OBJECTIVE: To know the non-compliance of hypertensive patients with scheduled visits, to define the patient profile, and to study the best method which identifies such patients. PATIENTS AND METHODS: A total of 174 hypertensive patients under pharmacological therapy in the Primary Care setting. Non-compliance with scheduled visits was assessed by review of the appointment book. The characteristics of patients, of the disease, and therapy associated with non-compliance were studied and six indirect methods were validates. RESULTS: The profile of the non-compliant patient corresponded to a housewife or with working activity (p = 0.01), obesity (p = 0.01), anxiety (p = 0.008), moderate-severe cardiovascular risk (p = 0.02), bad control of blood pressure values (p = 0.03), and physician's dependence (p = 0.001). Of the investigated methods, education level (EL) had the highest sensitivity (77.8%) (CI: 65.7%-89.9%) and medical judgement (MJ) the highest specificity (87.6; CI: 81.9%-93.3%), negative predictive value 80.1%, positive predictive value 51.5%, odds ratio of low compliance 3.1 and best Kappa index 0.28. CONCLUSIONS: Factors predicting non-compliance are few but delineate a possible profile. Only MJ is valid to detect non-compliance with scheduled visits.
OBJECTIVE: To know the non-compliance of hypertensivepatients with scheduled visits, to define the patient profile, and to study the best method which identifies such patients. PATIENTS AND METHODS: A total of 174 hypertensivepatients under pharmacological therapy in the Primary Care setting. Non-compliance with scheduled visits was assessed by review of the appointment book. The characteristics of patients, of the disease, and therapy associated with non-compliance were studied and six indirect methods were validates. RESULTS: The profile of the non-compliant patient corresponded to a housewife or with working activity (p = 0.01), obesity (p = 0.01), anxiety (p = 0.008), moderate-severe cardiovascular risk (p = 0.02), bad control of blood pressure values (p = 0.03), and physician's dependence (p = 0.001). Of the investigated methods, education level (EL) had the highest sensitivity (77.8%) (CI: 65.7%-89.9%) and medical judgement (MJ) the highest specificity (87.6; CI: 81.9%-93.3%), negative predictive value 80.1%, positive predictive value 51.5%, odds ratio of low compliance 3.1 and best Kappa index 0.28. CONCLUSIONS: Factors predicting non-compliance are few but delineate a possible profile. Only MJ is valid to detect non-compliance with scheduled visits.