Assessing the accuracy of hospital admission and discharge diagnosis of traumatic brain injury in a New Zealand hospital

AIMS: To investigate the accuracy of admission and discharge coding of traumatic brain injury (TBI) in a New Zealand hospital. METHOD: Prospective study of all patients over fifteen years of age admitted to Hutt Hospital over a six-month period with an actual or potential diagnosis suggesting TBI. RESULTS: During the six month period of study, 65 patients with the diagnosis of TBI were admitted to Hutt Hospital. Of these, 21 (32.3%) met the criteria for diagnosis of TBI ('Definite TBI'). A further eighteen patients, not admitted with a diagnosis of TBI, met the TBI criteria. Only 14/39 (35.9%) of 'Definite TBI' cases were identified at both admission and discharge. Discharge diagnosis of TBI identified correctly 26/39 (66.7%) of definite cases, with 34/60 (56.7%) cases with a discharge diagnosis of TBI not meeting our criteria for the diagnosis of TBI. Six out of 39 'Definite TBI' cases (15.4%) were not identified by either admission or discharge diagnosis. Thirty of the 65 patients (46.2%) admitted to hospital with the diagnosis of TBI showed clinical evidence of having taken alcohol, although only 12 had blood alcohol concentration measured. CONCLUSIONS: The admission and discharge diagnoses of TBI were not accurate when compared to a standard definition of TBI. For hospital discharge data to have any value, agreement on an operational diagnosis of TBI needs to be made, which should include measurement of the blood alcohol concentration. A suggestion for a diagnostic strategy is presented, along with ICD-10-AM codes that could be used to improve the current situation.