Background: Despite having high cervical cancer incidence and mortality rates, screening for cervical precancerous lesions remains infrequent in sub-Saharan Africa. The need to screen HIV-positive women because of the higher prevalence and faster progression of cervical precancerous lesions may be heightened by the increased access to highly-active antiretroviral therapy (HAART). Policymakers need quantitative data on the effect of HAART and screening to better allocate limited resources. Our aim was to quantify the potential effect of these interventions on cervical cancer mortality.
Methods and findings: We constructed a Markov state-transition model of a cohort of HIV-positive women in Cameroon. Published data on the prevalence, progression and regression of lesions as well as mortality rates from HIV, cervical cancer and other causes were incorporated into the model. We examined the potential impact, on cumulative cervical cancer mortality, of four possible scenarios: no HAART and no screening (NHNS), HAART and no screening (HNS), HAART and screening once on HAART initiation (HSHI), and HAART and screening once at age 35 (HS35). Our model projected that, compared to NHNS, lifetime cumulative cervical cancer mortality approximately doubled with HNS. It will require 262 women being screened at HAART initiation to prevent one cervical cancer death amongst women on HAART. The magnitudes of these effects were most sensitive to the rate of progression of precancerous lesions.
Conclusions: Screening, even when done once, has the potential of reducing cervical cancer mortality among HIV-positive women in Africa. The most feasible and cost-effective screening strategy needs to be determined in each setting.