| |
ADHERENCE in HIV: difficulties in evaluating
|
| |
| |
Is the Influence of Social Desirability on Patients' Self-Reported Adherence Overrated?
JAIDS Journal of Acquired Immune Deficiency Syndromes: Volume 35(2) 1 February 2004 pp 203-204
Wagner, Glenn PhD*; Miller, Loren G. MD, MPH†. *RAND, Santa Monica, CA. †Harbor-UCLA Medical Center and the Harbor-UCLA Research and Education Institute, Torrance, CA, and UCLA AIDS Institute, Los Angeles, CA Supported by NIDA grant R01DA13303
To the Editor:
Adherence is critical to the success of HIV antiretroviral therapy. Accordingly, clinicians need to be able to assess a patient's adherence accurately so as to inform optimal treatment decision making. Because physicians are poor predictors of a patient's adherence, accurate and reliable measures of adherence are needed to help clinicians identify patients who need assistance with their pill-taking behavior and to interpret the role of adherence in evaluating clinical outcomes. Electronic monitoring is considered by many to be the current state-of-the art method of evaluating adherence; yet, this technology is limited by its high cost and the restrictions it places on common adherence strategies (eg, pill boxes, pocketed doses). In contrast, self-report measures are much more practical because of their relatively low cost and ease of administration; however, self-reported rates of antiretroviral adherence average 10% to 20% greater than rates derived from electronic monitoring, and their ability to explain virologic outcomes is inferior to electronic monitoring, resulting in the consensus that self-reports overestimate adherence.
The reasons why patients often overestimate their adherence are poorly understood, and there are scarce data on the factors associated with self-report accuracy. We have previously found that patients who take all their doses are nearly always accurate in their selfreport (using electronic monitoring as the validity criterion), whereas only a small proportion (approximately 20%) of patients who miss doses are accurate in their self-report. Self-reports of missed doses are reliable in terms of the patient having actually missed a dose, but the number of doses missed is often inaccurate and underestimated. The aforementioned investigation was not powered to identify patient characteristics associated with self-report accuracy, but there was a trend toward those with lower cognitive functioning (eg, memory problems) being more likely to report adherence inaccurately.
Clinicians and researchers often assume that inaccurate self-reports of adherence are a result, at least in part if not mostly, of patients' social desirability (ie, patients desire to be seen as a good patient) and discomfort in disclosing nonadherent behavior. To our knowledge, there has been only 1 investigation that explored the relationship between the accuracy of self-reported adherence and treatment and social desirability, in which no association was detected, although this study investigated adherence to nonmedication treatments. We are not aware of any studies that have examined the relationship between social desirability and medication adherence.
Assuming that social desirability influences self-reports, we hypothesized that minimizing social desirability would enhance patients' accuracy of self-reported adherence. To test this hypothesis, we conducted a pilot study involving a 2-week observation of adherence among persons on HIV antiretroviral therapy. Participants were randomly assigned (using a 2:1 ratio) to either the experimental or control group. Participants in the experimental group were administered a standard 3-day recall self-report adherence measure (similar to that used in ACTG trials) at study end point (week 2), preceded by a discussion of the challenges and difficulties that are commonly associated with adherence to antiretrovirals (enhanced survey). In this brief discussion (approximately 5-10 minutes), the interviewer acknowledges the challenges of treatment and adherence, normalizes lapses in adherence, and encourages the respondent to discuss difficulties with adherence. It was hypothesized that this discussion would reduce discomfort or anxiety that the respondent may have regarding disclosure of nonadherence and therefore result in more accurate reporting of adherence. Participants in the control group did not receive this discussion prior to being administered the adherence measure (standard survey). Electronic monitoring caps (eDEM caps; AARDEX, Zurich, Switzerland) were used to monitor adherence to the antiretroviral with the most complex regimen (defined as the most frequently dosed medication), and these data served as the criterion for evaluating the accuracy of the self-report.
