The effect of lifestyle choices on emergency department use in Australia

Much attention has been paid to patient access to emergency services, focusing on hospital reforms, yet very little is known about the characteristics of those presenting to emergency departments.

By exploiting linkage of emergency records and a representative survey of the 45 and older population in Australia, we provide unique insights into the role of lifestyle in predicting emergency presentations.

A generalized linear regression model is used to estimate the impact of lifestyles on emergency presentations one year ahead. We control for extensive individual characteristics and area fixed-effects.

Not smoking, having healthy body weight, taking vitamins, and exercising vigorously and regularly can reduce emergency presentations and also prevent subsequent admissions from emergency. There is no evidence that heavy drinking leads to more frequent emergency visits, but we find a high tendency for heavy drinkers to smoke and be in poor health, which are both major predictors of emergency visits.

Targeted public health interventions on smoking, body mass and exercise may reduce emergency visits. Effective public health interventions which target body mass, exercise, current smoking and smoking initiation, may have the effect of reducing ED usage and subsequent admission. Individual-level data linking a survey of the population 45 and older in Australia with their emergency department (ED) records is exploited to provide unique insights into the role of lifestyle in predicting emergency care. Controlling for demographic and socioeconomic characteristics, as well as chronic conditions, we find that being a non-smoker, having a healthy body weight, taking vitamins, and doing a vigorous exercise at least once a week can prevent ED presentations. Being a non-smoker, taking vitamins and exercising also prevent subsequent admissions from ED. We do not find a similar protective effect from complying with dietary recommendations. There is no evidence that heavy drinking alone leads to more frequent ED visits, but we find a high tendency for heavy drinkers to smoke and be in poor health, which are both major predictors of ED visits. These results suggest that targeted public health interventions on smoking, body mass and exercise can reduce ED visits. The use of linked data provides important insight into the characteristics of potential ED users which in turn is valuable for the planning of health services.