Emergency Admissions and the Prescribing of Buprenorphine, Methadone, and Psychotropic Medications in People with Sickle Cell Disease: An Analysis of National Insurance Claims

Background

Although there have been efforts to address overdose risks and treatment gaps in general populations receiving opioids, there is has been little research examining potential drug-related poisoning risks in some of the highest-need and complex populations, including those with sickle cell disease (SCD). Chronic opioid medications are necessary and well-tolerated for the treatment of chronic pain, diffuse inflammation, chronic organ damage, and vaso-occlusive pain crises for people with SCD. Previous studies have found that patients with SCD are not more likely to develop addiction to opioids than the general population. However, amid the ongoing endemic of drug-related poisonings in the U.S., epidemiological data is scarce regarding the clinical characteristics of people who have SCD and physiologic opioid tolerance and dependence. There has been particularly limited work on the intersection of sociodemographic characteristics, substance use, and other mental health conditions in the SCD population. Concern has also been raised about the ongoing opioid epidemic contributing to increased stigma, and greater barriers to the management of acute and chronic pain in SCD. Studies have shown that preoccupation with opioid prescription monitoring hindered the ability of physicians to provide comprehensive SCD care. People with SCD often face more restrictive utilization management policies for opioids than peers with other forms of chronic pain (i.e., cancer-related pain). In order for us to develop better interventions for people with SCD, real-world data is urgently needed to describe the current treatment landscape and identify gaps in care for people with SCD and physiologic opioid dependence or tolerance. Using administrative claims data in the US, we aim to: 1) quantify the burden of hospitalization and emergency room admission for medication-related poisoning events, 2) describe the prevalence of opioid prescriptions, such as buprenorphine and methadone, in people with SCD and physiologic opioid dependence or misuse.

Methods

In this retrospective cohort study of people with physiologic opioid dependence or opioid use disorder (OUD), we analyzed ten years (1/2006-12/2016) of data from the Merative™ MarketScan® Commercial and Multi-State Medicaid Databases. The predictor variable was SCD (versus a comparator group of people with chronic pain) with diagnoses identified using ICD-9/10 codes. The outcome variables were 1) emergency admissions or hospitalization for non-fatal drug/medication-related poisoning events during enrollment, and 2) prescriptions for buprenorphine, methadone, and other opioids, identified via procedure codes and linked pharmacy records respectively. We used multivariable cox regression to estimate the relationship between SCD diagnosis and time until poisoning events. Multivariable poisson regression models were used to assess the relationship between SCD and buprenorphine, methadone, and other opioid prescriptions.

Results

Our overall cohort consisted of 25,448 people with ICD9-10 opioid dependence or misuse, including 313 with SCD and 25,135 with non-SCD chronic pain. In the SCD cohort, the median age was 30.8 years, and 82% were enrolled in Medicaid. Compared to the non-SCD chronic pain cohort, the SCD cohort had more hospitalizations or emergency admission for iatrogenic events (37.4% vs 9.8%, aHR=2.21[2.12-2.31]). With regards to medication receipt compared to peers with non-SCD chronic pain, the SCD cohort had a lower prevalence of buprenorphine prescriptions (8% vs 30.1%, aRR=0.20[0.18-0.21]) and a higher prevalence of methadone prescriptions for chronic pain (63.6% vs 6.8%, aRR=5.36[5.21-5.51]) and non-methadone/non-buprenorphine opioid prescriptions (80.5% vs 33.7%, aRR=2.20[2.15-2.25]).

Conclusions

We observed an elevated rate of emergency admission and/or hospitalization for drug/medication-related poisoning, as well as a significant burden of psychiatric comorbidity in a cohort of people with SCD and physiologic opioid dependence or OUD. The prescribing of potentially safer opioids, such as buprenorphine, was uncommon in the cohort. Further research is urgently needed to optimize opioid treatment strategies, combat stigma, and improve the comprehensive management of pain and co-occurring disorders in the SCD population.