Many countries use the cost–effectiveness thresholds recommended by the World Health Organization’s Choosing Interventions that are Cost–Effective project (WHO-CHOICE) when evaluating health interventions. This project sets the threshold for cost–effectiveness as the cost of the intervention per disability-adjusted life-year (DALY) averted less than three times the country’s annual gross domestic product (GDP) per capita. Highly cost–effective interventions are defined as meeting a threshold per DALY averted of once the annual GDP per capita. We argue that reliance on these thresholds reduces the value of cost–effectiveness analyses and makes such analyses too blunt to be useful for most decision-making in the field of public health. Use of these thresholds has little theoretical justification, skirts the difficult but necessary ranking of the relative values of locally-applicable interventions and omits any consideration of what is truly affordable. The WHO-CHOICE thresholds set such a low bar for cost–effectiveness that very few interventions with evidence of efficacy can be ruled out. The thresholds have little value in assessing the trade-offs that decision-makers must confront. We present alternative approaches for applying cost–effectiveness criteria to choices in the allocation of health-care resources.

Objective HIV-positive pregnant women are at heightened risk of becoming lost to follow-up (LTFU) from HIV care. We examined LTFU before and after delivery among pregnant women newly-diagnosed with HIV. Methods Observational cohort study of all pregnant women ≥18 years (N=300) testing HIV-positive for the first time at their first ANC visit between January–June 2010, at a primary healthcare clinic in Johannesburg, South Africa. Women (n=27) whose delivery date could not be determined were excluded. Results Median (IQR) gestation at HIV testing was 26 weeks (21–30). 98.0% received AZT prophylaxis, usually started at the first ANC visit. Of 139 (51.3%) patients who were ART-eligible, 66.9% (95%CI 58.8–74.3%) initiated ART prior to delivery; median (IQR) ART duration pre-delivery was 9.5 weeks (5.1–14.2). Among ART-eligible patients, 40.5% (32.3–49.0%) were cumulatively retained through six months on ART. Of those ART-ineligible at HIV testing, only 22.6% (95%CI 15.9–30.6%) completed CD4 staging and returned for a repeat CD4 test after delivery. LTFU (≥1 month late for last scheduled visit) before delivery was 20.5% (95%CI 16.0–25.6%) and, among those still in care, 47.9% (95%CI 41.2–54.6%) within six months after delivery. Overall, 57.5% (95%CI 51.6–63.3%) were lost between HIV testing and six months post-delivery. Conclusions Our findings highlight the challenge of continuity of care among HIV-positive pregnant women attending antenatal services, particularly those ineligible for ART.