|
The cost-effectiveness of adolescent hepatitis A vaccination in states with the highest disease rates |
Jacobs R J, Margolis H S, Coleman P J |
|
|
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 The vaccination of adolescents (15-year-olds) against hepatitis A.
Economic study type Cost-effectiveness analysis.
Study population The study population comprised the residents of the 10 USA states with the highest rates of reported hepatitis A between 1990 and 1997.
Setting The setting was primary care. The study was based in the states of Arizona, California, Idaho, Missouri, Nevada, New Mexico, Oklahoma, Oregon, Utah and Washington.
Dates to which data relate The effectiveness data were derived from sources published between 1984 and 1998. Also, from a database of hepatitis A incidence between 1990 and 1997, published by the Centres for Disease Control and Prevention National Notifiable Diseases Surveillance System. These data were modified by expert opinion to produce the final estimates of the parameters for the decision model. The direct cost data were derived from the resource use of hepatitis A treatment since January 1995 (published in 1997), combined with the expert panel estimates and unit costs from 1997. The indirect cost data were also derived from the resource use since January 1995, combined with the unit costs published between 1990 and 1996, and adjusted to 1997 levels. The price year was 1997.
Source of effectiveness data The effectiveness data were derived from a review of the literature, an expert panel and the authors' assumptions.
Modelling A decision model was used to estimate the relative consequences in terms of the health outcome and the cost of vaccination versus no vaccination. The decision model included a state transition model. The states were: death, infected and asymptomatic, infected and symptomatic, or not infected, given either vaccinated fully (two doses), partially (one dose) or not vaccinated.
Outcomes assessed in the review The model parameters were:
the annual hepatitis A infection incidence per 100,000;
the percentage of symptomatic infections;
the percentage of symptomatic patients hospitalised;
the number of liver transplantations among symptomatic patients;
the number of deaths among symptomatic patients; and
vaccination efficacy.
Study designs and other criteria for inclusion in the review No criteria were given. The study designs included a case series, and a database.
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 An expert panel survey was used.
Number of primary studies included Eighteen primary studies were included in the review.
Methods of combining primary studies An expert panel survey was used.
Investigation of differences between primary studies An expert panel survey was used.
Results of the review The results were presented according to age group. Those not presented to the expert panel are summarised below as the range between the two age groups, 15 to 19 years and 80 years and over.
The annual hepatitis A infection incidence (per 100,000) was 109 to 17.7.
The duration of symptoms was 38 days (95% confidence interval, CI: 32 - 44) to 43 days (95% CI: 34 - 52).
Methods used to derive estimates of effectiveness An expert panel survey was used. The panel comprised six hepatitis A experts, and three sequential surveys were conducted using the Delphi method. In the first survey, the experts were presented with the results of the literature review, from which they had to make estimates. In the second survey, the experts were presented with their peers' estimates, from which they had to reconsider their own estimates. These estimates were used to calculate the mean values and 95% CIs. "Consensus was considered to have been achieved when the 95% CI was +/- 15% of the mean value. The third survey was limited to items with larger CIs."
The authors also made some assumptions about the effectiveness.
Estimates of effectiveness and key assumptions The results were presented according to age group and are summarised below as the range (except for vaccination efficacy) between the two age groups, 15 to 19 years and 80 years and over.
The proportion of symptomatic infections ranged from 73% (95% CI: 67 - 78) to 90% (95% CI: 90 - 90). The proportion of symptomatic patients hospitalised ranged from 10% (95% CI: 9 - 11) to 43% (95% CI: 40 - 46).
The number of liver transplantations among symptomatic patients, per 10,000, ranged from 2 (95% CI: 1 - 3) to 0% (95% CI: 0 - 0).
The number of deaths among symptomatic patients, per 10,000, ranged from 18 (95% CI: 6 - 30) to 385 (95% CI: 356 - 414).
