J Clin Epidemiol Vol. 49, No. 8, pp. 865-868, 1996 Copyright 0 1996 Elsevier Science Inc. ELSEVIER 0895,4356/96/$15.00 PII SO895-4356(96)00114-X Out-of-Hospital Sudden Death Referring to an Emergency Department Roberto Manfredini ,‘I* Francesco Poduppi,’ Enrico Grandi, Cumelo Fersini,’ and Musk-no Galhani3 'FIRST INSTITUTE OF INTERNAL MEDICINE, UNIVERSITY OF FERRARA, I-44100 FERRARA, ITALY, 'INSTITUTE OF PATHOLOGICAL ANATOMY, HISTOLOGY, AND CYTOLOGY, UNIVERSITY OF FERRARA, I-44100 FERF.ARA,ITALY, AND 'EMERGENCY DEPARTMENT,ST. ANNA HOSPITAL OF FERRARA, I-44100 FERRARA,ITALY ABSTRACT. Over an 11-year period, autopsies were performed on 957 of 1038 nontraumatic deaths in the Emergency Department of the Central Hospital in Ferrara, Italy. Of these 957 cases, 732 (76.5%) met criteria for sudden death. In 100 (14%) of these cases, the death could be attributed to pulmonary embolism (55 cases), stroke (17), or rupture of aortic aneurysm (28). Acute myocardial infarction accounted for 403 (55%) of all sudden deaths. Severe coronary artery disease was found in 340 (84%) of these 403 deaths, with plaque fissuring or thrombi in 189 or 151 cases, respec- tively. Among the 229 sudden deaths for whom no immediate cause could be determined (3 1% of the total popula- tion), all had evidence of heart disease: 147 individuals had severe coronary artery disease, with plaque fissure ing or thrombi found in 72 or 43, respectively. The remaining cases with no immediate cause of death had evidence of a cardiomyopathy (61) or valvular disease (2 1). We conclude that acute myocardiil infarction accounts for the majority of cases of nontraumatic sudden death in our Emergency Department. Altogether, 84% of these patients had severe coronary artery disease. In approximately one-third of cases for whom no immediate cause of sudden death could be determined, all had evidence of heart disease, and about two-thirds had severe coronary artery disease. J CLIN EPIDEMIOL 49; 8:865-868, 1996. KEY WORDS. Sudden death, emergency department, autopsy, myocardial infarction, coronary artery disease INTRODUCTION The underlying etiopathological factors of sudden death (SD) are not fully known. The very unexpectedness and suddeness of the fatal event often make it very difficult, and sometimes impossible, to ob- tain exhaustive clinical histories of significant ECG records. Epidemiological studies in which the data were gathered only from death certificates often assumed that SD is mostly due to isch- emit heart disease [I]. However, when the causes of death reported on death certificates are compared with those from autopsy diagnosis in hospital patients, the results may be different [2,3]. Hence, the incidence of coronary artery lesions in patients who experienced SD is still under debate in the literature, owing to the great methodolog- ical differences between studies, e.g., use of death certificates or nec- ropsy; means of sampling the population, geographical, and ethnic differences; and prevalence of risk factors. We report here the results of a study carried out in the emergency department (ED) of a general hospital. It was a prospective study several years in duration that focused on a large population of pa- tients arriving at the ED and undergoing necropsy. SUBJECTS AND METHODS The study ran from 1 January 1983 to 31 December 1993 and was conducted in the ED of Ferrara General Hospital, Ferrara, Italy, a *Address for correspondence: Roberto Manfredini, First Institute of Internal Medicine, University of Ferrara, via Savonarola 9, I-44100 Ferrara, Italy. Accepted for publication on 29 January 1996. 1200-bed hospital with only one transport service for all out-of- hospital emergencies. Moreover, this hospital also serves as the sole teaching center for the University of Ferrara School of Medicine. Ferrara is unusually well suited to epidemiological studies: the popu- lation, in fact, is stable and well defined (approximately 150,000), and the ED has approximately 57,000 annual patient visits. Owing to the organization of the emergency service, the relatively restricted cachement area and its geographic features, the average time of ar- rival at the ED from anywhere in the city is around 15 min from the moment the call is received. This includes the time required to get to the patient, perform resuscitation if necessary, and transport him or her back to the ED. No subject who dies at home is brought to the hospital, unless spontaneous circulation is reinstated by resus- citation maneuvers. Thus, patients in the prehospital setting who did not have return of spontaneous circulation following cardiac arrest were not included in our series. The cases for this study were selected according to the follow- ing inclusion criteria: subjects undergoing SD, according to the Goldstein definition of “death occurring unexpectedly, within one hour from the onset of symptoms, in ordinary conditions of life, and in the presence of witnesses” [4]. The causes of SD included irrevers- ible cardiocirculatory collapse, cardiac arrest, and acute pulmonary edema. It must be stressed that, in applying the Goldstein definition in the prehospital setting, time elapsed is calculated from the first out-of-hospital event. Exclusion criteria were as follows: death occurred in the absence of witnesses or without a precise knowledge of the hour of the event;