Opioid dispensing declines for patients with schizophrenia, bipolar disorder

Prescription opioid dispensing decreased for patients with schizophrenia, as well as those with bipolar disorder, between 2015 and 2019, according to study results presented at NEI Congress.

“Patients with serious mental illness, such as bipolar disorder (BD) [or schizophrenia], may have an elevated risk of complications associated with opioid medications, including overdose and opioid misuse,” Brittany D. Roy, MPH, manager of epidemiology at Alkermes Inc. in Massachusetts, told Healio Psychiatry. “Declining trends in opioid dispensing rates have been reported in the general population as more precautions have been put in place; however, it is unknown whether similar decreases have occurred in populations with serious mental illness, such as bipolar disorder [or schizophrenia].

“Our objective was to utilize real-world data to assess prescription opioid dispensing patterns over time from 2015 to 2019 among individuals diagnosed with BD [or schizophrenia] as compared to an age- and sex-matched control population,” Roy added.

In the current retrospective, observational database study, Roy and colleagues analyzed health insurance claim data included the IBM MarketScan Commercial Database and the Multi-State Medicaid Database. Patients were 18 to 64 years old on the index data of the analysis year and had one or more inpatient or two or more outpatient diagnosis claims for BD or schizophrenia according to ICD-9 criteria during the cohort selection period. They had one or more claims before the baseline period to ensure diagnosis prevalence. Patients also had at least 6 months continuous enrollment in medical and pharmacy benefits during the baseline period and at least 12 months continuous enrollment in medical and pharmacy benefits during the analytic year.

The researchers matched each patient with bipolar disorder or schizophrenia to one control from the same databased based on age and sex during the same analysis year. They examined demographics and clinical characteristics, such as comorbidities and medication utilization, during the baseline period for each analysis year. Further, they extracted chronic or nonchronic prescription opioid dispensing data for each analysis year for patients with schizophrenia or BD and controls. They used logistic regression models, adjusted for relevant confounders (e.g., age, sex, race, year, pain diagnoses, medical comorbidities), to determine likelihood of receiving prescription opioid dispensing for patients with schizophrenia or BD compared with controls.

Roy and colleagues reported a sample size ranging between 49,907 and 58,200 patients with BD and corresponding numbers of matched controls in the commercial database, as well as between 49,734 and 69,987 in the Medicaid database across the 5 analysis years. Sample sized ranged between 4,732 and 7,784 patients with schizophrenia and corresponding numbers of matched controls in the commercial database and between 29,967 and 39,574 in the Medicaid database across the 5 analysis years.

In both databases, patients with BD or schizophrenia had higher rates of medical and psychiatric comorbidities, such as pain diagnoses, compared with matched controls. Chronic and nonchronic prescription opioid dispensing decreased significantly for patients with BD and for their matched controls, as well as overall for patients with schizophrenia and their matched controls, over time between 2015 and 2019 in both databases. After adjustment for relevant confounders, such as pain diagnoses, dispensing remained consistently higher for patients with BD vs. controls in both databases. Dispensing for those with schizophrenia vs. matched controls was higher in the commercial database but was similar in the Medicaid database.

Among patients with BD and controls, the researchers noted a decline in the average annual number of nonchronic opioid dispensing days between 2015 and 2019 across databases. In 2019, the most recent available analysis year, they noted 10 days or fewer as the average annual number of nonchronic opioid dispensing days for patients with BD and their matched controls in each database, which indicated exposures for a relatively short time period.

For patients with schizophrenia and matched controls, the proportion of those dispensed opioids chronically was low, at less than 5% in the commercial database and less than 10% in the Medicaid database. Nonchronic opioid dispensing was similar between patients with schizophrenia and matched controls across both databases.

“These results highlight the importance of integrated medical and psychiatric care for individuals with serious mental illness, including BD [or schizophrenia],” Roy said.