Changes in opioid and nonopioid prescribing trends do not appear to be associated with concurrent changes in pain intensity reported by older veterans with osteoarthritis (OA), according to a study published in Pain.

In this retrospective study, investigators employed an interrupted time-series, segmented regression model of aggregated monthly national data that was available from the Veterans Health Administration (VHA) corporate data warehouse to access the national repository of patient-level medical records. Data were part of an observational study that evaluated the safety and effectiveness of analgesics in veterans aged ≥50 years with arthritis of the knee or hip (n = 348,787) and included analgesic prescription name and pain intensity ratings. These data were aggregated monthly over the 5-year study period.

Related Articles

There was an increasing trend prior to the Opioid Safety Initiative (OSI) in all study analgesic prescriptions (β for time = 0.012; P =.003). After initiation of OSI, a significant step (0.319; P =.0001) and a pronounced negative change in slope (Δβ = -0.010; P =.0001) were revealed. The slope post-OSI was not significantly different from a flat trajectory. Prior to OSI, there was an increase in opioid prescribing (β for time = 0.018; P <.0001), as well as an increase in prescribing at the time of OSI (step change = 0.578; P <.0001). Prescription of nonsteroidal anti-inflammatory drugs remained unchanged and other study analgesics increased modestly over the study period. The slope of pain intensity in patients reporting pain was stable over the study period (β for time = 0.003; P =.613), and no changes related to OSI were revealed (step change = 0.015; P =.963). The percentage of patients reporting pain gradually increased through the study period (β for time = 0.089; P =.038), and there was no step changes (0.926; P =.271) or slope changes (Δβ = -0.029) associated with OSI. The effect size calculation was nonsignificant (-1.9%; P =.096). There was a significant stead upward trend in patients reporting moderate pain (β for time = 0.046; P =.047). There was no measure demonstrating any significant change associated with OSI.

This study is limited by the inability to reflect causality, as this was an observational study. The findings from VHA data may not be generalizable for other healthcare settings and patient populations.

Clinicians should be informed about the risks and benefits of nonopioid analgesic medications. The effect of changing rates of opioid and nonopioid analgesic medication prescribing on patient safety and pain intensity outcomes requires further research.

Follow @ClinicalPainAdv


Trentalange M, Runels T, Bean A, et al. Analgesic prescribing trends in a national sample of older veterans with osteoarthritis: 2012-2017. [published online January 29, 2019]. Pain. doi: 10.1097/j.pain.0000000000001509

This article originally appeared on Rheumatology Advisor