Vaccine 26 (2008) 4290–4297 Contents lists available at ScienceDirect Vaccine journal homepage: www.elsevier.com/locate/vaccine Comparison of self-reported and recorded vaccinations and health effects in Australian Gulf War veterans Helen Kelsall a,b,∗ , Dean McKenzie a , Malcolm Sim a , Karin Leder a , James Ross c,d , Andrew Forbes a , Jillian Ikin a a Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria, Australia b Cancer Epidemiology Centre, The Cancer Council Victoria, Melbourne, Australia c Defence Health Services Division, Australian Defence Force, Canberra, ACT, Australia d International SOS Australasia, Sydney, New South Wales, Australia article info Article history: Received 11 March 2008 Received in revised form 5 May 2008 Accepted 12 May 2008 Available online 2 June 2008 Keywords: Gulf War Vaccination Veterans’ health abstract Background: Vaccinations, and multiple vaccinations in particular, have been associated with adverse health effects in veterans of the 1990/1991 Gulf War. However, exposure assessment has usually been based on self-report and recall bias may have influenced the results. Methods: We investigated agreement between self-reported and recorded vaccinations and the relation- ship with health status in Australian Gulf War veterans. Results: Agreement between self-reported and recorded vaccinations was highest for plague (kappa = 0.80), and kappa coefficients were greater than 0.60 for polio and ‘other unlisted’ vaccines, between 0.41 and 0.60 for hepatitis B, hepatitis A, typhoid and pertussis, and less than 0.40 for the other listed vaccines. The associations of increasing number of self-reported vaccinations in dose response rela- tionships with total number of symptoms, functional impairment, and poorer physical health were not observed when based on recorded vaccination data, although the actual difference in estimates was small and statistically significant only for total number of symptoms. Conclusions: Vaccinations were not associated with adverse health effects when exposure assessment was based on recorded vaccinations. It would be prudent that future research studies should be based on recorded vaccination data. © 2008 Elsevier Ltd. All rights reserved. 1. Introduction Since 1992 there has been a sustained effort internationally to investigate the health complaints that have emerged in veterans of the 1991 Gulf War. A common finding of cross-sectional studies of UK, US, Australian and Canadian veterans [1–6] is that ill-health such as self-reported symptoms and multisymptom illnesses, func- tional impairment, poorer psychological health, and perception of poor health are more common among Gulf War veterans than their military comparison groups. Vaccination in defence forces plays an important role in pro- tecting the health and maintaining the operational effectiveness of service personnel. However, receipt of multiple vaccines and cer- tain specific vaccines, such as those against plague and anthrax, ∗ Corresponding author at: Department of Epidemiology and Preventive Medicine, Monash University, Alfred Hospital, Commercial Road, Melbourne, Vic- toria 3004, Australia. Tel.: +61 3 9903 0032; fax: +61 3 9903 0556. E-mail address: helen.kelsall@med.monash.edu.au (H. Kelsall). have been associated with multisymptom illness and other health effects among veterans of the 1990/1991 Gulf War from several countries [1,4,7,8], including our previously published finding of an association between increasing number of reported vaccinations and the total number of reported symptoms in a dose response relationship in Australian Gulf War veterans [1]. A study of UK Gulf War veterans who reported that they had their vaccination records to refer to showed that multiple vaccinations during, rather than before, deployment were associated with adverse health effects [9], although the overall effect of vaccines before and during deploy- ment was not statistically significantly different [10]. One potential limitation of the findings of these studies is that measurement of vaccination exposure was based on self-report, rather than objec- tive records. Several studies have attempted to assess the health impact of vaccination using self-reported vaccination status and to examine the factors that might influence the validity of self-reported data. Reported receipt of any routine vaccine or of plague vaccine was significantly associated with a multisymptom illness only in those UK Gulf War veterans who did not have their vaccination records to 0264-410X/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.vaccine.2008.05.032