Assessing the capacity and findings of routine programmatic data in Kenya to guide decision-making around contraceptives and antiretroviral therapy

The World Health Organization’s comprehensive strategic approach to prevention of mother-to-child transmission of HIV (PMTCT) covers the “4-prongs” of PMTCT programming: primary prevention of HIV; prevention of unintended pregnancies; prevention of vertical transmission from mother to infant; and provision of treatment, care, and support to mothers and infants [1, 2].

In resource-constrained settings, ensuring universal access to contraceptive options can be logistically challenging, and this is further complicated by the potentially reduced effectiveness of some long-term contraceptives in the presence of specific antiretrovirals. Clinician and patient contraceptive and antiretroviral therapy (ART) decisions are made within the larger context of standardized national guidelines and standardized essential medicines procurement at the national level, which, while simplifying procurement and improving cost-efficiency for the national health system [3], may not allow for a wide array of options available to individual patients.

This provision of ART and contraceptives becomes particularly problematic in situations where negative drug interactions are reported after the national introduction of new guidelines and commodities. Patel and colleagues [4] refer to the recent global “scare” about a potential association between dolutegravir use in pregnancy and neural tube defects [5] and how the subsequent rapid transition of some national programs to an efavirenz (EFV)-based ART regimen had the undesired knock-on effect of potentially reducing effectiveness of long-term contraceptive use in women who were on EFV, which they had reported in an earlier paper [6].

In this paper, Patel and colleagues comment on Kenya’s “robust national electronic medical record (EMR) system” [4] and follow-up their previous paper [6] by conducting a study with the dual intent of (1) conducting “a three-phase validation study” on the ability of strong national EMRs to provide accurate and reliable data to guide decision-making and (2) to better estimate the associations between long-term contraceptive methods, EFV-containing ART regimens, and incident pregnancy [4].

The three-phase validation included EMR review of 4 years of records from over 85,000 women living with HIV (WLHIV) (> 170,000 person-years of observation), 5000 random chart abstractions, and 1000 phone interviews with women who did and did not become pregnant using EFV and long-term contraceptive methods. Using these methods, Patel and colleagues planned to pool program data to determine if there is reduced contraceptive effectiveness in a real-world setting, while simultaneously confirming that the national program data used for analysis is reliable enough to guide future decision-making around contraceptive choices and ART regimens.