Highly active antiretroviral therapy: physician experience and enhanced adherence to prescription refill.

2003 
Objective: To identify patient and physician characteristics that may act as determinants of adherence to prescription refill of triple combination antiretroviral therapy. Methods: A population-based analysis of antiretroviral therapy-naive HIV-positive men and women in British Columbia, Canada, who initiated triple combination therapy between August 1 1996 and October 31 1998. Study participants were considered adherent if they were actually dispensed antiretrovirals ≥95% over the first year of therapy. Log-binomial regression was used to identify patient and physician characteristics associated with adherence to prescription refill. Results: Of the 886 individuals eligible for analysis, 495 (56%) were ≥95% adherent to prescription refill. In multivariate analysis, adherence was positively associated with increased age [adjusted relative rate (ARR) 1.19; 95% Cl: 1.07-1.32], having a diagnosis of AIDS (ARR 1.66; 95% Cl: 1.29-2.15), being male (ARR 1.79; 95% Cl: 1.27-2.53), and with greater experience of the treating physician (ARR 1.27; 95% Cl: 1.13-1.42). History of injection drug use was negatively associated with adherence to prescription refill (ARR 0.65; 95% Cl: 0.51-0.83), as was increased pill burden (per pill daily) (ARR 0.95; 95% Cl: 0.92-0.99). A sub-analysis of 316 patients who provided additional data regarding psychosocial characteristics indicated that adherence was positively associated with physician experience (ARR: 1.28; 95% Cl: 1.09-1.51) and being employed (ARR: 1.55; 95% Cl: 1.14-2.21), and negatively associated with a history of injection drug use (ARR: 0.61; 95% Cl: 0.43-0.85). Conclusion: While patient disease stage and personal characteristics may play an important role in patient adherence to prescription refill of complex therapeutic regimens, our findings indicate that HIV-experienced physicians may have greater success in maintaining patients on prescribed therapy.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    56
    Citations
    NaN
    KQI
    []