OBJECTIVES: To validate 4 indirect methods for detecting therapeutic non-compliance in 157 patients in a community health center Centro de Salud "Carrus" in Elche, Alicante. MATERIALS AND METHODS: The methods employed are self related compliance (SC), therapeutic gains obtained, Morinsky-Green Test, and the level of knowledge of Batalla disease (LK). For verification purposes, pills were counted, allowing the formation of 2x2 tables to calculate indices of validity. The statistical analysis employed is the calculation of 95% confidence intervals. RESULTS: SC yields significantly the lowest sensitivity (13.7 +/- 7.9%) and the highest specificity (85.7% +/- 7.5%). The LK presents the highest sensitivity 79.5 +/- 9.3%. SC underestimates non-compliance (32.5%) significantly, and the LK overestimates noncompliance (23.6%). CONCLUSIONS: The LK is a good method for detecting problems of low compliance and SC those of high compliance. It is recommended that both methods should be employed in a systematic fashion in clinical practice. However, the LK and SC cannot be used to calculate prevalences of therapeutic noncompliance in population-based studies.
OBJECTIVES: To validate 4 indirect methods for detecting therapeutic non-compliance in 157 patients in a community health center Centro de Salud "Carrus" in Elche, Alicante. MATERIALS AND METHODS: The methods employed are self related compliance (SC), therapeutic gains obtained, Morinsky-Green Test, and the level of knowledge of Batalla disease (LK). For verification purposes, pills were counted, allowing the formation of 2x2 tables to calculate indices of validity. The statistical analysis employed is the calculation of 95% confidence intervals. RESULTS: SC yields significantly the lowest sensitivity (13.7 +/- 7.9%) and the highest specificity (85.7% +/- 7.5%). The LK presents the highest sensitivity 79.5 +/- 9.3%. SC underestimates non-compliance (32.5%) significantly, and the LK overestimates noncompliance (23.6%). CONCLUSIONS: The LK is a good method for detecting problems of low compliance and SC those of high compliance. It is recommended that both methods should be employed in a systematic fashion in clinical practice. However, the LK and SC cannot be used to calculate prevalences of therapeutic noncompliance in population-based studies.
Authors: Carmen Valdés Y Llorca; Ernesto Cortés Castell; José Manuel Ribera Casado; Pilar de Lucas Ramos; José Luis Casteig Ayestarán; Amaia Casteig Blanco; Vicente Francisco Gil Guillén; Mercedes Rizo Baeza Journal: Int J Environ Res Public Health Date: 2021-04-19 Impact factor: 3.390