Partner HIV serostatus disclosure and determinants of serodiscordance among prevention of mother to child transmission clients in Nigeria

Background: Serodiscordance exists when the known HIV result of one member of a couple pair is positive while that of his/her partner is negative. In sub-Saharan Africa, in stable long-term couple partnerships (married or cohabiting), serodiscordance is a growing source of HIV-transmissions. This study aimed to ascertain across Nigeria, serodiscordance prevalence, partner HIV status disclosure and explore associations between suspected determinants and serodiscordance among PMTCT enrolled HIV positive pregnant women and their partners. Methods: A retrospective Quality of Care performance evaluation was conducted in July 2013 among 544 HIV positive pregnant enrolees of PMTCT services in 62 comprehensive facilities across 5 of Nigeria’s 6 geo-political zones. Data of client-partner pairs were abstracted from pre-existing medical records and analysed using chi-square statistics and logistic regression. Results: A total of 544 (22 %) of 2499 clients with complete partner details were analysed. Clients’ age ranged from 15 to 50 years with a mean of 30 years. Serodiscordant prevalence was 52 % and chi-square test suggests no significant difference between serodiscordant and seroconcordant clients and their partners (p = 0.265). Serodiscordant rates were closely associated trend wise with national HIV sero-prevalence rates and the median CD4+ count was 425 ul/mm3 (IQR: 290-606 ul/mm3). Similar proportion of clients (99 %) received testing and agreed to disclose status to their partners. Yet, there was no association between clients agreement to disclose HIV status to their partners and these partners getting tested and receiving results (p = 0.919). Significantly, 87 % of clients in concordant HIV positive relationships appeared to be symptomatic (WHO clinical stage 3 or 4) compared to 13 % clients in HIV-discordant relationships (p < 0.003). Client's age and CD4+ count did not aptly predict serodiscordance (Wald = 0.011 and 0.436 respectively). However, the WHO clinical staging appeared to be a better predictor of serodiscordance and concordance than other variables (Wald = 3.167). Conclusions: The results suggest that clinical staging (WHO) could be a better predictor of client- partner pair discordant or concordant HIV serostatus. Early partner testing and notification can avert seroconversion, hence properly designed and mainstreamed interventions that target serodiscordant couples are essential. Nigeria is organized into six geo-political zones, which encompass neighbouring states (see Additional file 1). The study was conducted in 5 of the 6 geo-political zones excluding the North East zone due to security challenges. In July, 2013, locations providing comprehensive HIV/AIDS treatment services to pregnant women and their partners for at least one year were enrolled. Eventually, 62 ARV (known as comprehensive service delivery points) spanning primary, secondary and tertiary level facilities were eligible. All pregnant women aged ≥ 15 years enrolled in PMTCT program in selected facilities in the study locations. Based on National PMTCT guidelines, pregnancy in the HIV positive woman is an indication for ARVs irrespective of CD4, VL or clinical stage [21]. A retrospective performance evaluation that measures Quality of Care (QoC) provided to HIV positive pregnant women and their exposed infants, particularly in the children’s first 2 years of life, in a 6-month review period (January 1st – June 30th, 2013). The QoC evaluation exercise included four (4) different categories of women (inclusion criteria) which were: Unbooked women are those who appear for the first time in the facility when in labour. Sample frame was developed by applying the four inclusion criteria stated above to each audited patient’s medical folders. The sampling method varied from site to site, depending largely on the site’s information management system. For Electronic Medical Record enabled sites (which includes patient IDs, up-to-date clinical and other follow up information) all eligible sample of PMTCT patients was generated and audited. For paper-based medical record sites, a systematic random sampling was used to generate the required sample of PMTCT patients audited. This was determined for each of the different sampling methods applied for the two categories of sites mentioned above using the HIVQUAL [22] sampling methodology (see Additional file 2). These relied on the following 5 simple steps; Where ‘’X” is the total sample of eligible population (Sample Frame) and ‘’Y” the sample size generated from HIVQual sample size determination table. The QoC performance evaluation process involved use of care and treatment service delivery indicators and entries in national data collection tools (DCTs) to guide data abstraction from randomly selected pre-existing patients’ medical records. The four inclusion criteria stated in section 2.4 were applied to each audited medical folders. For the PMTCT program component of the exercise, HIV negative pregnant women and HIV positive women with incomplete partner data were excluded from the sample of folders audited. Reliable and site specific data from the following registers - the Antenatal Care (ANC), HIV counselling and Testing, Delivery, Maternal, Partner, Early Infant Diagnosis, Child follow-up and ARV were reviewed and used to determine the pool of eligible sample of patients audited. The inclusion and exclusion criteria were applied on each step of the performance evaluation audit. Data was sorted, coded, keyed into Microsoft Access database and exported to SPSS 20.0 for analysis. Results were presented using charts, graphs and frequency tables. Descriptive statistics such as mean, frequencies and percentages was used to describe and summarize findings. Contingency tables were developed, logistic regression and chi-square statistics tested associations between variables and level of significance (α = 0.05). Ethical approval was obtained in line with the standard ethical procedures from the United States Centers for Disease Control and Prevention (CDC). The National Research Health Ethics Committee (NRHEC) of the Federal Ministry of Health Nigeria approved the primary survey. No clients were recruited for the PMTCT performance evaluation. Study abstraction forms collated data from pre-existing client medical records and facility registers. Within facilities, abstraction was performed only in secure areas and data with or containing links to unique identifiers of study subject were kept under lock and key, until these records were appropriately returned to the facility records sections. All individuals involved in data abstraction and analysis received training on confidentiality procedures and obtained Collaborative Institutional Training Initiative (CITI) human subjects certification.