Tiffany Feng, Eustina Kim, Diana Pham, Kienna Qin
UCLA
Our team explored which characteristics of each occupation (students, health professionals, and veterans) determined the use of benzodiazepines (BZDs), a type of drug that produces sedation and lowers anxiety levels, in the sample of participants in the United States during 2018.
BZDs are the most commonly prescribed medication in the U.S., and, unfortunately, are frequently abused and taken with other drugs. For our project, we created three multinomial logistic regression models to predict use/misuse of BZDs for each occupation.
Which variables contribute the most to accuracy? The plots below show the 30 most important variables.
NMU: Non-medical use
Having a history of drug use, mental illness, and prescriptions for opioids for treating pain are important in predicting the use of BZD for all three occupations.
The largest impacts for each occupation were:
Students: history of drug use
Veterans: mental illness and opioid prescriptions for pain
Health professionals: similar influences
Our data was not a representative sample and cannot be generalized to the larger population.