LSU: Faculty and Researchers Used Emergency Department Data to Estimate Prevalence of Smoking in Young Adults

Faculty and researchers at the LSU School of Medicine (Drs. Stephen Kantrow and Sarah Jolley), Louisiana State University Health Sciences Center School of Public Health (Ms. Xinnan Wang, Dr. Tung-Sung Tseng, Dr. Dodie Arnold, Dr. Claudia Leonardi, Dr. Richard Scribner, Dr. Edward Trapido, Dr. Hui-Yi Lin), Ochsner Health System (Eboni Price Haywood) and the Louisiana Public Health Institute (Dr. Lisanne Brown) used emergency department (ED) data to estimate the prevalence of smoking in young adults. Most state or national surveys of smoking are limited in size- especially when looking at county level data, and phone based surveys have had decreasing responses. School based surveys are helpful, but limited by age.

The researchers used data from electronic health records (EHRs) for five EDs within U.S. Census-defined metropolitan New Orleans (New Orleans–Metairie, LA) for persons 18-24 years old. (15 percent of young adults are reported to have used EDs in the past year). Smoking status was available for 55,777 persons (91 percent of the total Emergency Departments); 61 percent were women, 55 percent were black, 35 percent were white, and 8 percent were Hispanic. One third of patients were uninsured. Most smokers used cigarettes (95 percent ). Prevalence of current smoking was 21.7 percent for women and 42.5 percent for men. Smoking prevalence was highest for substance use disorder (58 percent ), psychiatric illness (41 percent ) and alcohol use (39 percent ), and lowest for pregnancy (13.5 percent ). In multivariable analyses, male gender, white race, lack of health insurance, alcohol use, and illicit drug use were independently associated with smoking. Smoking risk among alcohol and drug users varied by gender, race, and/or age.

The BRFSS estimated 29 percent prevalence during the same time, and had data on 597 subjects aged 18–30 years. Although ED data are likely to overestimate tobacco use, the large sample size is useful- especially for stratum-specific estimates- particularly in a demographically diverse population. Dr. Stephen Kantrow, the lead investigator, stated ”this approach provides smoking data for a large sample of young adults in one metropolitan area, and may support longitudinal studies of smoking in high and low risk populations.

Full article