Objectives: (1) To examine the variability in diagnosis of mild traumatic brain injury (mTBI) in primary care relative to that of an expert reviewer; and (2) to determine the incidence rate of mTBI in Ontario, Canada.
Method: Potential mTBI cases were identified through reviewing three months of Emergency Department (ED) and Family Physician (FP) health records. Potential cases were selected from ED records using the International Classification of Disease, 9th revision, Clinical Modification and External Cause codes and from all FPs records for the time period. Documented diagnoses of mTBI were compared to expert reviewer diagnosis. Incidence of mTBI was determined using the documented diagnoses and data from hospital catchment areas and population census.
Results: 876 potential mTBI cases were identified, 25 from FP records. Key indicators of mTBI were missing on many records (e.g., 308/876 records had Glasgow Coma Scale (GCS) scores). The expert reviewer disagreed with the documented diagnosis in 380/876 cases (kappa = 0.19). The expert reviewer was more likely to give a diagnosis if the GCS was 13-14, if there was documented loss of consciousness and/or post-traumatic amnesia, and/or if there was pathology found on an acute brain scan. Calculated incidence rates of hospital-treated mTBI were 426 or 535/100,000 (expert review--hospital diagnosis). Including family physician cases increased the rate to 493 or 653/100,000.
Conclusion: Health record documentation of key indicators for mTBI is often lacking. Notwithstanding, some patients with mTBI appear to be missed or misdiagnosed by primary care physicians. A more comprehensive case definition resulted in estimated incidence rates higher than previous reports.