom 13. Department of Public Health, Jichi Medical University, Shimotsuke, Japan 14. nstitute of Pathology, Medical Faculty, University of Ljubljana, Ljubljana, Slovenia 15. Department of Neuropathology/FLENI, Referral Centre for CJD and other TSEs, Institute for Neurological Research, Buenos Aires, Argentina 16. National Surveillance of CJD, Erasmus MC, Rotterdam, The Netherlands 17. Neurology Department, Kanazawa University Hospital, Kanazawa, Japan


Introduction
Creutzfeldt-Jakob disease (CJD) is a fatal neurodegenerative disease characterised by deposition of a pathological isoform of the normal cellular prion protein (PrP C ) [1].The annual CJD incidence worldwide is 1-2 per million population [2].CJD exists in various forms: genetic, caused by mutations in the PRNP gene encoding PrP C , acquired (variant and iatrogenic) and sporadic.Most cases have sporadic CJD (sCJD) -the cause of which is unknown.Occupational risk related to sCJD has been assessed in several case-control studies as a secondary study objective, with inconsistent results [3][4][5][6][7] and there have been occasional reports of health professionals with sCJD [8][9][10][11][12].
Occupation has not been included as a variable in all CJD surveillance protocols [13].Nonetheless, there is concern about potential occupational excess risk of sCJD among health professions, as shown by a recent study on guidelines in European Union (EU) Member States and Norway for the prevention of CJD transmission in medical settings.This study showed that 12 of the 17 contributing countries had specific recommendations targeted at minimising occupational exposure; eight of the 12 had systems for reporting or registering work-related incidents at hospitals or laboratories [14].
In March 2009, a CJD case was reported to the Spanish CJD registry, who was classified as having sporadic CJD.As the patient was an experienced general pathologist and neuropathologist, it was speculated that the disease might have been a result of the person's professional activities.The event was commented on in medical, scientific and mass media in Spain and elsewhere, e.g.[15].The patient died after a four-month disease course, characterised mainly by cognitive decline, ataxia and myoclonus.The disease prion protein subtype, i.e. strain, was confirmed histochemically and biochemically as MM1, the most common subtype [2].Risk factors for developing CJD, including blood transfusion, iatrogenic exposure (e.g. to dura mater, cadaveric pituitary-derived growth hormone) and mutations in the PRNP gene, were not identified.Assessment of the patient's routine hospital work indicated that the patient had had a history of minor injuries during postmortem examinations (personal communication, E. García-Albea, April 2009).
Following notification of this patient, the Spanish registry circulated a request for information to each national surveillance team participating in the European Creutzfeldt Jakob Disease Surveillance Network (EuroCJD), which dates back to 1993 and currently encompasses 25 collaborating centres in EU Member States and European Free Trade Association (EFTA) countries and a further eight in countries around the world, including Australia, Canada and Japan [16].These centres provide data from national registries either through the EuroCJD website or, as with Japanese data, at regular network meetings.The request asked for the following: (i) information on the diagnosis (year of birth and death, sex and place of residence) of reported cases of sporadic CJD among active or retired pathologists from 1996 onwards; and (ii) comments based on personal experience of occupational risk and CJD among health professionals, including technicians working at pathology laboratories.
There has been limited systematic research targeted at identifying occupational risk factors for sCJD in healthcare settings.This paper reports on the data supplied to the Spanish CJD registry in response to the request, and on the results of two literature reviews of sCJDone on case reports involving health professionals and the other on epidemiologically assessed healthcarerelated occupational risk of sCJD.

Individualised occupational data from national CJD surveillance teams
The Spanish CJD registry obtained answers in English to at least one of the requests for information from 21 national surveillance teams.The amount of information provided varied: in general, only data that had already been registered was reported; with regard to occupational history in CJD -recorded by profession or activity branch -several countries provided information on people in whom CJD had been excluded or on controls.
The data received were divided into two groups, for further analysis -one describing health professionals who were sCJD cases and the other describing health professionals among controls or non-cases.We did not attempt a formal epidemiological assessment of healthcare-related occupational risk of sCJD based on this information, for instance using a case-control design.

