Latent Classes of Brief Alcohol Intervention Recipients in Two Level-1 Trauma Centers
At-risk drinking is the primary predictor for injury in the US. Brief intervention has been developed to help reduce drinking and prevent alcohol-related injury risk behaviors. It is not clear, however, from the literature which patients experience the greatest improvements in drinking following brief intervention and hospital discharge. Understanding which patient populations are most likely to improve will aid social workers and other behavioral health care providers tailor intervention strategies to meet the needs of at-risk drinkers.
Methods
Data from two similar screening and brief intervention clinical trials (Study 1: N=737; Study 2: N=250) were used in this secondary analysis to characterize intervention recipients and to assess who experienced the largest improvements in alcohol use following discharge from a Level-1 trauma center. Latent class analysis was used to identify patient profiles based on alcohol-related risk behaviors using the Short Inventory of Problems+6. Demographic characteristics, history of injury care, and participant beliefs regarding the relationship of their alcohol use to their current injury were added as covariates to the models. Percent days abstinent and average weekly volume consumed were analyzed to examine improvements in drinking at six and 12 months.
Results
A five-class solution emerged from the data from Study 1 (AIC: 5042.20; ABIC: 5063.61; BLRT: p=0.00), and a four-class solution emerged from Study 2 (AIC: 2021.48; ABIC: 2032.73; BLRT: p=0.03). Four of the five classes in Study 1 possessed markedly similar conditional item probabilities to the four classes from Study 2. The four similar sets of classes were labeled: (1) multiple risks, (2) fighting and foolish risks, (3) accidents and injury, and (4) minimal risks. The largest increases in both studies for days abstinent were reported by the multiple risks class (Study 1: 6 months ηp2=0.15, 12 months ηp2=0.18; Study 2: 6 months ηp2=0.73, 12 months ηp2=0.65; p<0.05) and the minimal risks class (Study 1: 6 months ηp2=0.06; Study 2: 6 months ηp2=0.58, 12 months ηp2=0.41; p<0.05). Significant decreases in volume consumed for both studies were reported by the multiple risks class (Study 1: 12 months: ηp2=0.07; Study 2: 6 months ηp2=0.44, 12 months ηp2=0.48; p<0.05). In Study 2, large decreases in volume were also reported by the minimal risks (6 months ηp2=0.49, 12 months ηp2=0.29; p<0.05) and accidents and injury (6 months ηp2=0.52, 12 months ηp2=0.37; p<0.05) classes. The fighting and foolish risk class reported limited improvements.
Conclusions and Implications
Patients with histories of multiple alcohol-related risk behaviors and minimal risk appear to report the strongest positive changes following brief intervention and discharge from a Level-1 trauma center. Patients with histories of alcohol-related fighting seem to improve less. Future research will examine how to translate these findings into practice. With the ability to classify injured patients, social workers and other behavioral health professions will be able to build on positive prognoses for drinking improvements or adapt interventions to better serve those likely to improve less.