RESUMO
BACKGROUND: Choking incidents have long been recognized as a cause of death in people with mental illness. This study attempted to determine the rate of choking incidents and deaths among psychiatric inpatients and identify the risk factors associated with choking incidents. Some preventive measures are also suggested. METHODS: All choking incidents recorded over 3 years in 3 psychiatric wards of Chutung Veterans General Hospital (total of 210 beds) were retrospectively analyzed for demographic variables, psychiatric and medical diagnoses, and drug therapy at the time of the incidents. RESULTS: During the 3-year period, 16 incidents were recorded, involving 11 patients at a rate of 1 incident every 451.4 months of hospitalization per person. Men were 3 times more likely to experience choking incidents than women, and the mean age of choking patients was higher than that of all patients (59.7 vs. 44.4 years). Among the 16 incidents, 3 fatal cases were noted, giving an annual choking death rate of 5.05 per 1,000 persons hospitalized. A high re-choking rate of up to 40.0% was noted, and patients with organic mental disorder had 3.4 times the choking incidence of all patients. The mean antipsychotic dosage of the choking cases was similar to that of all inpatients, but the mean hypnotic dosage of the fatal choking cases was significantly higher. Up to 62.5% of the choking accidents occurred at breakfast, and mantou (a type of steamed bun) was the most common food (9/16) that resulted in the accidents. Of the 11 patients, 9 were rated as "poor" or "very poor" with regard to self-care before the accidents. CONCLUSION: Compared with studies in Western psychiatric inpatients, we have a lower choking incident rate but higher death rate due to choking. We found that older age, male sex, higher dosage of hypnotics, previous choking attacks, organic mental disorder, poor self-care, breakfast time and mantou were possible risk factors associated with choking, in which older age, poor self-care, and higher dosage of hypnotics for fatal cases were statistically significant.
