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.

Universal and selective antenatal HIV screening in pregnant women. The universal policy involved offering HIV screening to all pregnant women while the selective strategy involved offering screening only to women at high risk of infection.

Screening.

Cost-benefit analysis and cost-effectiveness analysis.

Pregnant women in the UK.

Secondary care. The economic study was carried out in the UK.

Effectiveness data were obtained from literature published between 1997 and 1999. Resource use data and their corresponding dates were not reported. Cost data were obtained from studies published between 1997 and 1999. The price year was 1996-7.

Effectiveness data were derived from a review of the literature.

A decision model was constructed to estimate the costs and effects associated with each strategy including the threshold value for the prevalence of HIV above which the cost-effectiveness of universal screening as opposed to selective screening was assured.

Probabilities and other factors influencing the effectiveness of the screening programme before and after maternal diagnosis were as follows:

false negative rate (initial screen) and false positive rate of test algorithm,

uptake of testing in higher risk groups with an "on request" policy and selective testing,

uptake of testing in low risk groups with an "on request" policy,

uptake of testing in low risk groups with universal policy,

average percentage of antenatal population in higher risk groups (London and outside London),

prevalence of undiagnosed HIV in higher risk groups (London and outside London),

prevalence of undiagnosed HIV in low risk groups (London and outside London),

average prevalence (range) of undiagnosed HIV in UK districts (London and outside London),

probability of termination if HIV detected during pregnancy,

rate of vertical transmission with and without risk reduction strategies,

life expectancy of an HIV infected child when the mother's infection was not known during pregnancy (years),

efficacy of antiretroviral therapy,

life expectancy of an HIV infected child when the mother's infection was identified during pregnancy (years),

life expectancy of an uninfected child,

life expectancy of HIV infected woman whose infection was diagnosed and not diagnosed during pregnancy (years).

Not reported.

Not reported.

Not reported.

Not reported.

A total of 8 studies or reports were included.

Not reported.

Not reported.

The values adopted for the probabilities and other factors affecting the effectiveness of the screening programme before and after maternal diagnosis were as follows:

false negative rate (initial screen) and false positive rate of test algorithm, 0;

uptake of testing in higher risk groups with "on request" policy and selective testing, 0.25 and 0.20, respectively;

uptake of testing in low risk groups with "on request" policy, 0.05;

uptake of testing in low risk groups with universal policy, 0.95;

average percentage of antenatal population in higher risk groups in London and outside London, 14.95 and 5.55, respectively;

prevalence of undiagnosed HIV in higher risk groups (London and outside London), 0.98 and 0.18, respectively;

prevalence of undiagnosed HIV in low risk groups (London and outside London), 0.014 and 0.002, respectively;

average prevalence of undiagnosed HIV in UK districts (London and outside London), 0.16 (range: 0.021 - 0.28) and 0.013 (range: 0.0021 - 0.038), respectively;

probability of termination if HIV detected during pregnancy, 20% (range: 0 - 25%);

rate of vertical transmission with and without risk reduction strategies, 7.6% (range: 1.7% - 12.7%) and 26.5% (range: 21.1% - 31.9%);

efficacy of antiretroviral therapy, 0.67 (pessimistic and optimistic values, 0.8 and 0.6) for the infected child and 0.6 (0.8 and 0.4) for the infected woman;

life expectancy of HIV infected child when mother's infection not known during pregnancy, 10.09 years (8.60);

life expectancy of HIV infected child when mother's infection identified during pregnancy, 11.66 years (9.87);

life expectancy of an uninfected child, 77 years (39.51);

life expectancy of HIV infected woman whose infection diagnosed and not diagnosed during pregnancy, 18.25 years (13.96) and 17.63 years (13.5), respectively.

These data formed the input parameters for the model

The measure of benefits was discounted life expectancy (years) which was turned into monetary benefit by assuming a willingness to pay of 10,000 per life year gained.

Costs were discounted. Quantities of resource use were not reported separately from the costs. Cost components were reported separately. The cost analysis covered the costs of termination, initial counselling and management of HIV positive women during pregnancy, additional medical management of women with HIV during pregnancy, normal birth, risk reduction strategy (including delivery), diagnosis and lifetime costs for an HIV-infected child, management of women diagnosed and not diagnosed with HIV infection, cost of the screening test, and the average cost of the offer and pretest discussion. The perspective adopted in the cost analysis was that of the health care system. The source of cost data was research published between 1997 and 1999. 1996-7 price data were used. The costs incurred due to the additional time spent identifying women at higher risk were not included in the total cost of selective strategy.

Not considered.

UK pounds sterling ().

Sensitivity analyses were performed based on the most optimistic and the most pessimistic conditions for antenatal testing. Threshold analysis was performed to identify the prevalence thresholds above which universal policy would turn out to be cost-effective relative to the selective policy. The prevalence threshold values were then compared with the estimated values for the prevalence of undiagnosed HIV infection in pregnant women in different districts of the UK to find out how many districts should adopt the universal policy.

The estimated discounted (2% discount rate) life expectancy of an HIV-infected woman and her child when infection was diagnosed during pregnancy was 13.959 years and 37.254 years, respectively. The corresponding values in the case of the infection not being diagnosed were 13.504 years and 31.317 years. The total difference between the diagnosed and undiagnosed cases was 6.392 years, which was turned into monetary value by multiplying it by 10,000 (6.392 times 10,000).

The estimated discounted (6% discount rate) costs of an HIV-infected woman and her infant when infection was diagnosed during pregnancy were 104,344 and 6,086, respectively. The corresponding values in the case of undiagnosed infection were 79,961 and 15,636. The total difference in costs between diagnosed and undiagnosed cases was 14,833.

Net benefit (the value of life according to a willingness to pay of 10,000 times the number of life years gained minus the additional downstream costs) was used as the measure of cost-benefit. The net benefit of adopting an antenatal screening programme was 49,090 (range: 12,300 - 59,000). With a test cost of 0.60 and uptake of 95%, and up to 4 minutes pretest discussion, the universal screening could be adopted in almost all of the UK.

Universal screening with a pretest discussion should be adopted throughout the United Kingdom as part of routine antenatal care as long as test costs can be kept low and uptake high.

A justification was given for the choice of the comparator. You, as a database user, should consider whether this is a rational choice.

Although the authors adopted sound methodological techniques, the internal validity of the estimates of benefit can not be objectively assessed due to the lack of a comprehensive literature review and quality assessment of the primary studies included in the review. However, uncertainties in the data were comprehensively explored in the sensitivity analyses.

Quantities of resource use were not reported separately from the costs. Adequate details of the methods of cost estimation were given.

The degree to which the authors' conclusion are justified cannot be fully assessed due to the fact that the complete set of results obtained from the sensitivity analyses reported in other studies conducted by the authors was not presented. The issue of generalisability to other settings or countries was not fully addressed. The authors highlighted how the validity of economic evaluation in this area could be increased (see below).

Universal screening with pre-test discussion is the optimal choice in the UK as long as the test costs can be kept low and uptake is high. Future studies should investigate other relevant factors in economic evaluation such as anxiety caused by testing, the reduction of infection upon diagnosis, quality adjustment to life years, and societal costs of caring for orphaned children.

UK Department of Health.