Antiretroviral Medication Compliance in Patients with AIDS

Abstract

The paradigm of AIDS patient care has evolved to that of a chronic disease that is manageable with combination antiretroviral therapy. Intermittent adherence to antiretroviral regimens, however, has been associated with the selection of HIV mutations, resulting in drug-resistant virus. Medication compliance has become a vital component in the care of HIV-infected patients. This study was designed to assess the degree of medication compliance with zidovudine (ZDV) over a 2-month period among a convenience sample of 23 ambulatory patients with HIV infection. Enrollment took place during February to March 1995, when monotherapy with ZDV was considered the standard of care. Medication compliance was assessed by three methods: patient self-report determined by questionnaire, pharmacy refill records from the 3 months immediately before entry in the study, and an electronic monitoring system (Medication Event Management Systems [MEMS], Aprex Corp., Fremont, CA), which records the date and time of each opening of the medication vials. MEMS was utilized among a subgroup of eight participants over a 2-month period. Despite greater than 95% (22 of 23) of the subjects reporting that they believed ZDV was life prolonging, a majority took < or = 75% of the prescribed doses of ZDV as identified by both refill and MEMS methods. The mean percentage compliance over a 2-month period of observation for MEMS and pharmacy refill record review methods was 66% and 78%, respectively (p = 0.294). Among the subgroup of eight in the MEMS evaluation, 88% reported taking their ZDV according to the directions, all of the time. These results demonstrate that reliance upon patient self-report of medication compliance is less accurate than MEMS or pharmacy refill records, and that reliance on self-reporting could lead to erroneous assumptions of the patient's true drug compliance. In addition, the study suggests that pharmacy refill records may provide a method of assessing compliance that is equivalent to MEMS but is less experimental, and can easily be followed and interpreted by pharmacists and other clinicians caring for patients infected with HIV. Most importantly, relatively poor compliance rates demonstrated by MEMS raises serious concern for widespread development of HIV resistance to the more complicated, multiple-drug regimens in present use.

Perceived compliance with AZT dosing among a sample of African-American drug users

“…AZT was the most commonly reported medication among participants. Although other antiretroviral medications were reported, as shown below, consistent with a ® nding reported by Jeffe et al 14 , AZT was the most commonly recognized and reported antiretroviral. Parametric (Pearson's r) and non-parametric (Spearman's rho) correlational analyses were used to assess bivariate associations between self-reported compliance with AZT dosing and sociodemographic, drug use, sexual behaviour, and health belief variables.…”

A Randomized Study of Serial Telephone Call Support to Increase Adherence and Thereby Improve Virologic Outcome in Persons Initiating Antiretroviral Therapy

“…Targeting interventions to patients likely to have lower adherence may be a more useful approach. Adherence self-report measurements may overestimate adherence to study regimens, compared with more objective measurements, such as pill counts or medication-event monitors [42][43][44]. However, there has been a high correlation between self-reported and objective adherence measurements, and lower adherence by self-report has been correlated with lower virologic responses [43][44][45][46][47][48].…”

Use of Electronic Monitoring Devices to Measure Antiretroviral Adherence: Practical Considerations

“…Self-report, pharmacy records, and pill counts tend to overestimate patient adherence by anywhere from 20 to 30% (Burney et al, 1996;Farmer, 1999;Frick et al, 1998;Liu et al, 2001;Schwed et al, 1999;Wagner and Rabkin, 2000). Clinicians also tend to overestimate medication adherence and inadequately detect poor adherence (Miller et al, 2002).…”