Thirty-four participants on HIV antiretroviral therapy enrolled in the study; mean age was 41 years, 82% were male, and 74% were nonwhite. At baseline, mean self-reported CD4 count was 398 (18% <200) and 48% reported having an undetectable viral load. Average length of time on the current antiretroviral regimen was 12 months, and all were on twice-daily dosing regimens. Four participants did not complete the study (1 died of heart failure, 3 were lost to follow-up). Of the 30 study completers, 21 were randomized to receive the enhanced survey (experimental group), whereas the other 9 were administered the standard survey (control group). The sample's mean electronically monitored adherence rate was 66% (SD = 33) over the entire 2-week observation period and 65% over the past 3 days; mean self-reported adherence (past 3 days) was 83% (SD = 26), which was highly correlated with electronic monitoring (past 2 weeks: R = 0.66, P < 0.001; past 3 days: R = 0.69, P < 0.001). Twelve (40%) participants took all doses (100% adherence) in the past 3 days according to electronic monitoring. Twelve (57%) of the 21 participants in the experimental group accurately reported their adherence compared with 3 (33%) of 9 in the control group (P = 0.21). Among the participants who had missed doses in the last 3 days of the study period (12 in the experimental group, 6 in the control group) as measured by electronic monitoring, however, a similar proportion of participants in the experimental and control groups accurately reported at least 1 missed dose (75% and 67%, respectively) and the total number of missed doses (25% and 17%, respectively).
After the adherence assessment, the electronic monitoring data were reviewed with the participants in both groups; those who inaccurately reported their adherence were asked to provide reasons for the discrepancy between the electronic monitoring data and their self-report. Among the 14 participants who reported taking more doses than what was indicated by the electronic caps, 6 (43%) said that they had taken the dose(s) and questioned whether the cap was malfunctioning, whereas the other 8 (57%) stated that they must have mistakenly thought they had taken the dose(s) due to forgetfulness. In instances where the functioning of the cap was questioned, the project staff tested the cap for a 3-day period and the cap correctly recorded all openings in all cases. Only 1 participant reported taking fewer doses than what was indicated by electronic monitoring, and he attributed the discrepancy to poor memory.
At the closing interview, all participants (30 completers) were asked if they felt any discomfort in talking to their provider about missed doses or if they had ever intentionally not told their provider about missed doses. Only 5 (17%) participants reported being uncomfortable telling their provider about missed doses, and only 6 (20%) reported ever falsely telling their provider they had not missed doses. When provided with a list of reasons why some patients do not tell their provider about missed doses, 42% cited poor memory for not accurately disclosing missed doses, whereas concerns about making their provider angry and concerns that their provider would change their treatment were each cited by 4 (13%) participants. Finally, 90% of the participants stated that they were equally comfortable disclosing missed doses to their provider as they were to other clinic staff such as a nurse or case manager or even a friend or family member.
These findings are somewhat mixed with regard to the potential influence of social desirability on accuracy of self-reported adherence. Although the study's power was limited by the small sample size, the lack of difference between the experimental and control groups with regard to accuracy, particularly in the subgroup of patients who had missed doses, may suggest that the challenges to adherence discussion prior to the adherence assessment may be of limited value in diminishing socially desirable responses or discomfort with disclosing missed doses. Alternatively, the effects of this challenges to adherence discussion may have been irrelevant if social desirability did not play a significant role in self-reported adherence among these participants, as supported by the large number of subjects who reported being comfortable disclosing nonadherence to their providers. Social desirability may have been a factor among participants who attributed the discrepancy between their self-report and electronic monitoring data to the cap malfunctioning, however. What is more clear is that patients seem to have difficulty tracking even recent (past 3 days) pill-taking behavior. The repetitive daily routine of pill taking may contribute to the poor self-monitoring and inaccurate cognitive perception of adherence demonstrated by many patients.
The need for accurate and practical adherence measures for widespread clinical use calls for the development of improved self-report methodology and methods to assist patients in accurately recalling pill-taking behavior. Our results do not discount the potential role of social desirability on the accuracy of self-reported adherence, but our findings do suggest that social desirability may not be as significant a factor as suspected. It is possible that the imperfections of human memory may be responsible for a significant proportion of inaccurate reports of medication adherence. There needs to be further research on the reasons behind inaccurate self-reported adherence. Improved self-reported measures will forward the ability of clinicians and investigators to assess adherence accurately and to identify patients who need intervention to improve medication adherence.
Glenn Wagner, PhD
Loren G. Miller, MD, MPH
|
|
| |
| |
|
|
|
