Title

Potentially inappropriate medications do not clinically impact trauma recidivism in the elderly

Document Type

Abstract

Publication Date

10-2019

Abstract

Study Objective

The study evaluated the association between potentially inappropriate medications (PIM) as defined by Beers’ Criteria and the risk of trauma recidivism in the elderly. The American Geriatric Society (AGS) modified the list of PIMs in 2012 to include 19 categories of medications with limited effectiveness in the elderly but with a positive association with complications such as delirium, bleeding, fracture and falls, a major cause of elderly trauma. We hypothesized that the presence of PIM increases the risk of trauma recidivism.

Methods

This case-control study reviewed patients over age 65 evaluated at a single Level 1 Trauma center from February 1, 2013 through August 1, 2018. We performed a 1:1 propensity score matched analysis using age and sex between injured patients with one trauma evaluation at any level of trauma activation and those with more than one trauma evaluation during the study period. We used a receiver operating characteristic (ROC) curve to determine the number of PIM with a maximum predictive value for recidivism. Area under the curve (AUC) with 95% confidence intervals (CI) were calculated. We also evaluated if there were specific categories of PIM that were associated with recidivism using Chi square analysis. A p value of <0.05 was deemed significant.

Results

During the study period, there were 11,089 trauma evaluations meeting study criteria. 8252 patients had one trauma evaluation only and 2837 patients had more than one trauma evaluation for a 26.2% recidivism rate. Propensity matching resulted in 2324 patients in each group. Patients with one trauma evaluation and recidivists had similar mean number of total medications (12.34, standard deviation (SD) 6.71 vs 12.67, SD 6.83, p = 0.09) but those with one trauma evaluation had a lower mean number of PIM (11.09, SD 6.74 vs 11.88, SD 6.99, p =0.001). ROC curve analysis showed that the maximum predictive threshold was 11 PIM (AUC = 0.545, 95% CI=0.529-0.560). In addition, analgesics, antidepressants, antiplatelets, antipsychotics, anxiolytics, central nervous system medication and stimulants were each significantly associated with recidivism (p <0.001 for all categories). Recidivists were on a greater mean number of medications in these seven categories than non-recidivists (6.45, SD 3.034 vs 6.15, SD 3.038, p = 0.001)

Conclusions

PIM are statistically associated with trauma recidivism in the elderly. However, of the 19 categories of PIM identified by the AGS, only seven were associated with recidivism. Moreover, as the mean number of PIMs differed by less than one between the two groups, the clinical significance of our findings is doubtful. We conclude that PIMs do not clinically influence trauma recidivism in the elderly.

Publication Title

Annals of Emergency Medicine

Volume

74

Issue

4S

Open Access

Share

COinS