HIV Testing among Antenatal Care–Attending Pregnant Women and Male Partners in Cambodia: Primary and Secondary Data Analyses Using Cambodia Demographic Health Surveys (2005, 2010, and 2014), Case-Control Study and In-Depth Interview
Introduction:Cambodia has set a goal to test at least 95% of all pregnant women for HIV. Thanks to Cambodia’s “Boosted Linked Response” strategy, the HIV testing rate among pregnant women attending antenatal care (ANC) has gradually increased but plateaued out at approximately 90% for the past several years. Moreover, since 2012, the strategy has failed to test more than one-fourth of male partners of ANC-attending pregnant women. Therefore, we examined factors associated with HIV testing during ANC among pregnant women and their male partners. In addition, we also explored men’s barriers to attending ANC with their pregnant partners, and their perspectives on three HIV testing alternative strategies (home-based HIV testing, free-of-charge pregnant woman–delivered HIV self-testing, and out-of-pocket community pharmacy–delivered HIV self-testing.
Methods: In Study 1, we pooled together three Cambodia Demographic Health Surveys (2005, 2010, and 2014) and adopted Anderson’s Behavioral Model of Health Services to guide our data analysis. The study population consisted of all Cambodian women aged 15–49 years with one or more live births in the three years preceding each survey who attended ANC for the most recent birth (weighted N=11,181). In Study 2, we conducted a Case-Control study. The study population consisted of men who attended ANC with their pregnant partners from September 2020 to December 2020. The outcome was defined as declining or accepting an HIV test as part of ANC. 132 cases and 264 controls were recruited from three government-run ANC health facilities in Phnom Penh (the National Maternal and Child Health Center, Chaktomuk Referral Hospital, and Posenchey Health Center). An Extended Theory of Planned Behavior (ETPB) was applied to guide our data analysis. In Study 3, a qualitative study was nested within the Case-Control study. The study population consisted of 30 men (10 HIV testing decliners, 10 acceptors, and 10 ANC non-attendees). All participants were recruited using consecutive sampling from ANC and postpartum departments of the National Maternal and Child Health Center.
Results: In Study 1, HIV testing rates as part of ANC increased significantly by year (15.5% in 2005, 46.2% in 2010, and 77.4% in 2014, p<0.001). Women who received adequate pre-test counseling had consistently greater odds of being tested for HIV than those who did not: aOR=17.3 [95% CI: 12.4–24.0] in 2005, aOR=7.2 [95% CI: 5.9–8.9] in 2010, aOR=8.5 [95% CI: 6.5–11.2] in 2014, and aOR=8.9 [95% CI: 7.7–10.3] in the pooled dataset. In Study 2, male partners with a low intention to test and low perceived behavioral control (PBC) had greater odds of declining an HIV test (low intention with aOR=3.2 [95% CI: 1.8–5.6] and low PBC with aOR=1.8 [95% CI: 1.1–2.8]). A low intention to test was predicted by an absence of perceived risk of HIV infection (aOR=2.0 [95% CI: 1.1–3.6]), unsupportive subjective norms (aOR=2.3 [95% CI: 1.4–3.9]), and an absence of partner communication about HIV testing (aOR=3.0 [95% CI: 1.5–5.7]). In Study 3, barriers to attending ANC visits included individual-level factors (being unable to take time off work, waiting outside the health facility watching over an older child, not wanting to pay for a parking fee, and a negative attitude toward male ANC attendance), relationship-level factors (a negative couple dynamic, meaning men taking healthcare decisions alone as they considered themselves heads of the family), community-level factors (a negative attitude toward male ANC attendance), and health system-level factors (not being invited by providers to come inside the ANC consultation room). Concerns about home-based HIV testing included worries that neighbors would be suspicious, not being home at the time of the home visit, an inability to verify the identity of the testing team, the feeling of being coerced into testing, feeling offended by a home visit, the potential for contamination of testing instruments, the potential for confidentiality breaches by the testing team, and the potential for improper management of biological specimens. For pregnant woman–delivered HIVST, perceived concerns included a lack of pre-and post-test counseling, questionable accuracy of the test kit, instructions of use in foreign languages, not being capable of using the kit correctly, and questions about why the provider or his partner wanted him to be tested. For community pharmacy–delivered HIVST, perceived concerns included low availability, feeling shy or embarrassed when purchasing the kits, cost, and potential lack of technical assistance provided by pharmacy staff.
Conclusion: HIV testing uptake among ANC-attending pregnant women in Cambodia can be increased by delivering high-quality pre-test counseling. Therefore, health system-level interventions should include regular refreshment training to counselors at ANC sites without understaffing issues and lower the ratio of ANC clients to counselors at ANC sites plagued by understaffing issues. Among male partners, trans-theoretical approaches could be adopted to increase their intention to test by altering their risk perception, subjective norms, and partner communication. Potential community-level interventions using mass media, peer educators, and small-group education sessions may be useful in changing men’s behavior. Home-based HIV testing and HIV self-testing have the potential to become complementary services to the current ANC-based HTC.