Title:
COST-EFFECTIVENESS OF UNIVERSAL AND SELECTIVE HIV
SCREENING OF PREGNANT WOMEN IN QUÉBEC
Authors:
Remis, R.S., Vandal, A.C.,
STD/AIDS Prevention and Control
Program, Regional Public Health Department, Montreal, Quebec, Canada
ABSTRACT:
Objectives:
To determine the cost per HIV infection prevented among newborns through a universal and selective program in Québec to screen pregnant women for HIV infection and offer zidovudine to prevent maternal-infant transmission.
Methodology:
We stratified pregnant women into three groups: women born in an endemic country (END), injection drug users (IDU's) and others (primarily women born in Canada and infected by heterosexual transmission). For each group, we estimated the proportion with CD4+ counts less than 700/microliter and the proportion of HIV-infected women aware of their serostatus. The efficiency of HIV transmission from mother to infant was varied by CD4+ count (indicating the stage of maternal infection) based on the European Collaborative Study and modified by zidovudine administration according to the results of the ACTG 076 clinical trial. The model was refined by factoring in HIV testing and zidovudine acceptance rates, rates of therapeutic abortion among newly diagnosed HIV-infected women and by assuming lower seroprevalence among test acceptors. A frequency distribution was assigned to each parameter using the best available data. Approximate distributions of the model outcomes, including cost-effectiveness, were determined using Monte-Carlo simulation. The model was evaluated for HIV screening of all pregnant women (universal screening) and of END and IDU women only (selective screening); in the latter case, the rates at which HIV testing would be offered and accepted were assumed to be lower than for universal screening.
Results:
Under universal and selective screening for IDU and END respectively, the expected number of prevented HIV infections in newborn infants were 3.1 (with 95% confidence interval [CI] 2.1; 4.4) and 2.3 (1.5; 3.3) and costs per transmission prevented were $242,000 ($172,000; $353,000) and $56,000 ($39,000; $81,000).
Discussion:
In spite of the uncertainties in the parameters, the model provided reasonably precise estimates for the cost-effectiveness of HIV screening among pregnant women. Although selective screening appears to provide much greater efficiency and would prevent most HIV infections in newborn infants, any policy decision must also take into account ethical, political and logistic considerations. The estimations derived by our study will be useful in the current deliberations concerning policies to provide HIV screening for pregnant women.
Address:
R.S. Remis, 1616 René-Lévesque ouest, Suite 300A, Montréal, Qc., Canada H3H 1P8
Tel: (514) 932-3305 Fax: (514) 932-1502 Email: md22@musica.mcgill.ca
A.C. Vandal, Department of Statistics, University of Auckland, Private Bag 92019, Auckland, New Zealand
Tel.: 9 373-7599/5713 Fax: 9 373-7018 Email: vandal@stat.auckland.ac.nz
Return to list of publications
Retour à la liste des publications