|
An economic evaluation of universal vaccination against hepatitis B virus |
Fenn P, Gray A, McGuire A |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology Vaccination programmes against Hepatitis B using a genetically engineered yeast-derived vaccine, Engerix B.
Economic study type Cost-effectiveness analysis.
Study population Hypothetical infants, children (age 6 years) and adolescents (age 11 years) based on the UK Public Health Laboratory Service (PLHS) and a study by the US Center for Disease Control (CDC).
Setting Community. The study was carried out at Universities in Nottingham, Oxford and London in the UK.
Dates to which data relate It was assumed that future birth cohorts and age specific non Hepatitis B Virus (HBV) related deaths remained constant at the 1990 level. Epidemiological data were derived from the 1987 US study and age specific incidence rates from 1990 PHLS data on reported infections. All costs have been converted to 1992/1993 prices using the Hospital and Community Health Services Pay and Price Index as supplied by the UK Department of Health.
Source of effectiveness data Effectiveness data were derived from a review of the literature and estimates.
Modelling A Markov model was used to simulate the cost-effectiveness of vaccination on the transition to infection, subsequent disease states, and death. The Markov cycle was 1-25 years.
Outcomes assessed in the review The review assessed data on acute HBV infections, the age-specific all-cause death rate, and life expectancies with and without vaccination for each cohort.
Study designs and other criteria for inclusion in the review Not used. Epidemiological studies based on UK population surveys were sought, but the research found was based on a US population.
Sources searched to identify primary studies Criteria used to ensure the validity of primary studies Methods used to judge relevance and validity, and for extracting data Number of primary studies included Two reports with primary data were found and included in the review.
Methods of combining primary studies Investigation of differences between primary studies Results of the review The US study reported that only 8.7% of actual infections are reported. UK data recorded an incidence of approximately 1.1 cases per 100,000 population. Assuming that these cases represent 8.7% of actual infections, then the actual annual rate in the UK would be 12.6 per 100,000 population (approximately 7,000 infections per annum). These data were used as baseline estimates in the model. The estimated age-specific infection rates based on PHLS data (adjusted for underreporting) were the basis for the time dependent transition probabilities to infection in the Markov model.
Methods used to derive estimates of effectiveness Assumptions about effectiveness were also made by the authors.
Estimates of effectiveness and key assumptions To determine the extent to which the reduced incidence of infections translates into a reduced prevalence of chronic carriers, and consequently a further reduction in the likelihood of infection, it was assumed that the likelihood of infection was a linear function of the prevalence of chronic carriers. Other assumptions were: future birth cohorts and age-specific non-HBV related death rates remain constant at the 1990 level; a maximum period of 25 years for the evaluation of universal HBV vaccination; that the vaccine was administered according to either a 0, 1, 2 month primary schedule, or a 0, 1, 6 month primary schedule; full compliance, achieving a protective effect in 90% of cases; a necessary booster after 5 years; non-responders will revert to the infection rates estimates for the unvaccinated population; 25% of acute HBV infections developed acute type B hepatitis, of whom 1% die from fulminant hepatitis, implying an excess transition probability from acute infection to death of 0.0025; 10% of those infected over the age of 5, and 50% of those infected under the age of 5, become chronic carriers in the subsequent period.
Measure of benefits used in the economic analysis The benefit measures used were the avoidance of acute HBV infections and the increase in life expectancy following from the reduced number of infections and chronic sequelae. Life expectancy with and without vaccination was estimated for each cohort by summing the expected time a given individual spent in non-terminal states. For each alternative vaccination programme modelled, gains in life expectancy were summed over all vaccinated cohorts, after the combined effects of vaccine protection and herd immunity were simulated within a cyclical Markov process. Sensitivity analyses were carried out with respect to the assumed degree of underreporting and the assumed percentage of those infected becoming chronic carriers in the following period. The outcome was defined as incremental life years gained. Results were reported for both discounted (6%) and undiscounted benefits.
Direct costs The costs associated with each vaccination were estimated in relation to the number of cohorts vaccinated. It was assumed that each primary vaccination dose and booster cost 8.66, and that for all vaccinations apart from the primary infant schedule there would be an additional cost of administration per dose of 10.63, the cost of a GP consultation. The costs of in-patient care were calculated by taking the average cost for an in-patient day in an acute non-teaching hospital multiplied by the average length of stay for the disease group of interest (hepatitis, malignant neoplasm of liver and intrahepatic bile ducts, chronic liver disease and cirrhosis), and then multiplying by the average annual number of cases. To estimate the net costs associated with each programme, an estimate of the treatment costs which would be averted as a result of the reduced incidence and prevalence of Hepatitis B was subtracted. Unit costs per patient for treating acute Hepatitis B were estimated to be 1,520, for treating primary liver cancer, 2,262 and for treating cirrhosis, 1,862. By means of a sensitivity analysis 45 days of terminal care for patients with hepatic cancer were also allowed. Costs were discounted at an annual discount rate of 6%. Costs and quantities were not reported separately.
