Validating injury-related diagnoses by physicians: an analysis of 62,269 hospitalizations from a large hospital in Changsha, China.

OBJECTIVE: To assess the accuracy of injury-related hospitalization diagnoses by physicians in China.

METHODS: 62,269 hospitalizations between 2014 and 2016 at a large hospital in Changsha, China were assessed. We considered three types of diagnostic errors: under-reporting, lack of cause specificity, and misclassification across injury causes. Diagnosis and coding of diseases were ascertained by professional coders based on the tenth International Classification of Disease (ICD-10). Chi square tests examined the proportion difference across subgroups of patient demographics, injury intent, and injury cause.

RESULTS: The 62,269 records included 1011 injury-induced hospitalizations, but physicians' diagnoses only reported 405 (40%) of the 1011 hospitalizations as injury-related. The proportions of under-reporting, lack of cause specificity, and misclassification errors, respectively, were 50.5%, 7.9% and 1.5%. The proportion of diagnostic errors was relatively similar across patient sex, intent of injury, and cause of injury subgroups, but varied substantially across patient age groups.

CONCLUSION: Physicians' diagnoses frequently omitted injury-induced hospitalizations. Such errors will impact injury research and policy-making when they are undetected and uncorrected.