Abstract
Adherence to antiretroviral (ARV) treatment during adolescence and young adulthood is a significant clinical issue for the current management of the HIV/AIDS epidemic in Romania. Understanding patients' own perceptions of their adherence behaviours and related psychological barriers is instrumental for developing robust interventions, and developing psychometrically sound instruments is essential for measuring adherence in this population. We adapted to Romanian an internationally validated questionnaire for the evaluation of ARV treatment adherence. We subsequently conducted a cross-sectional survey to examine its psychometric properties and investigate the relations between self-reported aspects of adherence and established indicators of adherence and health status: Pill count, doctor's assessment of patient's adherence and viral load. Results suggest that low self-reported adherence is particularly associated with experiencing side effects and emotional distress, as well as perceptions of high treatment difficulty and time demands, low self-efficacy, low treatment efficacy and low treatment satisfaction. Perceptions of improvements in health status were overall associated with increased adherence, but feeling good physically sometimes preceded non-adherence behaviours. The questionnaire proved psychometrically sound according to classical test theory criteria (e.g. Cronbach's α = 0.77, significant associations with adherence and health status indicators). Addressing adherence barriers in clinical practice with this population may help reduce their potential impact on behaviours.
Acknowledgements
This study is part of the project “Cognitive-Behavioral Determinants of Adherence among HIV-positive Adolescents in Romania”, supported by a grant from BIPAI and the Abbott Fund and developed by the Baylor Black Sea Foundation at the Centrul Clinic de Excelenţǎ – Spitalul Clinic de Boli Infecţioase Constanţa – Baylor College of Medicine – Texas Children's Hospital – Abbott Fund. The team of investigators for this project is: R.B. Wanless (Baylor International Pediatric AIDS Initiative, Houston), A.L. Dima (School of Medicine, University of Southampton, UK), A.-M. Schweitzer, L. Vlahopol, E.A. Caraveţeanu, I. Stochiţǎ, R. Diaconiţǎ (Baylor Black Sea Foundation at the Centrul Clinic de Excelenţǎ – Spitalul Clinic de Boli Infecţioase Constanţa – Baylor College of Medicine – Texas Children's Hospital – Abbott Fund, Constanţa, Romania), R.K. Amico (Applied Health Research, Michigan; Center for Health, Intervention, and Prevention, University of Connecticut, USA), E. Remor (Facultad de Psicología, Universidad Autónoma de Madrid), S. Ruta (Institutul de Virusologie Ştefan Nicolau, Romania), and S. Rugina (Spitalul Clinic de Boli Infecţioase, Constanţa, Romania). Part of this data was presented as a poster session at the 24th Conference of the European Health Psychology Society “Health in Context”, Cluj-Napoca, Romania, 1–4 September 2010.
Notes
1. Researchers interested in the CEAT-VIH and user's manual should contact the original author at [email protected]
2. Kendall τ is reported to provide better estimates of the population parameters compared to Spearman's ρ (Field, 2005; Howell, 2010). We also computed Spearman's ρ and Pearson's r where appropriate, and the results were extremely similar.
3. Time intervals between study procedures ranged from 0 to 479 days, therefore criterion validity was also examined via partial correlations controlling for time intervals between measurements (Table ).
4. Item no. 5 had low item-total correlation (0.18, p = 0.03) and low correlations with the other four items (–0.04 to 0.13), reducing the subscale's Cronbach's α to 0.55. It was therefore excluded from this analysis.