BACKGROUND: Longitudinal population-based studies of long-term opioid therapy (L-TOT) in chronic non-cancer pain (CNCP) patients are sparse. Our study investigated incidence and predictors for initiating L-TOT and changes in self-rated health, pain interference and physical activities in long-term opioid users.
METHODS: Data were obtained from the national representative Danish Health and Morbidity Surveys and The Danish National Prescription Registry. Respondents with no dispensed opioids the year before the survey were followed from 2000 and from 2005 until the end of 2012 (n = 12,145). A nationally representative subsample of individuals (n = 2015) completed the self-administered questionnaire in both 2000 and 2013. Collected information included chronic pain (≥6 months), health behaviour, self-rated health, pain interference with work activities and physical activities. Long-term users were defined as those who were dispensed at least one opioid prescription in six separate months within a year.
RESULTS: The incidence of L-TOT was substantially higher in CNCP patients at baseline than in others (9/1000 vs. 2/1000 person-years). Smoking behaviour and dispensed benzodiazepines were significantly associated with initiation of L-TOT in individuals with CNCP at baseline. During follow-up, L-TOT in CNCP patients increased the likelihood of negative changes in pain interference with work (OR 9.2; 95% CI 1.9-43.6) and in moderate activities (OR 3.7; 95% CI 1.1-12.6). The analysis of all individuals indicated a dose-response relationship between longer treatment duration and the risk of experiencing negative changes.
CONCLUSIONS: Individuals on L-TOT seemed not to achieve the key goals of opioid therapy: pain relief, improved quality of life and functional capacity.
SIGNIFICANCE: Long-term opioid therapy does not seem to provide pain relief, improvement in HRQOL and physical capacity in CNCP patients in a general population.