Is it possible to diagnose the therapeutic adherence of patients with COPD in clinical practice? A cohort study

BACKGROUND: Therapeutic adherence of patients with chronic obstructive pulmonary disease (COPD) is poor. It is therefore necessary to determine the magnitude of non-adherence to develop strategies to correct this behaviour. The purpose of this study was to analyse the diagnostic validity of indirect adherence methods. METHODS: Sample: 195 COPD patients undergoing scheduled inhaled treatment attending 5 Primary Care Centres of Malaga, Spain. Variables: Sociodemographic profile, illness data, spirometry, quality of life (St. George Respiratory Questionnaire: SGRQ), and inhaled medication counting (count of dose/pill or electronic monitoring) were collected. The patient's knowledge of COPD (Batalla test:BT),their attitude towards treatment (Morisky-Green test: MGT) and their self-reported therapeutic adherence (Haynes-Sackett test: HST) were used as methods of evaluating adherence. The follow-up consisted four visits over one year (the recruitment visit: V0; and after 1 month:V1; 6 months:V2; and 1 year:V3). RESULTS: The mean age was 69.59 (95% CI, 68.29-70.89) years old and 93.8% were male. Other findings included: 85.4% had a low educational level, 23.6% were smokers, 71.5% mild-moderate COPD stage with a FEV1 = 56.86 (SD = 18.85); exacerbations per year = 1.41(95% CI, 1-1.8). The total SGRQ score was 44.96 (95% CI, 42.46-47.46), showing a mild self-perceived impairment in health. The prevalence of adherence (dose/pill count) was 68.1% (95% CI, 60.9-75.3) at V1, 80% (95% CI, 73-87) at V2 and 84% (95% CI, 77.9) at V3. The MGT showed a specificity of 67.34% at V1, 76.19% at V2 and 69.62% at V3. The sensitivity was 53.33% at V1, 66.66% at V2 and 33.33% at V3.The BT showed a specificity of 55.1% at V1, 70.23% at V2 and 67.09% at V3. The sensitivity was 68.88% at V1, 71.43% at V2 and 46.66% at V3. Considering both tests together, the specificity was 86.73% at V1, 94.04% at V2 and 92.49% at V3 and the sensitivity was 37.77% at V1, 47.62% at V2 and 13.3% at V3. CONCLUSIONS: The prevalence of treatment adherence changes over time. Indirect methods (dose/pill count and self-reported) can be useful to detect non-adherence in COPD patients. The combination of MGT and BT is the best approach to test self-reported adherence.