Assessing the validity of North York dental diagnostic codes

We set out to assess the consistency of diagnostic codes in the North York Public Health Department (NYPHD) database with dental charts, report prevalence of the conditions, assess providers compliance with writing the diagnosis, and assess the predictive power of the diagnostic codes. The NYPHD database and a sample of 220 dental charts were compared to assess consistency of the diagnostic codes. Procedure and diagnostic codes were compared to assess providers! compliance with writing the diagnosis. Bivariate odds ratios and logistic regression analyses were used to assess the strength of the diagnoses in predicting procedures. Ninety-one percent of the diagnoses from the selected charts matched those in the database. Carlos in dentin was the most common diagnoses. Clinicians recorded diagnostic codes for almost 90% of the cases. Logistic regression showed highly significant association between diagnoses and relevant procedures as indicated by program guidelines. The results of the study indicate that the database was consistent with the charts, and the diagnoses predicted the care as indicated by the program guidelines.