International classification for primary care (ICPC) coding by family medicine trainees

Abstract

Objectives: To determine the ability of Portuguese doctors in the second half of their family medicine residency to code with the International Classification of Primary Care - revised version (ICPC-2). Study design: Cross-sectional. Setting: South of mainland Portugal, Azores e Madeira. Participants: Residents in the 29th and 30th courses of the Family Medicine Residency Coordination of South Portugal. Methods:A convenience sample of 100 residents was contacted by email, and asked to code three clinical vignettes written by the authors. The correct coding of reasons for encounter, diagnoses, and processes was determined by the authors with the help of a national ICPC-2 expert. We made a descriptive analysis of the proportion of correct answers, failures of classification, and most frequent errors. The answers of residents with and without previous training in ICPC-2 usage were compared. Results: The participation rate was 47.0%. 56.5% of reasons for encounter, 75.9% of diagnoses, and 48.8% of processes were correctly classified. Residents frequently coded in excess parts of the visit that are not, by the ICPC-2 definition, reasons for encounter; they frequently left out the code health maintenance/preventive medicine from problems and reasons for encounter; and, in processes, they seldom code the complete and partial medical examination. Residents who had previous training in ICPC-2 usage correctly coded more reasons for encounter (72.3% vs 55.1%) and problems (85.0% vs 74.8%). Conclusion: Residents appear to be better at coding diagnosis than reasons for encounter or processes. Errors in the coding of preventive activities and medical examination are frequent. Residents who had training in ICPC-2 usage had better results in coding of reasons for encounter and problems.