Long-term vaccination efficacy (age 10 to 75 years) was 95 to 69% for the full (two-dose) vaccine, and 62 to 13% for the partial (single-dose) vaccine.
Two assumptions were made. First, the ratio of the actual number of symptomatic cases to the reported cases in each age group would be three to one. Second, only those patients younger than 60 years would be transplant candidates.
Measure of benefits used in the economic analysis The measure of benefits was the years of life saved (YOLS). This was derived using modelling.
Direct costs The direct costs were broken down into vaccination and treatment according to diagnosis-related group per hospital stay, physician visits and outpatient procedures, and drugs.
The resource quantities for vaccination were derived from the assumption that 80% of adolescents would receive at least one dose, 80% of those vaccinated would receive the second dose and, under the Vaccines for Children Programme, 70% of adolescents would be eligible for vaccination. The resource quantities for physician visits and outpatient procedures came from the case series. The resource quantities for hospital days, physician visits, tests and treatment drugs came from the literature review and subsequent expert panel survey.
The unit costs for vaccination were $11.08 (range: 3 - 15.19) under the Vaccines for Children Programme, and $23.08 per dose and $13.7 to $17.56 for administration otherwise. The unit costs for treatment were for hospitalisation, physician visits, outpatient procedures, and drugs. The cost of hospital stay was derived using 1997 reimbursements for diagnosis-related groups 205 and 206 (diseases of the liver, except malignancy, cirrhosis, and alcoholic hepatitis). The cost of physician visits and outpatient procedures were derived using Medicare fees. The drug costs were estimated by reducing average wholesale prices by 20%, assuming that the lowest cost generic drugs were used. The resource quantities were given by number of hospitalisations and liver transplantations only. The discount rate was 3% in the base-case. The price year was 1997.
Statistical analysis of costs Hepatitis A treatment and work loss costs were presented as a mean value with the 95% CI. The total costs were given for the 10 USA states, pooled, as point estimates only. No statistical tests of the difference were reported.
Indirect Costs The indirect costs were days missed from paid employment and lost housekeeping services. For employment, the quantities were obtained from the case series, and the unit costs were obtained from age- and gender-specific estimates of workforce participation and earnings. It was assumed that 55% of infections occur in males, and annual earnings were for 250 workdays. For housekeeping, the quantities were obtained by assuming a 50% reduction in housework while hepatitis A symptoms were present. The unit costs were estimated by applying age-, gender- and employment status-specific values. All unit costs were adjusted to 1997 levels using the Employment Cost Index. The discount rate was 3% in the base-case.
Sensitivity analysis One-way sensitivity analyses were conducted on the pooled 10-state population. These were conducted by:
varying the discount rate between 0 and 5%;
varying the vaccine efficacy between its 95% CIs;
using the lower 95% CI for efficacy for 20 years and 0% efficacy thereafter;
substituting mortality estimates from the Viral Hepatitis Surveillance Programme for the expert panel estimates;
varying hospitalisations and liver transplantation rates between the 95% CIs;
varying work loss and symptom duration between the 95% CIs; and
substituting infection rates for the total USA for the 10 states with the highest infection rates.
Also, the number of quality-adjusted life-years was estimated by assuming "a 50% reduction in utility for 1 month per symptomatic infection".
Estimated benefits used in the economic analysis The number of years of life lost was 1,486 without vaccination and267 with vaccination.
The number of years of life saved by vaccination was 1,219.
The discount rate was 3%.
Cost results The total health system costs were $17,333,000 without vaccination and $34,062,000 with vaccination.
The total societal costs (including indirect costs) were $46,781,000 without vaccination and $39,730,000 with vaccination.
The discount rate was 3%.
Synthesis of costs and benefits An incremental cost-effectiveness analysis was carried out to combine the costs and the benefits. From the societal perspective, vaccination was the dominant strategy, having decreased the number of years of life lost and the costs.
From the perspective of the health system, the incremental cost per discounted YOLS was $13,722.
