Presence of chronic liver disease, prior use of gabapentin and being discharged to a long-term hospital or rehabilitation center are among some of the factors associated with a new opioid use disorder diagnosis.
A new retrospective cohort study evaluated opioid-naive patients who were prescribed at least one opioid during an inpatient hospital visit to determine baseline factors associated with a new diagnosis of opioid use disorder (OUD). The ultimate goal was to better understand factors associated with the disorder in order to inform pain management strategies in this setting.
The team of investigators, led by Elizabeth C.S. Swart, BS, UPMC Centers for High-Value Health Care and Value-Based Pharmacy Initiatives, UPMC Health Plan, alluded not only to the 2.5 million individuals living with OUD in the US, but also the economic impact of this epidemic.
In the US, the cumulative societal cost of opioid use disorder is nearly $80 billion and has been associated with the rise of illicit drug use. Investigators acknowledged the numerous methods of response to this crisis, including policy and treatment approaches aimed at harm-reduction, medication-assisted treatment (MAT), Good Samaritan laws and increased availability of naloxone.
“Other efforts to curtail the epidemic have mainly focused on reducing inappropriate prescribing,” they wrote, “which have led to decreased opioid prescriptions; however, rates of OUD, nonmedical opioid use, and opioid-related overdose deaths remain prevalent.”
While many studies have evaluated inpatient opioid administration, this study explored baseline characteristics including medication use pre-admission, during hospitalization, and at discharge to identify associations with a new diagnosis of opioid use disorder in the year following hospitalization.
"To our knowledge, this is one of the first studies to evaluate patient- and medication-related risk factors in the development of OUD in opioid-naïve inpatients exposed to opioids during hospitalization," investigators wrote. "The inpatient setting is a critical setting for opioid exposure and can serve as a conduit to long-term opioid use."
Investigators developed a dataset of electronic health records (EHR), pharmacy claims, and medical claims that reflected data of patients aged 14 or older who were discharged between 2014-2017 from UPMC Health Plan, a vertically integrated healthcare delivery and finance system. Patients who were diagnosed with OUD, or prescribed an opioid prior to the study onset were excluded.
Out of the final sample of 23,033 opioid-naive patients, 2.1% were diagnosed with the disorder within a year of receiving their first opioid in the hospital. Within that population, investigators discovered that a substantial number of those individuals were young and white, residing in high opioid geo-rank regions and had a history of nonopioid related drug disorder, tobacco use, and gabapentin use.
When examining the 6 months prior to hospital admission among those with a new OUD diagnosis, investigators observed that the average number of prescriptions filled were higher, the number of days’ supply was higher, and prescriptions filled for gabapentin and skeletal muscle relaxants were higher, compared with those without a new diagnosis.
"Notably, although a baseline history of skeletal muscle relaxant use was more common in those who were diagnosed with OUD, in our multivariable model it was not associated with subsequent OUD" they wrote. "This is contrary to similar studies which looked at factors influencing long-term opioid use among opioid-naïve patients."
Adjusted multilevel mixed-effects logistic regression models were used to indicate the factors associated with a new opioid use disorder diagnosis within 12 months of discharge.
“Overall, presence of chronic liver disease and being discharged to a long-term hospital, rehabilitation center, or special needs facility were significant factors for patients in the surgery cohort who developed OUD (P< .0001). For patients who did not receive surgery, opioid geo-rank was significant at P< .0001 for those diagnosed with OUD compared to those who were not.”
General characteristics that were associated with higher odds of a diagnosis included being between the ages of 25–34, compared to those 65 years of age and older. When looking at race, White patients were associated with higher odds of a diagnosis.
Medical history and comorbidites played a significant role. Previous history of chronic liver disease was associated with higher odds of an OUD diagnosis in the surgery cohort. Those with nonopioid-related drug disorder (excluding cannabis and cocaine), tobacco use disorder, and mental health conditions also demonstrated higher odds of diagnosis.
Investigators noted that the use of benzodiazepines or skeletal muscle relaxants in the 12 months prior to index hospitalization event was not associated with an OUD diagnosis; however, use of gabapentin 12 months prior to index date was.
Additionally, patients discharged to long-term hospitals, rehabilitation centers, or special needs facilities instead of going home were more likely to receive a diagnosis.
"Future research is needed to evaluate if identification of high-risk patients can mitigate risk of subsequent OUD through proactive effective pain management strategies that include reducing opioid exposure when feasible," investigators concluded.
The study, "Patient and medication-related factors associated with opioid use disorder after inpatient opioid administration" was published in the Journal of Hospital Medicine.