Hospital initiation of opioids and long-term prescribing among older adults in primary care - a cohort study

Abstract

Abstract Introduction The upwards trend in opioid prescribing poses concern for older adults given risks associated with prolonged opioid use.[1] With patients 65 years or older being predominantly prescribed stronger opioids, minimising duration of therapy is important to reduce falls, fracture and hospitalisation risk.[2] Studies assessing the associations between initial prescription attributes and duration of opioid use have yet to be conducted in an Irish setting. Aim This study aims to assess long-term opioid prescribing in opioid-naive patients initiated on opioid therapy in hospital, and patient and discharge prescription factors associated with long-term opioid prescribing. Methods This retrospective cohort study included approximately 40,000 patients aged ≥65 years from 44 GP practices during 2012-2018 in Ireland. Using GP record and hospital discharge data, individuals initiating an opioid at hospital discharge who were opioid naive (no opioid prescription in the previous 365 days) were identified. The primary analysis excluded cancer-related hospitalisations based on ICD-10 classification. Among non-cancer-related hospitalisations, Cox regression analysis assessed associations between patient and discharge prescription factors (opioid drug, duration, tapering instructions, as needed use specified) with the duration of opioid continuation post-discharge. Results Overall, 975 non-cancer-related opioid-naive patients were initiated on opioids at discharge (48.4% male, mean age 77.9 years). Of the 975 patients, 141 (14.5%) were prescribed 2 opioids and 10 (1.0%) were prescribed 3 opioids. Forty-one percent (n=403) continued opioid therapy following discharge, and 8.2% (n=80) were continually prescribed for >365 days. Of those who discontinued therapy within 365 days, the mean time to discontinuation was 106 days. Initial prescription factors including morphine (HR 0.43, 95%CI 0.23-0.80); duration ≥14 days (HR 0.58, 95%CI 0.39-0.86); and no duration stated (HR 0.46, 95%CI 0.32-0.66) (Figure 1) had statistically significant associations with long-term opioid therapy among those with non-cancer hospitalisations, adjusting for other factors. Conclusion Opioid type and initial prescription characteristics such as duration were associated with longer duration of use. Although limited by lack of information on the specific indication for opioid use, this study is the first in the Irish setting to evaluate the prescribing factors which may be modified to reduce their contribution to prolonged opioid therapy post-discharge.