OBJECTIVES: Making an accurate clinical diagnosis in the field can be a great challenge with pediatric out-of-hospital cardiac arrest (OHCA). We aimed to compare the etiology of pediatric OHCA by pre-hospital clinical diagnosis with etiology by coroner's diagnosis and autopsy. DESIGN: As part of the Ontario Pre-hospital Advanced Life Support (OPALS) study, we conducted a prospective cohort study including children below age 19 with OHCA during an 11-year period. Prehospital clinical diagnosis was determined by blinded review and deaths were then matched with provincial coroner's office records. The agreement between prehospital clinical diagnosis and autopsy diagnosis was derived by consensus review. Inter-observer agreement was evaluated using kappa values. RESULTS: For the period 1992-2002, there were 414 cardiac arrests in children <19 years of age that matched coroner's records. Mean age was 5.9 years (S.D. 6.4 years) with 39.4% of cases under 1 year of age. Etiology by clinical diagnosis was medical 49.5%, trauma 36.0% and undetermined 14.5%. The overall kappa for clinical diagnosis compared to coroner's diagnosis was 0.62. The kappa for medical cases was 0.53, trauma was 0.93 and 'undetermined' was -0.01. Medical clinical diagnosis had a lower agreement with the coroner's diagnosis (62.4%) compared with trauma (96.0%), RR 0.65, 95% CI [0.58, 0.73]. The poorest kappas by diagnosis were for neurological (0.39), respiratory (0.42), 'other' medical (0.56), SIDS (0.58) and cardiac (0.63). The commonest coroner's diagnoses in the 'undetermined' clinical diagnosis category were: pneumonia (17.6%), seizure or post-seizure (11.8%), arrhythmia (9.8%) and aspiration (5.9%). CONCLUSION: Even in an ideal situation, a clinician in the field might be unable to determine the etiology of pediatric cardiac arrest in 14.5% of cases. There is poorer agreement for 'medical' compared to 'trauma' cases. This is the largest study to date comparing clinical diagnosis of the causes of OHCA in children to the 'gold-standard' of coroner's diagnosis.
OBJECTIVES: Making an accurate clinical diagnosis in the field can be a great challenge with pediatric out-of-hospital cardiac arrest (OHCA). We aimed to compare the etiology of pediatric OHCA by pre-hospital clinical diagnosis with etiology by coroner's diagnosis and autopsy. DESIGN: As part of the Ontario Pre-hospital Advanced Life Support (OPALS) study, we conducted a prospective cohort study including children below age 19 with OHCA during an 11-year period. Prehospital clinical diagnosis was determined by blinded review and deaths were then matched with provincial coroner's office records. The agreement between prehospital clinical diagnosis and autopsy diagnosis was derived by consensus review. Inter-observer agreement was evaluated using kappa values. RESULTS: For the period 1992-2002, there were 414 cardiac arrests in children <19 years of age that matched coroner's records. Mean age was 5.9 years (S.D. 6.4 years) with 39.4% of cases under 1 year of age. Etiology by clinical diagnosis was medical 49.5%, trauma 36.0% and undetermined 14.5%. The overall kappa for clinical diagnosis compared to coroner's diagnosis was 0.62. The kappa for medical cases was 0.53, trauma was 0.93 and 'undetermined' was -0.01. Medical clinical diagnosis had a lower agreement with the coroner's diagnosis (62.4%) compared with trauma (96.0%), RR 0.65, 95% CI [0.58, 0.73]. The poorest kappas by diagnosis were for neurological (0.39), respiratory (0.42), 'other' medical (0.56), SIDS (0.58) and cardiac (0.63). The commonest coroner's diagnoses in the 'undetermined' clinical diagnosis category were: pneumonia (17.6%), seizure or post-seizure (11.8%), arrhythmia (9.8%) and aspiration (5.9%). CONCLUSION: Even in an ideal situation, a clinician in the field might be unable to determine the etiology of pediatric cardiac arrest in 14.5% of cases. There is poorer agreement for 'medical' compared to 'trauma' cases. This is the largest study to date comparing clinical diagnosis of the causes of OHCA in children to the 'gold-standard' of coroner's diagnosis.
Authors: Dianne L Atkins; Siobhan Everson-Stewart; Gena K Sears; Mohamud Daya; Martin H Osmond; Craig R Warden; Robert A Berg Journal: Circulation Date: 2009-03-09 Impact factor: 29.690