Case reports of sCJD among health professionals
Countries with available registry data on cases' occupations sent individualised data on physicians with neuropathologically confirmed or probable sCJD or other types of CJD [17,18].Some countries provided such data on other health professions.In the few instances in which occupation as a pathologist was identified, professional experience or job duration at a laboratory or department was specified.The results were tabulated, using the original definitions from the countries'

Box
Search terms used in first step of two literature searches on sporadic Creutzfeldt-Jakob disease (sCJD) in health professionals and analytical studies on occupational risk of sCJD for health professions and selection criteria used in a second step, reported 1 January 1989-1 October 2011

MEDLINE
The search strategy was based on the following medical subject headings (MeSH) terms: • prion diseases/prions/Creutzfeldt-Jakob syndrome; and • health occupations/allied health occupations/ occupational groups/occupations/occupational dentistry/case control studies.

Embase
The search strategy was based on the following Emtree thesaurus terms: • prion/prion disease/prion protein/Creutzfeldt Jakob disease/ Creutzfeldt Jakob disease agent; and reports.No standard occupational classification was used for grouping response results and each case was assigned to one occupational category.Frequently, the occupational categories corresponded to a combination of professional profiles, e.g.specialities and work types (clinical, administrative, laboratory, etc.).In such cases, the category most likely to involve direct contact with human tissue or patients was selected.
Healthcare-related occupations among controls and non-cases Some CJD surveillance teams with a sufficient sample size provided data on occupation of people with suspected sCJD who were finally classified as not having CJD (non-cases) and also of those in control groups.
Five EuroCJD countries with large populations -Germany, Italy, Japan, Spain and the United Kingdom (UK) -provided this type of data, both published and unpublished.These countries supplied data on physicians who were controls Italy and Japan also provided information on other health professionals who were non-cases.Information on different categories of health professionals was available for British controls.Occupation was usually categorised on the basis of original records at registries.In a few instances, reporting physicians or relatives were consulted about the predominant activity, e.g.general practice vs radiology, of the non-cases.

Literature reviews
The first step in the literature reviews sought to identify reports of sCJD among health professionals, whether reported as case studies or drawn from analytical studies published during 1 January 1989 to 1 October 2011.We carried out several searches in MEDLINE and Embase using the medical subject headings (MeSH) and Emtree thesaurus terms listed in the Box, to identify case studies on CJD in health professionals and

Diff erent documents identifi ed n=715
Failed to meet inclusion criteria n=671 No prion/CJD disorders n=170

Not occupation related n=214
Hospital/public health n=89

Secondary publication n=198
Potentially valid reports included f or f ull-text review n=44  Reviewed reports on occupational risk assessment n=9 Excluded as the study focused on non-occupational exposures (genetic, biochemical, diagnostic, etc.) n=33 Reports possibly addressing but not reporting results on healthcare-related occupations n=4 Risk not assessed due to lack of health professions among cases n=1 Reports providing valid risk-based data for healthcare-related occupations n=4

Figure 2
Literature review of analytical studies on occupational risk of sporadic Creutzfeldt-Jakob disease for health professionals, 1982-1 October 2011 analytical studies on occupational risk of CJD for health professions.The initial searches yielded a total of 715 different documents.
In a second step, two independent assessors applied predefined sets of inclusion and exclusion criteria (Box) to the titles of the retrieved documents or, where available, to their abstracts.
Documents that met the inclusion criteria were processed further for full-text analysis in order to obtain the case description or to assess health-related occupational risk of sCJD.
After the selection criteria for had been independently, though not always unanimously, applied to the 715 documents by two reviewers, EAC and JPC, 671 were rejected and 44 selected for further analysis by both reviewers (Figures 1 and 2).