Statistical analysis of costs Sensitivity analysis Sensitivity analyses were carried out with respect to the assumed degree of underreporting of the incidence of actual infections. The assumption that 10% of those infected over the age of 5, and 50% of those infected under the age of 5, become chronic carriers in the subsequent period was subjected to sensitivity analysis. A sensitivity analysis was performed to allow for the costs of 45 days of terminal care for patients with hepatic cancer and to allow for the use of nursing time rather than GP time.
Estimated benefits used in the economic analysis Benefits were not reported separately.
Cost results Costs were not reported separately.
Synthesis of costs and benefits Baseline results with benefits discounted ranged from 188,015 to 301,365 per life year gained (LYG), depending on the duration of programme and vaccination strategy. Under these assumptions there is very little difference in cost-effectiveness between the adolescent and combined (infant and adolescent) strategies (ranges from 194,292 per LYG to 233,370 per LYG and 192,016 per LYG to 231,115 per LYG respectively), but the child vaccination strategy is relatively less cost effective (250,302 per LYG to 301,365 per LYG) and the infant strategy is relatively more cost-effective (188,015 per LYG to 227,130 per LYG). With benefits undiscounted, the comparable range is from 5,234 per LYG to 13,034 per LYG. Again, the infant programme offers the best cost-effectiveness ratios (ranging from 5,234 per LYG to 7,002 per LYG), and the child programme offers the worst (ranging from 9,646 per LYG to 13,034 per LYG). Sensitivity analysis revealed that the results were particularly sensitive to the assumed incidence and reporting rate of infections; results were also sensitive to the epidemiological uncertainties over the transition probabilities in the chronic stage of the disease. Results were less sensitive to the assumptions concerning hospital costs and vaccine administration costs. However, the effects of changes in administration costs were pronounced for the non-infant programmes.
Authors' conclusions A strategy of vaccinating infants was found preferable to other strategies, irrespective of the assumptions over discounting, due to the administrative savings from incorporating the vaccine into current procedures.
CRD COMMENTARY - Selection of comparators A justification was given for the comparators used, namely the authors sought to model the effects of universal vaccination over time on the specified cohorts - the "herd immunity" effect. However, selective vaccination for high risk groups as comparators might have been useful.
Validity of estimate of measure of benefit The authors did not state that a systematic review of the literature was undertaken. Effectiveness estimates were combined using narrative methods. The estimate of effectiveness was derived credibly from the primary studies. Estimation of benefits was modelled, using a Markov model, which is more appropriate than decision tree analysis.
Validity of estimate of costs The cost perspective was not specified and indirect costs were not included. Costs and quantities were not reported separately. A sensitivity analysis of quantities was conducted and the ranges used appear to be appropriate. A sensitivity analysis of prices was conducted and the authors performed sensitivity analyses to allow for 45 days of terminal care and nursing time rather than GP time. The authors reported results for both discounted and undiscounted benefits.
Other issues The authors did make comparisons of their findings with those from other studies, although the main focus of the report was a comparison of the 4 comparators included in this report. The issue of generalisability to other settings was addressed. The study focussed on 4 populations and this was reflected in the authors' conclusions.
Implications of the study A strategy of vaccinating infants seems preferable to a child vaccination programme, an adolescent vaccination programme, and a combined child and adolescent vaccination programme.
Source of funding Financial assistance from SmithKline Beecham UK.
Bibliographic details Fenn P, Gray A, McGuire A. An economic evaluation of universal vaccination against hepatitis B virus. Journal of Infection 1996; 32(3): 197-204 Other publications of related interest US Center for Disease Control. Hepatitis B surveillance report no. 51. US Center for Disease Control,1987.
US Center for Disease Control. Up-date on hepatitis prevention. US Center for Disease Control,1987.
Indexing Status Subject indexing assigned by NLM MeSH Age Factors; Child; Cost-Benefit Analysis; Great Britain; Hepatitis B /economics /prevention & Hepatitis B Vaccines /administration & Humans; Infant; Life Expectancy; Markov Chains; Vaccination /economics; Vaccines, Synthetic /administration & control; dosage /economics; dosage /economics AccessionNumber 21997006147 Date bibliographic record published 31/10/2000 Date abstract record published 31/10/2000 |
|
|
|