Evaluation of adherence to antiretroviral therapy in pediatrics

Author(s)

Aline Santarem Ernesto

• Document type: Master's Dissertation
• Press: Campinas, SP.
• Institution: Universidade Estadual de Campinas (UNICAMP). Faculdade de Ciências Médicas
• Defense date: 2011-08-29

Examining board members

Marcos Tadeu Nolasco da Silva; Jorge Andrade Pinto; José Dirceu Ribeiro

• Advisor: Marcos Tadeu Nolasco da Silva

Abstract

Background: The Acquired Immunodeficiency Syndrome (Aids) is currently a chronic disease, manageable by Highly Active Antiretroviral Therapy (HAART). In a setting of universal access to treatment, patient adherence arises as a limiting and challenging issue. This study aimed to evaluate the prevalence of nonadherence to HAART in a cohort of children and adolescents, using complementary instruments, and also identify and understand associated factors. Patients and Methods: Observational, analytical, prospective, cross-sectional study. The study population comprised 108 Human Immunodeficiency Virus (HIV) -infected patients on HAART (60 boys), from 8 to 19 years-old. Adherence was evaluated by a standardized questionnaire, pharmacy refill data (PRD) and a self-efficacy scale. Patients or caregivers were interviewed (whoever was in control of medicine administration). Patients who received less than 95% of prescribed doses in the 24 hours of 7 days before the interview, or who had a record of an interval of more than 37 days between refills, were considered nonadherent. The self-efficacy scale provided a continuous score, varying from 0 to 100. The association between adherence outcomes and independent variables related to demographical, clinical, immunological, virological and psychosocial conditions was estimated. Statistical analysis was performed with the use of Odds Ratios (OR) for comparison between categorical variables, Mann-Whitney test for comparison between continuous variables and categories, and Spearman Correlation Coefficient (rs) for comparison between continuous variables. Results were considered statistically significant if p _ 0.05. Confounding variables were controlled by multivariate analysis with logistic regression. The study was approved by the Human Research Ethics Committee of the State University of Campinas Faculty of Medical Sciences. Results: Nonadherence prevalence varied from 11.1% (nonadherent patients in 3 instruments), 15.8% (24-hour self-report), 27.8% (7-day self-report), 45.4% (PRD) and 56.3% (at least one of the outcomes). Self-reports from 24 hours and 7 days, when compared to PRD, showed low sensitivity (29% and 43%, respectively) and high specificity (95% and 85%, respectively). The following independent variables showed statistically significant association with nonadherence on univariate analysis, according to each instrument: Twenty-four hour self-report: difficulty of ministration by caregiver (OR = 9.11 ; 95%CI = 2.87 - 28.98); lack of religious practice by patient (OR = 2.76; 95%CI = 0.92 - 8.32); medication intolerance (OR=4.61; 95%CI =1.47 - 14.42); per capita income (median US$ 137.91 vs US$ 208.33; p = 0.016); number of nucleoside/nucleotide analogues (NRTIs) with resistant mutations (median 6 vs 1; p = 0.016); Seven-day self-report: difficulty of ministration by caregiver (OR = 2.91; 95%CI = 1.05 - 8.12; HAART ministration by the patient (OR = 3.59; 95%CI = 1.47 - 8.78); caregiver with 8 or less years of school attendance (OR = 3.25; 95%CI = 1.03 - 10.30); patient with 8 or more years of school attendance (OR = 3.70; 95%CI = 1.41 - 0.70); patient age (median 13.94 vs 12.94; p = 0.03); per capita income (median US$ 131.67 vs US$ 201.39; p = 0.009); Pharmacy refill data: difficulty of ministration by caregiver (OR = 3.19; 95%CI = 1.11 - 9.17); HAART ministration by the patient (OR = 2.70; 95%CI = 1.15 - 6.33); lack of virological control (OR = 3.70; 95%CI = 1.67 - 8.33); missed consultations in the former 6 months (OR = 3.27 (95%CI = 1.38 - 7.78); lack of religious practice by patient (OR = 2.47 (95%CI = 1.10 - 5.57); lack of religious practice by caregiver (OR = 3.19; 95%CI = 1.36 - 7.50); caregiver working outside the home (OR = 2.27; 95%CI = 1.05 - 4.92); per capita income (median US$ 166.67 vs US$ 222.22 ; p = 0.014); The following independent variables showed statistically significant association with nonadherence on multivariate analysis, according to each instrument: Twenty-four hour self-report: medication intolerance (OR = 9.11; 95%CI = 2.87 - 28.98); Seven-day self-report: difficulty of ministration by caregiver (OR = 2.91; 95%CI = 1.05 - 8.12); HAART ministration by the patient (OR = 3.59; 95%CI = 1.47 - 8.78); socioeconomical classes C+D (3.54; 0.97 - 2.85); Pharmacy refill data: lack of virological control (OR = 3.73; 95%CI = 1.68 - 8.31); missed consultations in the former 6 months (OR = 3.27 (95%CI = 1.38 - 7.78); lack of religious practice by caregiver (OR = 3.19; 95%CI = 1.36 - 7.50); The self-efficacy score had a median of 95.20 (11.90 - 100) and showed significant association with difficulty of ministration by caregiver (median 78.5 vs 95.2; p = 0.001), lack of virological control (median 90.4 vs 100; p = 0.001), HAART ministration by patient (median 89.8 vs 95.2; p = 0.05), missed consultations in the former 6 months (median 86.3 vs 100, p < 0.001), clinical categories N, A or B ( median 90.47 vs 100; p = 0.018), lack of religious practice by patient (median 90.4 vs 95.2, p = 0.037), being orphan (median 95.2 vs 90.4 p = 0.05), CD4/CD8 ratio (rs = 0.220; p = 0.025), number or antiretroviral classes with resistance (rs = 0.583; p < 0.001), number of NRTIs with resistance (rs = 0.44; p = 0.009), per capita income (rs = 0.302; p = 0.001), PedsQL score, emotional domain (rs = 0.265; p = 0.007). Conclusion A high prevalence of HAART nonadherence was observed in the study population, being pharmacy refill data the most sensitive measurement. Additionally, an association was observed between adherence outcomes and self-efficacy scores. The instruments employed showed complementarity in the recognization of nonadherence risk factors. Aiming to eliminate confounding variables, seven factors were identified as associated to lack of adherence: medication intolerance, difficulty of ministration by the caregiver, ministration of medicines by the patient, lower socioeconomic class, lack of virological control, lack of religious practice by the caregiver and missed consultations (AU)

• FAPESP's process: 09/06948-0 - Adherence to antiretroviral therapy in adolescents
• Grantee: Aline Santarem Ernesto
• Support Opportunities: Scholarships in Brazil - Master