ABSTRACT
BACKGROUND: Narrow definitions of long-term opioid (LTO) use result in limited knowledge of the full range of LTO prescribing patterns and the rates of these patterns. OBJECTIVE: To investigate a model of new LTO prescribing typologies using latent class analysis. DESIGN: National administrative data from the VA Corporate Data Warehouse were accessed using the VA Informatics and Computing Infrastructure. Characterization of the typology of initial LTO prescribing was explored using latent class analysis. PARTICIPANTS: Veterans initiating LTO during 2016 through the Veteran's Administration Healthcare System (N = 42,230). MAIN MEASURES: Opioid receipt as determined by VA prescription data, using the cabinet supply methodology. KEY RESULTS: Over one-quarter (27.7%) of the sample fell into the fragmented new long-term prescribing category, 39.8% were characterized by uniform daily new LTO, and the remaining 32.7% were characterized by uniform episodic LTO. Each of these three broad sub-groups also included two additional sub-groups (6 classes total in the model), characterized by the presence or absence of prior opioid prescriptions. CONCLUSIONS: New LTO prescribing in the VA includes uniform daily prescribing, uniform episodic prescribing, and fragmented prescribing. Future work is needed to elucidate the safety and efficacy of these prescribing patterns.
