Munoz-Moreno JA, Tuldra A, Fumaz CR, Ferrer MJ, Perez-Alvarez N, Garrido N, Clotet B; International Conference on AIDS (15th : 2004 : Bangkok, Thailand).

Int Conf AIDS. 2004 Jul 11-16; 15: abstract no. WePeB5844.

Lluita contra la Sida Fdn. - HIV Unit. Germans Trias i Pujol University Hospital, Barcelona, Spain

Background: Multiple factors have been associated to adherence (ADH) to antiretroviral treatment in HIV infected patients (PTS). Interventions and recommendations have also been determined to improve this ADH. In this study we describe variables that have been suggested to be studied and we provide different predictive models. Methods: 457 PTS were evaluated. ADH was measured with SERAD questionnaire. This questionnaire permits to obtain ADH percentages according to PTS' self-report. PTS were considered adherent if they took >=95% of medication prescribed. These variables were controlled: age, gender, risk practice, 16 personality (P) scales and 5 general traits (with 16PF-5 questionnaire), depression (BDI test), toxic substances' consumption (TSC), weeks on treatment, daily number of pills (NP), CD4 cell count, viral load and previous antiretroviral treatments. A logistic regression model was elaborated and different classification trees were composed. Results: The univariate model provided these significant variables: TSC, depression, NP and age. Relevant P scales were: emotional stability, sensitivity, abstractedness and apprehension, and anxiety and independence. 4 multivariate models were obtained: NP and sensitivity scale (82,9% of correct classification), NP and serious depression (83,6%), TSC, NP and independence (82,3%) and NP, age, sensitivity and independence (83,5%). The classification trees were established as predictive models for ADH. The more significant predictive factors in these 3 models were the following P variables: perfectionism, tension, sensitivity, abstractedness, vigilance and warmth, and independence, tough-mindedness, anxiety and extraversion. The age was also a relevant predictive variable for ADH. Conclusions: This study demonstrates that predictive models for ADH to antiretroviral treatment may be composed with reliability. We can conclude that variables such as the NP, TSC and age, or P dimensions such as the independence, tough-mindedness, anxiety and extraversion may provide significant information to predict this ADH.

Publication Types:

• Meeting Abstracts

Keywords:

• Acquired Immunodeficiency Syndrome
• Antiretroviral Therapy, Highly Active
• Anxiety
• CD4 Lymphocyte Count
• Data Collection
• Drug Therapy, Combination
• HIV Infections
• HIV Seropositivity
• Humans
• Models, Biological
• Questionnaires
• Viral Load
• drug therapy
• therapy

Other ID:

• GWAIDS0039555

UI: 102283771

From Meeting Abstracts