The current treatment for human immunodeficiency virus (HIV) infection requires strict adherence to the medication regimens to be most effective. Patients must take their medication at least 90 to 95 percent of the prescribed times to prevent development of viral resistance. Golin and colleagues performed a study to examine the factors that affect adherence to antiretroviral therapy.

They chose four categories of factors, including (1) patient factors, such as educational level, attitude toward medications, and substance abuse; (2) regimen factors, such as regimen complexity and the fit of the regimen in the patient's normal daily activities; (3) clinical interactions, such as continuity of care issues and the doctor-physician relationship; and (4) social/environmental factors, such as social support and stress.

This prospective trial followed 140 adult participants for a 48-week course and measured their adherence for at least two four-week periods. Adherence was measured every four weeks using electronic medication bottle caps that recorded every time a pill bottle was opened, pill counts, and self-reports. Participants were questioned before the study to determine their beliefs regarding the importance of taking their medication and their ability to adhere to the prescribed regimens. The results were assessed using descriptive, bivariate, and multivariate models.

The average adherence to the medication regimen was 71 percent of all prescribed doses, which is comparable to adherence in other HIV medication studies but considerably lower than the 90 to 95 percent adherence rate required to prevent treatment failure. In the final multivariate model, data from a total of 1,030 four-week periods from 117 patients revealed that black ethnicity, lower income, lower education, greater alcohol use and active drug use, and higher dose frequency negatively correlated with adherence.

The use of more medication reminders was independently associated with increased adherence. In the bivariate models, a greater dose frequency was associated with lower adherence levels. Factors in these models that were not associated with adherence included patient gender, whether the patient was in a committed relationship, whether the patient was working, number of children, literacy level, self-efficacy, beliefs about antiretroviral medications, total number of pills and number of antiretroviral medications, and fit of the regimen with the patient's lifestyle. In the category of physician-patient relationship factors, only trust in the physician correlated with adherence.

The authors conclude that most patients need aggressive modification to improve their adherence to antiretroviral regimens. Substance abuse, which was shown to have a negative effect on adherence, is amenable to counseling and treatment programs. Increasing the number of medication reminders, such as pill boxes or paging systems, may improve adherence. Regimens that reduce the daily dosing frequency potentially would be helpful. In attempting to improve adherence, physicians should remember that some factors, such as behavior counseling or access to medical care, did not influence adherence. The authors state that the association between nonadherence and some of the factors, such as black ethnicity, remain unclear.