Responses to a 1 month self-report on adherence to antiretroviral therapy are consistent with electronic data and virological treatment outcome

Abstract

Many questionnaires on adherence to antiretroviral therapy are in use, but the validity of patients' responses has not been tested. The Medication Adherence Self-Report Inventory (MASRI) has been developed and tested for its validity against objective measures and treatment outcome. Prospective study comparing questionnaire responses with MEMS TrackCap (MC, a medication event monitoring system), pill count (PC) and plasma HIV viraemia in a publicly funded specialist HIV clinic. Patients self-medicating antiretroviral therapy who were not cognitively impaired and were able to read and understand English. Mean adherence by MC of the 78 subjects was 92.9% (SE, 1.8%) and by PC 96.8% (SE, 1.4%). Agreement between MC and responses to items about doses missed 1, 2 or 3 days ago was low (kappa = 0.23 (P < 0.03), 0.44 (P < 0.001) and 0.28 (P < 0.01) respectively). This improved when these responses were summated (kappa = 0.46;P < 0.001) and was similar to that for recall of non-adherence over the preceding 2 weeks (kappa = 0.54; P < 0.001). Mean self-reported adherence by visual analogue scale (VAS) over the preceding month was 93.3% (SE, 1.2%). This was strongly associated with both MC (r = 0.63; P < 0.001) and PC (r = 0.75; P < 0.001). On multivariate analysis, the strongest association between a MASRI item and MC was for the VAS. Both the 2 week recall and VAS items were inversely associated with viral load (P = 0.01). There was no association between dose timing (measured MC or questionnaire) or 3 day self-report and viral load. The MASRI provides a means of measuring patient adherence that is valid when compared with objective measures.