Measuring Patient Adherence

IN RESPONSE:

We appreciate the thoughtful comments of Dr. Liu and his coauthors. We understand that the authors feel that their adherence score reflects a composite of several approaches to measure adherence. However, the authors first used electronic monitor data and then filled in missing data points with pill count and, finally, interview data. Our editorial aimed to help the reader understand that electronic monitor data served as the centerpiece of this approach. If a researcher had nearly complete adherence data on his or her study participants from electronic monitors, Liu and colleagues' approach would require minimal additional data from pill counts or patient interviews. Other potential weaknesses of electronic monitors, such as patients “decanting” multiple doses at once, are not addressed by Liu and colleagues' approach. Liu and colleagues' letter reiterates their preference for using electronic monitors to measure adherence and their reluctance to trust patient self-report. In our editorial, we described a study that observed a significant relationship between self-reported adherence to antiretroviral therapy and viral load (1). We believe that carefully collected self-report data still offer promise in evaluating patient adherence to medications. However, strenuous efforts are needed to reduce the known biases of self-report data, such as social desirability, time frame covered by the questions, type of interviewer, language barriers, and poor recall. In many clinical settings, self-report is the best measure we have. In research settings, self-report remains a far less costly, complex, and intrusive method of determining how patients take their medications. Vitolins and coworkers (2) recently reviewed the strengths and weaknesses of multiple adherence measures but concluded that the best option is to use self-report with other “objective” measures. They acknowledged that it is necessary to make the best of a difficult situation but did not dismiss self-report. Similarly, we hope that researchers will continue to investigate ways to improve and supplement self-report as a measure of adherence.