Case reports of sCJD among health professionals
Of the 44 documents selected for full-text review, 34 were excluded as the studies did not examine healthrelated occupations (Figure 1).Four studies that failed to include specific categories of health professionals with sCJD or in which the diagnosis of CJD was not validated were also excluded [3,6,7,19].Six studies -four case reports [10][11][12]20] and two case-control studies, which provided information on health-related occupations in sCJD case series [5,21] -were selected for data extraction.Five case studies and one case-control study retrieved from personal records before 1989 were also included [4,8,9,17,22,23].Thus, the final analysis of 12 reports included data on individual health professionals from case reports [8][9][10][11][12]17,20,22,23] and numbers of health professionals with sCJD from three reports on case-control studies [4,5,21].These 12 reports included sCJD cases fulfilling diagnostic criteria for neuropathologically confirmed sCJD or for probable sCJD (people in the latter category were only included in case-control studies) [5,24].Where health professions were listed in the case series of a large case-control study and numbers were not reported, only one individual, e.g. a dentist, was counted [5].

Epidemiologically assessed healthcarerelated occupational risk of sCJD
Of the 44 documents selected for full-text review (the same 44 mentioned above), 40 fulfilled the inclusion criteria.After analysis of the texts, seven analytical studies on occupation and risk of sCJD remained for potential data extraction [3,[5][6][7]19,21,25].Those excluded were multipurpose case-control investigations that made no mention of occupation in the results, occupation-unrelated meta-analyses, genetic case-control studies and public-health occupational profiles derived from empirical data.Four analytical studies reported before 1989 were reviewed: two were included [4,26] and two rejected [27,28].Nine documents [3][4][5][6][7]19,21,25,26] provided data on occupational risk but only five of these addressed healthcare-related occupations [5][6][7]19,26].Due to the low numbers (absence of exposed cases) in one study [26], riskbased data for health professions were only available from four case-control studies [5][6][7]19].Reported associations for healthcare-related occupational risk obtained from these four epidemiological studies and raw negative findings from the above-mentioned study [26] were tabulated.

Individualised occupational data from national CJD surveillance teams
Health professionals among registered sCJD cases A total of 202 health professionals were listed among 8,321 cases of sCJD registered by 21 respondent countries participating in EuroCJD (Table 1).Of these, 65 (32%) were physicians and 137 were other healthcare workers.The highest numbers by medical speciality were general practitioners (n=9), surgeons (n=7), internists (n=7), dentists (n=4), ophthalmologists (n=3) and pathologists (n=2).The proportion of physicians or dentists among all registered sCJD cases was 65/8,321 (0.8%).

Health professionals among non-cases or controls
Table 2 shows individual data reported for health professions among non-cases or controls in five countries (Germany, Italy, Japan, Spain and the UK).Among 83 healthcare workers, 15 were physicians, six of whom had unknown specialisations, and three were surgeons.The percentage of physicians and dentists among CJD cases in Germany, Italy, Japan, Spain and the UK combined was 0.7% (34/4,949 (Table 1).This was similar to the proportion in the combined controls 0.5% (15/2,968).

Reported sCJD in health professionals
Individual occupational profiles of reported healthrelated professionals with sCJD are outlined in Table 3.

Health-related occupational risk of sCJD
The nine analytical papers on occupations and sCJD identified [3][4][5][6][7]19,21,25,26] tended to focus on healthcare and animal care-related occupations, with Cocco et al.'s study furnishing detailed data on other occupations [19].This study used a large number of non-validated CJD diagnoses from death records in the United States and controls selected after exclusion of persons with neurological diseases reported as the cause of death [19].The main findings for healthcare-related occupations from five papers are summarised in Table 4.While three of four studies on health professions did not demonstrate excess risk [5][6][7], statistically significant findings -for persons working at physicians' offices -were solely reported by Cocco et al. [19].