The sensitivity analyses showed that, from the societal perspective, vaccination was dominant except for the following scenarios:
when the discount rate was 5%;
when the lower 95% CIs were used to estimate efficacy for 20 years; or
when the lower United States hepatitis A incident rates were used, which was the least cost-effective scenario at $35,619 per discounted year of life saved.
The sensitivity analyses showed that, from the health system perspective, the most cost-effective scenario was with a discount rate of 0%, at $2,997 per discounted year of life saved. The least cost-effective scenario was using lower United States incidence rates, at $54,191 per discounted year of life saved.
The number of quality-adjusted life-years saved was estimated to be 2,872.
Authors' conclusions "Hepatitis A vaccination of adolescents in states with high disease rates would reduce costs to society. Although health system costs would increase, cost-effectiveness is comparable to other recommended vaccines and superior to many commonly used medical interventions."
CRD COMMENTARY - Selection of comparators The use of no vaccination as the comparator was justified by it being current practice. You should consider if this is appropriate to your own setting.
Validity of estimate of measure of effectiveness The effectiveness was measured in terms of a summary measure, the YOLS. This was estimated using a decision model with input parameters derived either directly from the literature, via an expert panel assessing the literature estimates, or from the authors' assumptions. This process was generally reported comprehensively, the authors building on the flaws of previous modelling work (for example, stratifying parameter values by age group). However, selection bias might have occurred when selecting the sources, since no details were provided of the review of the literature or of how the variables to present to the expert panel were chosen. Also, it was unclear how the 95% CIs were derived from the expert panel, and ending the iterative process when the CI was +15% of the mean value appears arbitrary. However, the authors acknowledged the limitations of their method of estimation and made some allowance by using the least favourable values.
Validity of estimate of measure of benefit The YOLS is an appropriate measure of benefit to compare with technologies where quality of life is not an important issue. The authors estimated the effect of using a preference-based method of valuing quality of life attributes. They showed that the benefit would only increase, although they did not reduce the value of extra infection-free years, for example due to ageing.
Validity of estimate of costs The method used to calculate the costs was reported comprehensively. In terms of generalisability, sensitivity analyses were conducted on appropriate variables to reflect important possible variations between settings, for example different economic perspectives. It is difficult to know to what extent sensitivity analyses should have been conducted, since the impact of variability or uncertainty depends on the precise setting and the timing of a decision on implementation. Ideally the authors' model should be available to others for testing, but more details could have been given on the actual resource quantities used.
Other issues Appropriate comparisons with other studies were made. Generalisability was discussed, including the use of sensitivity analyses. The authors also acknowledged several limitations of their study. For example, the method of effectiveness estimation, lack of account of herd immunity, migration, variability in the geographical focus of vaccination, and the lack of longer-term studies. The resource quantity results could have been reported more comprehensively. The authors' conclusions were in keeping with the population studied.
Implications of the study "Although health system costs would increase, cost-effectiveness is comparable to other recommended vaccines and superior to many commonly used medical interventions." While the economic perspective of the study was that of the United States, the study seems to have been presented with sufficient transparency to consider the implications for other settings.
Source of funding Part funded by an unrestricted research grant from SmithKline Beecham Pharmaceuticals (Philadelphia, PA, USA).
Bibliographic details Jacobs R J, Margolis H S, Coleman P J. The cost-effectiveness of adolescent hepatitis A vaccination in states with the highest disease rates. Archives of Pediatrics and Adolescent Medicine 2000; 154(8): 763-770 Indexing Status Subject indexing assigned by NLM MeSH Adolescent; Adult; Cost-Benefit Analysis; Decision Support Techniques; Female; Hepatitis A /economics /prevention & Humans; Male; United States; Vaccination /economics; control AccessionNumber 22000001344 Date bibliographic record published 30/11/2002 Date abstract record published 30/11/2002 |
|
|
|