The risk factors for concurrent use of benzodiazepines or death among non-cancer patients with long-term opioid analgesics use

• Main Authors: Po-FengLee, 李柏鋒
• Other Authors: Chung-Yi Li
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/45fj83

碩士 === 國立成功大學 === 公共衛生研究所碩士在職專班 === 107 === Introduction Over the past 20 years, the consumption of opioid analgesics has significantly increased in many countries of North America and Europe. There has been international concern over the rise in opioid analgesics addiction and black-marketing, as well as opioid intoxication and mortality. Although the average opioids consumption in Taiwan is still much lesser than U.S. and Australia, it rose by 41% between 2002 and 2014. This was because Taiwanese government implemented more strict policies on prescription of opioid analgesics, as well as the introduction of new opioid analgesics. However, there were only few articles conducted to address prescriptions of opioid analgesics in Taiwan. This study aimed to investigate the trend of long-term opioid analgesics use in Taiwan and the risk factors for subsequent intoxication or death among non-cancer patients with long-term opioid analgesics use. Goal of study We aimed to investigate (1) sociodemographic characteristics; (2) risk factors of concurrent benzodiazepine use; (3) risk factors of opioid intoxication; (4) risk factors of death; (5) age and sex standardized mortality ratio (SMR); and (6) the distribution of cause of death, among non-cancer patients with long-term opioid analgesics use. Material and methods Our study design was a longitudinal and retrospective cohort study; and the database analyzed was all claim data, between 2000-2013, of a randomly selected population-based cohort (2-million people) provided by the Ministry of Health and Welfare. Our study cohort consisted of 12,990 adults with non-cancer diagnosis and long-term opioid analgesics use between 2001 and 2012. Long-term opioid analgesics use was defined as consecutive opioid analgesics use for more than 14 days, or intermittent opioid analgesics use more than 28 days in a 90 days period. We performed survival analysis (Cox regression model) to identify risk factors for subsequent concurrent benzodiazepine use and death. Defining the whole population as reference population, we also calculated age and sex standardized mortality ratio with indirect standardization method. Results We identified 12,990 beneficiaries to be our study cohort. In this cohort, 7,826 (60.2%) enrollees were male, and 8715 (67.1%) enrollees were aged between 45 and 64 years. Totally 2,852 (22%) enrollees were noted to have combined use of benzodiazepine, and 2,327 (81.6%) out of these 2,852 enrollees were identified within 1 year after the definition of long-term opioid analgesic were met. Risk factors of concurrent benzodiazepine use were female gender, middle age (45-64 years old), lower socioeconomic status, higher Charlson’s Comorbidity Index (CCI) score, and mental illness. We identified only 5 enrollees with opioid intoxication, and this is very likely to be under-estimated. There were 558 deaths (4.3%) noted in our cohort during follow-up, corresponding to a mortality rate of 11.6 / per 1,000 person-years. The risk factors of death included male gender, older age, combined benzodiazepine use, higher CCI score, living in rural area, lower income, affective disorder, and alcohol use disorder. Compared with the whole population in 2006, the all-cause SMR of our cohort between 2001 and 2013 is 1.41 (95% confidence interval=1.29-1.53). The sex-specific SMR was higher in female enrollees than in male enrollees. Conclusion The number of enrollees with long-term opioid analgesics use and combined benzodiazepine use, respectively, was increasing between 2001 and 2012. We also found that sociodemographic characteristics, comorbidity, and concurrent benzodiazepine in enrollees with long-term opioid analgesics use were significantly associated with death. Our cohort had a higher all-cause SMR compared with general population. Moreover, since it is difficult to identify opioid intoxication events correctly by International Classification of Disease codes in the claim data of National Health Insurance, establishing other valid surveillance systems to minitor intoxication events are needed and should be considered. Key words: prescription opioids; benzodiazepine; mortality; risk factors