Opioid utilization patterns in United States individuals with sickle cell disease

To the Editor: Sickle cell disease (SCD) is the most commonly inherited blood disorder in North America, and impacts approximately 100 000 people in the United States (U.S).1 Affected individuals generally live on average 30 years less than the general population.2 These individuals suffer acute, intermittent painful episodes (vaso-occlusive crisis [VOC]), as well as fatigue, chronic organ damage and chronic pain.2 Currently, the National Heart Lung and Blood Institute and Centers for Disease Control and Prevention recommend aggressive treatment with opioids during acute VOC and as a therapeutic option for chronic pain. Despite these guideline recommendations for aggressive opioid treatment for acute pain in SCD, the high opioid doses often used for SCD-related pain may result in a negative perception of affected persons by some healthcare professionals.3 Additionally, the current national opioid crisis is impacting overall opioid prescribing including for the SCD population.3 Despite the critical need for and concerns about opioid-based pain management, there is limited evidence on current opioid treatment patterns among individuals with SCD and any potential link to the U.S. opioid epidemic. This study aims to describe opioid utilization in individuals with SCD within the context of healthcare utilization, and the U.S. opioid epidemic among a published broader U.S. population. Since individuals with SCD are likely to have public health insurance coverage, and literature suggests that utilization (including opioid use) may differ between privately and publicly insured indidviduals,4 our study examined utilization in both commercially-insured individuals and those with Medicaid coverage. This was a retrospective observational study using administrative claims data with detailed methods provided in the Supporting Information Appendix SA1. All reported measures are for the year prior to SCD indication. For each annual cohort, approximately 2884 Commercial and 5612 Medicaid SCD individuals met the final inclusion criteria (Supporting Information Table S1). Commercial individuals were older but less likely to have prespecified comorbid conditions including chronic pain and acute chest syndrome as described in detail in Supporting Information Table S2. Across payers, the proportion of individuals with SCD with any opioid use was significantly larger in the 18-30-age group (vs <18 years), and remained constant after age 30, with significantly higher use observed among Medicaid individuals (Supporting Information Figure S1 and Supporting Information Table S3). Average utilization and morphine equivalent dose (MED) per patient exhibited comparable trends across payers (Figure 1). Across payers, MED was significantly higher in the ≥31 age group than in the 18-30 age group, while individuals with opioid use remained constant (Supporting Information Figure S1). Medicaid individuals age 18-30 years (vs <18 years) had a significant 3-fold increase in the number of ED visits, VOCs and inpatient admissions. Commercial individuals also had a significant 2-fold increase in the number of VOCs, ED visits, and inpatient admissions between these age groups (Figure 1). The Commercial cohort (vs Medicaid) had a similar proportion of individuals who used hydroxyurea (21% vs 27%) and a significantly larger proportion who had a hematologist/oncologist visit in the last year (43% vs 7%, P < .001); (Supporting Information Table S4). Across all payers, only a small proportion of individuals had ≥90 days of hydroxyurea supplied (Supporting Information Table S4). Across payers, those with a hematologist/oncologist visit and ≥90 days of hydroxyurea supplied decreased from the youngest to the oldest age groups (Supporting Information Table S5 and Supporting Information Figure S2). There has been a marked increase in opioid prescriptions dispensed over time in the broader U.S. population5 (Supporting Information Figure S3A), which substantially exceeds the underlying growth of the U.S. population.6 In contrast, opioid use was substantial but stable over time across both cohorts of individuals with SCD (Supporting Information Figure S3B). Of note, opioid overdose related deaths among individuals with SCD was also markedly lower compared with other disease states (Supporting Information Figure S4). Opioids remain the recommended treatment for SCD individuals with acute pain. To the best of our knowledge, this study is the first comprehensive assessment of opioid use trends and concurrent healthcare utilization in this population reported by age and payer. Consistent with a recent study in this population, opioid use was greater among older individuals,7 and differed markedly between payers. These findings may reflect the increasing incidence of SCD complications as affected individuals age. The greater incidence of VOCs among Medicaid individuals potentially reflects greater severity of SCD symptoms in this population. In our sample, both Commercial adults and all Medicaid-covered individuals generally did not consistently fill hydroxyurea prescriptions or have relevant specialty care from a hematologist/oncologist. Determining the reason behind these apparent gaps in care was beyond the scope of this study; however, the limited use of a SCD specialist suggests that adults with SCD may have less access to care than affected children, which is consistent with literature indicating that a comprehensive care plan exists for pediatric individuals.2 This is of particular concern since the severity of SCD complications increases with age and adults may also lack insurance coverage which is an additional barrier to obtaining proper disease management.2 The simultaneous increase in the number of VOCs, healthcare utilization and opioid medication use among SCD individuals age >18 compared with SCD individuals <18 suggest SCD-related pain is more common among adults and further confirms the cumulative effect of this disease.7 Although SCD adults have high opioid use (vs children), total opioid use across payers has remained constant over time (2008-2013), while reported opioid use in a broader U.S. population has increased to epidemic levels.5 Methodological differences preclude direct comparison of results for the broader U.S. population to results from our study,5 but our findings offer some reassurance regarding use of SCD. The results of this study should be interpreted within the context of certain known limitations. The potential for misclassification of SCD is present as patient selection relied on administrative claims data which is for billing purposes only, and data from this source are subject to coding limitations in addition to possible data entry errors. To increase the specificity of SCD identification, we excluded individuals with only a "diagnostic" SCD claim, thereby only included those with a confirmed SCD diagnosis. VOCs were based off ICD-9 codes not chart data, and may not reflect a true VOC. Opioid use and MED were based on available data in recorded outpatient pharmacy claims, and only represent dispensed opioids and do not reflect differing usage based on individual needs or provide insight into individuals' actual opioid consumption. This study provides a comprehensive overview of opioid dispensing and healthcare utilization among SCD individuals. In sharp contrast to increasing opioid use trends in the general U.S. population, this population has generally demonstrated stable opioid use over time. Of concern, this broader epidemic is shaping physician prescribing patterns and possibly limits access of SCD individuals to guideline-recommended opioid therapy for SCD-related pain. These results likely reflect both changes in access to care, and health care coverage as individuals age. Current results support prior literature which indicates a worsening of the disease during life, and describe novel findings on the differing experiences of individuals with Medicaid and Commercial insurance. The development and approval of new disease-modifying medications targeting the pathophysiology of SCD may be a path to decrease opioid use in this population by decreasing commonly observed acute and chronic pain, and SCD complications, especially among adults. The authors would like to acknowledge Biswarup Mukherjee for his contributions to this project. SKB is affiliated with Thomas Jefferson University and consulted with Global Blood Therapeutics on this project. JK is affiliated with the Medical University of South Carolina and consulted with Global Blood Therapeutics on this project. IA and RH are employees of Global Blood Therapeutics. SW (consultant through Wade Outcomes Research and Consulting), and VN (employee) are affiliated with Truven Health Analytics, which was paid by Global Blood Therapeutics for the research study and assistance in developing this manuscript. CD is affiliated with Emory University and consulted with Global Blood Therapeutics on this project. TABLE S1 Annual cohorts TABLE S2 Baseline Patient Characteristics TABLE S3 Measures of opioid utilization TABLE S4 Measures of annual healthcare utilization for all patients TABLE S5 Measures of annual healthcare utilization across age groups FIGURE S1 Average opioid utilization FIGURE S2 Hydroxyurea utilization by age groups FIGURE S3 Opioid use in general U.S. population (A) vs SCD population (B) FIGURE S4 Opioid-related deaths among patients with SCD and other diseases Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.