627 
Adherence to Antiretroviral Therapy (ART) and Its Principal Determinants in HIV-infected Adults in India
S Pujari*1, A Sarna2, A Sengar3, R Garg4, S Katke1, R Prasad5, and J van Dam2
1Ruby Hall Clin and Grant Med Fndn, Pune, India; 2Horizons/Population Council, New York, NY, USA; 3Northern Railway Hosp, Delhi, India; 4Employees State Insurance Corp, Delhi,India; and 5Horizons/Population Council, New York, NY, USA
Background: High levels of adherence (> 95%) are
required for successful ART. A diagnostic study was undertaken to explore
existing levels of adherence to therapy among HIV+ persons currently on ART in Pune and Delhi, India
Methods: Using a cross-sectional study design 310 HIV+
patients on ART at 1 private-sector
(i.e., mostly paying out-of-pocket) and 3 public-sector (free ARV medications)
hospitals were interviewed using a semi-structured questionnaire adapted from AIDS
Clinical Trial Group (ACTG) adherence instruments. Key themes of analysis
included adherence to ART, reasons
for missed doses, factors influencing adherence (i.e., financing mechanisms,
social support, disclosure, alcohol and drug use, depression according to the
BDI_II index, trust in provider, clinical stage, and improvement in CD4 counts).
Adherence was measured using self-report; and the primary outcome measure was
adherence < 90% over 4-day recall (lower adherence). In the private-sector
hospital careful assessment of patient readiness before starting ART and physician-based adherence support was
provided as part of routine care. Chi square and T test were used to assess
significance between the groups for categorical and continuous variables.
Independent predictors of lower adherence were determined using logistic
regression analysis.
Results: Adherence among 310 patients (252 paying, 58
free) was studied. Mean 4-day adherence was 96.4% and 80.5% (p = 0.001) and 89.7% vs
60.3% patients reported a 4-day adherence > 90% (p = 0.000) among the 2 groups respectively. Commonly-cited reasons
for missing doses over past 7 days included being away from home (26%), busy
with other things (21.9%) and running out of pills (19.8%). On univariate analysis age group (26 to 45 years), less than
university education, being unemployed, obtaining free treatment, severe
depression, baseline CD4 count > 200/mm3, hospitalization > 2
times, and taking 4-or-more medicines were associated with lower adherence.
However, only obtaining free treatment (adjusted odds ratio (OR), 4.48, 95%
confidence interval (CI) 2.1 to 9.3, p
= 0.000), severe depression (adjusted OR 4.91, 95% CI 2.0 to 11.8, p.000), and
age group 26 to 45 years (adjusted OR 7.7, 95% CI 1.1 to 39, p = 0.014) were associated with lower
adherence in multivariate analysis.
Conclusion: Overall adherence to ART
is good among this group of patients. Provision of free treatment without
adequate adherence support may compromise the success of ART scale-up programs. Designing
interventions to support clients at risk for non-adherence need special focus.
Keywords: adherence; antiretroviral therapy; HIV infected persons
