Should British soldiers be vaccinated against hepatitis A? An economic analysis T.O. Jefferson *~, R.H. Behrens* and V. Demicheli* We conducted a stud), to analyse the eJficieno, of introducing vaccination against hepatitis A to the schedule .['or troops in the British Army. The study design included a cost-effectiveness analysis ( CEA ) of cost per case avoided comparing active and passive immunization and a cost-benefit analysis (CBA). The study population comprised all British Arm)' personnel as well as those soldiers assumed to be deployed to high-risk areas a variable number o/'times in 5 years. The average cost of one case of hepatitis A avoided b)' vaccination was calculated and compared with the average cost of achieving the same outcome by passive immunization. In a 5-year J'our-exposure scenario at a low incidence and using a 3% discount rate, avoiding one case of hepatitis A by vaccination would cost £52865, aga#lst £97305 by passive immunization. The equivalent cost-benefit ratios are 13.4 ./'or gammaglobulin and 7.2 .['or vaccine. For fewer exposures the break-even point Jor vaccine is two exposures in 4 ),ears. Although our estimates are sensitive to direct costs and relative(v sensitive to the estimated incidence, vaccinating troops against hepatitis A appears to be a more £fficient procedure than passive immunization, especiall)' as a long-term invesmlent in troops likely to effect several operational deployments. Given the d(~culO' of forecasting which troops would deploy, the best-buy strategy may be vaccination off troops most likely to deploy repeatedly. Keywords: Hepatitis A; prevention; economic analysis: British Army The importance of infection with the hepatitis A virus (HAV) for military formations is well known. In the Pacific campaign US Army units experienced incidence rates as high as 2600 per 100000 average strength, while in Vietnam the incidence rate rose from 570 to 670 once passive immunization was discontinued 1. More recently, outbreaks of hepatitis A have occurred in Norwegian troops deployed on UN duties in Gaza and Lebanon. These troops, although partly passively immunized, had an incidence of hepatitis A as high as 898 per 100 000 in Gaza according to Hesla 2 or 460 per 100000 according to Tormans et al. 3, who calculated an aggregate incidence rate for both deployments. As a new vaccine against hepatitis A has recently been developed (Havrix, SmithKline Beecham), we decided to analyse the costs and benefits of adding immunization against hepatitis A to the schedule of British troops vaccinations as an aid to decison-making. At present the British Army is actively immunizing personnel and dependants travelling on duty to areas *Medical Directorate, Headquarters British Army of the Rhine, British Forces Post Office 140, London, UK. *Hospital for Tropical Diseases Travel Clinic, 180-182 Tottenham Court Road, London W1P 9LE, UK. ++Universityof Pavia, Via Bassi 21, 27100 Pavia, Italy. ~To whom correspondence should be addressed. (Received 16 December 1993; revised 3 March 1994; accepted 15 March 1994) with a high risk of hepatitis A. We conducted our economic analysis on Army personnel only, thus excluding dependants. METHODS Our economic analysis had two aims: to calculate the cost of avoiding one case by vaccination and to compare it with the cost of obtaining the same result by the injection of human immunoglobulin; and to calculate the total costs and benefits of vaccination and to evaluate the efficiency of introducing such a policy. The disease To achieve both aims it was necessary to estimate the incidence of potentially avoidable cases among British soldiers. We achieved this by analysing the incidence of hepatitis A in 1991, the year of the Gulf War. We then used this estimate (21 cases per 100000 population at risk) as a low incidence estimate and tenfold greater incidence as a high estimate to simulate a higher risk caused by worse operational conditions. Using the low and high incidence estimates we forecast the expected number of cases of hepatitis A over 5 years, the timespan of our study. We held the risk constant in both simulations. 0264-410X/94/15/1379-435 :~" 1994 Butterworth-HeinemannLtd Vaccine 1994 Volume 12 Number 15 1379