Autopsy and premortem diagnostic discrepancy review in an Irish tertiary PICU.
Abstract Objective: 1) Review ante- and post-mortem diagnoses and assign a Goldman error classification. 2) Establish autopsy rates. Design: A retrospective analysis of autopsies performed on patients who died in Paediatric intensive care unit (PICU) between November 13th 2012 and October 31st 2018. We reviewed medical and autopsy data of all patients and Goldman classification of discrepancy between ante- and post-mortem diagnoses was assigned. Setting: Tertiary PICU. Patients: All patients that died in PICU within the designated timeframe. Interventions: Goldman error classification assignment. Measurements and main results: 396 deaths occurred in PICU from 8,329 (4.75%) admissions. 99 (25%) had an autopsy, 75 required by the coroner. All were included in the study. Fifty-three were male and 46 females. Fifty-three patients were transfers from external hospitals, 46 from our centre. Forty-one were neonates, 32 were <1 year of age, and 26 were >1 year of age. Median length of stay was 3 days. Eighteen were post cardiac surgery, and three post cardiac catheter procedure. Major diagnostic errors (Class I/II) were identified in 14 (14.1%), 2 (2%) Class I, and 12 (12.1%) were Class II errors. Class III and IV errors occurred in 28 (28.2%) patients. Complete concordance (Class V) occurred in 57 (57.5%) cases. Conclusion: The autopsy rate and the diagnostic discrepancy rate within our PICU is comparable to those previously reported. Our findings show the continuing value of autopsy in determining cause of death and providing greater diagnostic clarity. Given their value, post-mortem examinations, where indicated, should be considered part of a physician’s duty of care to families and future patients.