Health Complexity Assessment in Primary Care: a validity and feasibility study of the INTERMED tool

Abstract

Background: While considerable attention has been devoted to patients health complexity epidemiology, comparatively less attention has been paid to tools to identify and describe, in a personalized and comprehensive way, complex patients in primary health care (PHC). Objective: To evaluate INTERMED tools validity and feasibility to assess health complexity in PHC. Design: Cross-sectional psychometric study. Setting: Three Brazilian PHC Units. Participants: 230 patients above 18 years of both sexes. Measurements: Spearmans rho assessed concurrent validity between the whole INTERMED and their four domains (biological, psychological, social, health system) with other well-validated instruments. Pearsons X2 measured associations of the sum of INTERMED current state items with use of PHC, other health services and medications. Cronbach Alpha assessed internal consistency. INTERMED acceptability was measured through patients views on questions and answers, understanding and application length as well as objective application length. Applicability was measured through patients views on its relevance to describe health aspects essential to care and INTERMEDs items-related information already existing in patients health records. Results: 18.3% of the patients were complex (INTERMED 20/21 cut-off). Spearman correlations located between 0.44 - 0.65. Pearson coefficients found were X2 = 26.812 and X2 = 26.883 (both p = 0.020) and X2 = 28.270 (p = 0.013). Cronbach Alpha was 0.802. All patients views were very favorable. Median application time was 7 minutes and 90% of the INTERMED interviews took up to 14 minutes. Only the biological domain had all its items described in more than 50% of the health records. Limitations: We utilized the cutoff point used in all previous studies, found in research performed in specialized health services. Conclusion: We found good feasibility (acceptability and applicability), and validity measures comparable to those found from specialized health services. Further investigations of INTERMED predictive validity and suitability for routine PHC use are worthwhile.