29 Assessing point-prevalence and documentation of head injury in the north London forensic service

Abstract

Objectives/Aims Traumatic brain injury (TBI) is highly prevalent in prison populations, but little is known about TBI prevalence amongst mentally disordered offenders in secure settings. The North London Forensic Service (NLFS) is a tertiary forensic secure psychiatric service, serving a catchment population of over 1.4 million. We aimed to estimate the point-prevalence of head injury amongst inpatients at NLFS, and audit the quality of documentation of head injury. Method We performed a retrospective case note review of the NLFS electronic records system to identify all documented head injuries in all NLFS inpatients on a selected day. To increase detection, we reviewed neuroradiology reports, and medical records were electronically screened for the following text-strings: ‘brain injury’, ‘TBI’, ‘head injury’, ‘road traffic accident’, ‘loss of consciousness’, ‘LOC’, ‘concussion’, ‘MRI’, ‘GCS’ and ‘boxing’. For each injury identified, we assessed whether amnesia and altered consciousness at the time of the injury were documented, in line with established criteria for assessing severity of TBI (Department of Veterans Affairs/Department of Defence criteria). Injury mechanism and date, patient age, and admission length were recorded. Authors AJB and RM co-rated the findings, and disagreements were resolved by consulting a third party. Results 100% of inpatient records were screened (n=194, 6% female, 94% male, mean age 40.2 years). 58 documented head injuries were identified, involving 50 patients (25.8%). 43 head injuries occurred prior to admission to NLFS. 16 patients (8.2%) had head injuries associated with clearly documented impairment of consciousness and/or amnesia, suggestive of TBI. 13.7% of all recorded head injuries included explicit documentation of both consciousness level and amnesia (33% when limited to head injuries occurring during admission to NLFS). The most common identified mechanism was assault (n=30). Mean admission length was 1306 days, and was not significantly different in patients with a history of head injury (p=0.825, Mann-Whitney U ). 34% of inpatients had undergone neuroimaging. Abnormalities were identified in 32% of reports, with leukoaraiosis and generalised cerebral involutional changes being most frequently reported. Conclusion These findings highlight a local need to improve quality of documentation in assessment of patients with head injuries, and provide an estimate of point-prevalence of head injury and TBI at NLFS. The high frequency of assault observed in our sample is comparable to previously published findings in prison populations.