Discussion
Despite a number of case reports of sCJD in physicians and technicians, the findings of this EuroCJD survey do not suggest an increased risk of sCJD in health professionals, nor do analytical studies show a clear excess risk for health-related professions.Methodological limitations of analytical studies in which occupational data were frequently provided by informants who were probably aware of the sCJD diagnosis [3][4][5][6][7]26] argue in favour of a cautious interpretation of the positive association reported for persons working at physicians' offices [19].Consequently, the main finding of this literature review and complementary EuroCJD observation is that health professionals, including medical staff, are not at greater risk of developing sCJD.However, this cannot exclude the possibility that there may be an occupational risk in specific circumstances, for example, for people in contact with high-risk central nervous system tissue, and appropriate precautions, as recommended by national authorities, should therefore be followed, particularly regarding laboratory work.
Although in some studies occupation was specifically analysed [19,25] and occupation may be the subject of specific inquiry in some surveillance systems, a limitation of some registries and scientific studies is that occupation may not have been systematically recorded.
When occupation was recorded, it is unlikely that a framework for consistent occupational data collection was used, so that neither registries nor case-control studies have incorporated the classic epidemiological double approach.Recording of occupation may not identify specific chemical or biological exposures, which would require data for professions (job titles, medical specialisations) being cross-referenced with branches of activity (laboratory, administrative or clinical patient-contact work).The lack of registered surveillance data that combine profession with activity (e.g.contact with human tissue), when compared with the descriptions from previous case reports and the incident in Spain, illustrates the limits of the validity of available data for analytical purposes and precludes formal use of statistical testing.Although our study does not provide evidence of an excess risk of sCJD in health professionals, the fact that the data collected were mainly linked to medical speciality rather than actual activity might have concealed an excess risk of sCJD for some specific health professionals.
A case-control study seeking to examine the putative occupational risk posed by surgical injuries should have a biologically clear working hypothesis and a custom-tailored methodology.Matrices designed by linking medical speciality and surgical/forensicanatomical/pathological activity, in which the health   2), the study size that would afford the necessary statistical power for a proper examination of the specific practices of health professions is higher than that provided by existing CJD registries in any one country.
Since complementary analyses would be needed for professional and activity categories defined in terms of temporal references that have not been explored to date, such as 'ever employed' or 'currently employed', as well as duration of employment, requirements for study size and collaboration would be even higher.
In conclusion, a wide spectrum of medical specialities and health professions are represented in sCJD registries.Although selection due to higher ascertainment may lie behind the case reports of certain professions involved in clinical management or care of patients with sCJD, the biological significance of these observations remains uncertain and available data do not indicate an increased risk of sCJD in health professionals.However, the methodological issues mentioned above indicate the need for caution in drawing conclusions from the data and large-scale studies with specific causal hypotheses are needed in order for further research to be undertaken into the potential link between health professions and sCJD .Acknowledgments a The information on health-related occupations from this study was not included in the meta-analysis [6].
studies with no (n=33) or insuffi ciently described (n=1) individual health occupational data Reports assessing occupational health risk excluded due to non-validated diagnosis or devoid of sCJD cases among health prof essionals Reports issued bef ore 1989 included f rom personal records Reviewed case and case-series reports on sCJD among health professionals sCJD: sporadic Creutzfeldt-Jakob disease.

Figure 1
Figure 1Literature review of case reports of sporadic Creutzfeldt-Jakob disease among health professionals, 1979-1 October 2011

Table 2
Occupational profile of non-cases or controls obtained through the European Creutzfeldt Jakob Disease Surveillance Network (EuroCJD), 1980-2009 (n=2,968)

Table 3
[17,24]ibed in case-control studies and, in general, fulfilling criteria as probable or confirmed sporadic Creutzfeldt-Jakob disease (sCJD)[17,24].b Mention of clinical features, genetic study or country of origin is frequently made in Slovakian cases given the high incidence of genetic forms and the relevance of genotyping for correct classification of CJD.

Table 4
Summary of methods and main results of analytical epidemiological research into healthcare-related occupations and sporadic Creutzfeldt-Jakob disease, literature review, 1982-1 October 2011 CJD: Creutzfeldt-Jakob disease; EuroCJD: European Creutzfeldt Jakob Disease Surveillance Network; GP: general practitioner; OR: odds ratio UK: United Kingdom; USA: United States of America.