Cluster Analysis of Outpatient Prescription Medications Following Combat-Related Amputations: A Retrospective Study.

Abstract

Objectives: The objectives are: 1) to conduct cluster analysis to explore patterns of outpatient prescription medications during the first year following combat-related amputations, and 2) to analyze patient characteristics associated with prescription medication clusters.

Design: Retrospective study of Department of Defense casualty records and outpatient prescription medications during the first year postinjury. Hierarchical cluster analysis of American Hospital Formulary Service (AHFS) medication classes.

Setting: Military amputation rehabilitation program.

Participants: 1,651 US service members who sustained major limb amputations in Operations Iraqi and Enduring Freedom (OIF/OEF), 2001-2017, and had outpatient prescription medications.

Interventions: n/a.

Main outcome measures: Prescription medication clusters, patient characteristics.

Results: Cluster analysis identified three prescription medication profiles: Cluster 1) Baseline (n = 682) consisted of six AHFS classes: cathartic/laxatives, anticonvulsants, semisynthetic opioids, antidepressants, NSAIDs, combination opioid analgesics. These six classes were part of all three profiles (i.e., baseline medications). Cluster 2) Baseline plus Insomnia (B+I) (n = 502): atypical antipsychotics, anxiolytics/sedatives, methadone, benzodiazepines, analgesics/antipyretics and baseline medications. Cluster 3) Baseline plus Severe Injury (B+SI) (n = 467): anticoagulants, 5-HT 3 receptor antagonist, multivitamins, anxiolytic/sedatives, analgesic/antipyretics and baseline medications. The Baseline cluster was characterized by early injury years (85%, 2001-2008), moderate/serious injuries (85%), single amputation, chronic pain, and posttraumatic stress disorder. Cluster B+I was characterized by early injury years (62%), minor injuries (16%), low rates of traumatic amputations, and postinjury complications. Cluster B+SI was characterized by later injury years (88%, 2009-2017), serious/severe injuries (62%), multiple amputation (42%), TBI (43%), high rates of complications, and prescriptions.

Conclusions: Cluster analysis identified three outpatient prescription medication profiles during the first year postinjury. All profiles included analgesics and psychotherapeutics but differed on specific insomnia medications, multivitamins, and anticoagulants. Medication profiles had significant associations with injury years, injury severity, and postinjury complications. The results inform future research and prescription medication practices.