Which barriers affect morbidity registration performance of GP trainees and trainers?

Jip de Jong, Mechteld R. M. Visser, Margreet Wieringa-de Waard

Research output: Contribution to journalArticleAcademicpeer-review

4 Citations (Scopus)

Abstract

Diagnosis coding percentages in the specialty training of general practitioners (GPs) are generally high, but not perfect, indicating barriers against coding still exist, possibly influencing the validity of data based on electronic patient records (EPRs). To study the relationship between barriers to coding diagnoses with the International Classification of Primary Care (ICPC) of GP trainees and trainers and their self-reported and actual coding performance. A questionnaire was developed, and returned by 71 (of 73, 97%) GP trainees and 103 (of 108, 95%) GP trainers, affiliated to the GP Specialty Training of the Academic Medical Center, University of Amsterdam. Their barriers to ICPC coding and self-reported coding performance were compared with EPR-derived data extractions that were collected during one year. Mean coding percentages were 88.3 (SD=11.5) and 82.3% (SD=19.0) (trainees/trainers). Most participants reported always registering ICPC codes for consultations and home visits, specifically in those situations pre-specified in the questionnaire. Telephone consultations, repeat prescriptions and administrative actions were coded less frequently. Most participants never or rarely experienced coding barriers, an exception being 'insufficient refinement of the ICPC system'. Most motivation and ICPC-related barriers correlated with self-reported and actual coding performance. Regression analyses showed that 'ICPC coding is unpleasant to use' predicted both trainees' and trainers' coding percentages. The trainers' coding percentage was also predicted by 'no personal gain from ICPC' and 'coding is difficult'. The mean coding percentages we found were high, but could further be improved by increasing GPs' motivation and by making ICPC coding more user-friendly. EPR-derived data seem biased by non-coded telephone consultations only
Original languageEnglish
Pages (from-to)708-716
JournalInternational Journal of Medical Informatics
Volume82
Issue number8
DOIs
Publication statusPublished - 2013

Cite this