Predictors of adherence to clinical appointments among caregivers of infants on Hiv care in Kakamega county referral hospital, Kenya

Abstract

Poor attendance in scheduled clinical appointments in HIV care affects transmission of HIV from an infected mother to the infant (MTCT), clinical processes, and treatment outcomes. Whereas the increased morbidity and mortality associated with HIV care discontinuation is well established, few studies have reported appointment adherence among caregivers of infants in HIV care. Predictors of adherence to appointment in caregivers of infants on HIV care remain unexplored, while the MTCT rate in Kakamega is higher than surrounding Counties despite > 80% of the pregnant women receiving a HIV test. This study sought to determine predictors of clinical appointment adherence among caregivers of children ≤18 months on HIV care at Kakamega Referral Hospital. Specific objectives were; to assess socio-demographic factors, clinical factors and develop a prediction model for clinical appointment adherence among caregivers of infants in HIV care. Using a descriptive cross-sectional study design and through systematic random sampling technique, total of 156 caregiver-infant pairs were selected. Caregivers were interviewed using structured questionnaire, and their appointment records used to deduce adherence to appointment. Socio-demographic and clinical factors influencing adherence to appointment were analyzed using descriptive statistics, odds ratios, and univariate binary logistic regressions. An appointment adherence predictive model was developed using multivariate logistic regression and further validated using Nagelkerke’s R2 and the Receiver Operating Characteristic (ROC) curve. Overall, 43.6% of the participants reported ≥90% status of appointment adherence. Socio-demographic variables that showed significant association with adherence to clinical appointment were number of people staying with caregivers (p=0.025): age of 15-24 years (p=0.016); staying with someone (p=0.040); staying with ≥2 people (p=0.025); ≥2 forms of reminders (p <0.001); husband as a reminder (p=0.011); those with a child aged of 2 – 5 months (p=0.002); and those with HIV exposed or infected infants (p<0.001). Clinical variables that showed significant association with adherence to clinical appointment were: early ANC starters (p<0.001); optimal attendance of ANC visits (p=0.036); less than 10 clinical appointments (p=0.047); self- reporting >80% adherence (p=0.035); and with child birthweight of 2,500-3,500 grams (p=0.019). There were five variables that significantly predicted adherence to clinical appointment. These were: caregivers who stayed with ≥2 people were six times likely to be adherent (AOR 6.09, 95%CI=1.248-29.685, p=0.026); those with ≥2 forms of reminders three times adherent (AOR 3.31, 95%CI=1.106-9.881, p=0.032); early ANC starters four times adherent (AOR 4.14, 95%CI=1.526-11.242, p=0.005); with a selfreporting rate ≥81% fourteen time more adherent (14.39, 95%CI=1.322-156.647, p=0.029); whose infants had a normal birthweight four times more adherent (4.76, 95%CI=1.283-17.632, p=0.020). In conclusion, there is a challenge in the clinical appointment adherence among caregivers of infants in HIV care at the Referral Hospital. Seven socio-demographic and five clinical predictors of appointment adherence among caregiver-infant pair are identifiable of which five are of highest prediction for adherence to appointments. It is recommended that caregiver social support systems and early and serial interaction with the health system to improve clinical appointment adherence be strengthened. Results will be used to plan intervention strategies to optimize appointment management of caregivers at the hospital. The results will inform the health policy makers and further enhance new areas of research.