Association Between Formulary Coverage and Use of Abuse-Deterrent Prescription Opioids, Risk for Abuse or Overdose, and Associated Healthcare Resource Utilization

Prescription opioids are effective medications for the management of pain^1-3; however, misuse or abuse of prescription opioids can lead to serious consequences, including addiction, overdose, and death. In 2016, more than 11.5 million persons misused prescription opioids in the United States,⁴ including 2.5 million young adults between the ages of 18 and 25 years.^5,6 In 2017, approximately 36% of opioid overdose deaths in the United States involved the use of prescription opioids.^7,8

Substance abuse related to prescription opioids imposes a substantial economic burden on the US healthcare system. Weiss and colleagues estimated that the national rate of prescription and illicit opioid abuse–related hospitalizations increased by 64.1% from 136.8 inpatient stays per 100,000 persons in 2005 to 224.6 inpatient stays per 100,000 persons in 2014.⁹ A total of 142,557 emergency department visits (15.7 per 10,000 visits) between July 2016 and September 2017 were attributable to opioid overdose.¹⁰

In response to the high prevalence and substantial healthcare burden associated with prescription opioid abuse, the US Food and Drug Administration (FDA) issued an Opioids Action Plan that emphasizes expanding access to abuse-deterrent formulation (ADF) opioids as an important goal in the fight against opioid abuse.^11-13 ADF opioids maintain effective pain control while reducing the risk for abuse by manipulation of the formulation.¹⁴ Most currently marketed ADF opioids use physical and chemical barriers or agonist and antagonist combinations to deter abuse, although other technologies have also been used or are in development.¹⁴

Despite the potential for ADF opioids to reduce opioid misuse and abuse, their uptake has been low relative to the overall use of prescription opioids.¹⁵ This may partially result from limited formulary coverage and utilization management requirements (ie, prior authorization, step therapy) imposed on these drugs. Other factors, such as physician prescribing habits or patient preference, may have also contributed to the slow uptake.

The objectives of this study were to evaluate the associations between formulary coverage of ADF opioids and health plan members’ utilization of ADF opioids, and their risks for opioid abuse or overdose and associated hospitalizations, emergency department visits, and related medical expenditures.

Methods

In this cross-sectional, multiyear panel study we used administrative claims data from 2015 and 2016 from the Inovalon Medical Outcomes Research for Effectiveness and Economics (MORE2) Registry. The MORE2 Registry is a multipayer claims database that contains member enrollment information and medical and pharmacy claims for approximately 287 million unique and deidentified health plan members from more than 150 commercial, managed Medicaid, and Medicare Advantage plans across the United States. Claims data in the registry were linked with each member’s plan benefit design data that include insights into formulary policies for the coverage of ADF opioids.

The study sample included continuously enrolled adult (aged ≥18 years) health plan members who filled at least 1 prescription drug claim for a Schedule II opioid in the first 6 months of 2015 or 2016, including any oral form of morphine, hydrocodone, hydromorphone, oxycodone, oxymorphone, tapentadol, meperidine, levorphanol, opium, codeine sulfate, or buccal, nasal, oral, oral transmucosal, transdermal, or sublingual forms of fentanyl.

Three ADF opioids available in the US market during the study period were considered in the analysis, including morphine sulfate and naltrexone hydrochloride (HCl) extended-release (ER; Embeda, which is to be discontinued in 2020), hydrocodone bitartrate ER (Hysingla ER), and reformulated oxycodone HCl controlled-release (CR; reformulated OxyContin). Oxycodone ER (Xtampza) was excluded from the analysis, because it was approved by the FDA near the end of the study period. Other opioids with tamper-resistant properties that do not have FDA-approved abuse-deterrent labeling were included and were considered in this study as non-ADF opioids.

Excluded from the study were plan members with evidence of opioid abuse or overdose in the baseline period, which was defined as having at least 1 medical claim (based on International Classification of Diseases, Ninth Revision [ICD-9]/ICD, Tenth Revision [ICD-10] codes) with a diagnosis of opioid, heroin, methadone or synthetic opioid abuse, intentional or accidental poisoning in the primary or secondary position (see Appendix Table), and/or any administration or prescription for potential opioid abuse treatment, including naloxone monotherapy, na­loxone combination drugs, naltrexone (except morphine sulfate and naltrexone HCl), buprenorphine, or methadone. Also excluded were patients for whom plan benefit design data were unavailable (Figure 1). Finally, members who disenrolled from or changed their health plan midyear were also excluded from the study.

Because a health plan’s formulary can change from year to year, the unit of analysis was the member-year. Individuals could contribute 2 observations to the study sample if they met the eligibility criteria in both years. The first prescription fill for a Schedule II opioid in each calendar year (ie, the measurement year) was defined as the index fill. The calendar year before the measurement year served as the baseline period for capturing the member’s demographics, comorbidities, and baseline healthcare resource utilization (HCRU).

Members were assigned to a study group based on their health plan formulary coverage for ADF opioids. Two analyses based on formulary coverage were constructed. The first analysis (Definition A) considered whether and which ADF opioid drugs were covered by the plan formulary, regardless of tier placement or utilization management restrictions. Three member cohorts were formed based on this definition: (1) no ADF opioids, which included members in plans that excluded all ADF opioids from the plan formulary; (2) oxycodone HCl CR only, which comprised members in plans that included oxycodone HCl CR only on the plan formulary; and (3) oxycodone HCl CR plus ≥1 ADF opioids, which comprised members in plans that included oxycodone HCl CR and at least 1 other ADF (morphine sulfate and naltrexone HCl ER, hydrocodone bitartrate ER, or both) on the plan formulary.

The second analysis (Definition B) incorporated tier placement and utilization management restrictions (ie, prior authorization and step therapy) placed on the covered ADF opioids. Each of the 3 ADF drugs was classified based on whether the plan covered the drug on a preferred formulary tier without restrictions. Three member cohorts were constructed based on Definition B, including (1) no ADF opioids were covered on a preferred tier without restrictions; (2) oxycodone HCl CR only covered on a preferred tier without restrictions; and (3) oxycodone HCl CR plus ≥1 ADF opioids covered on a preferred tier without restrictions.

A given ADF opioid was required to meet all 3 criteria (ie, covered on the formulary, on a preferred tier, and without restrictions) to be considered covered; otherwise the ADF opioid was considered to not be covered, based on Definition B. A few health plan members were enrolled in plans that covered only morphine sulfate and naltrexone HCl ER, hydrocodone bitartrate ER, or both on the preferred tier without restrictions. As such, these individuals did not fit into any of the 3 cohorts described above. They were excluded from the analyses because the sample size was too small to generate meaningful results.

Study Outcomes

The study outcomes were assessed for each measurement year. The primary study outcomes included the use of an ADF opioid, defined as the likelihood of filling ≥1 prescriptions for an ADF opioid, and the diagnosis of opioid abuse or overdose, defined as the risk for a medical claim that indicates opioid abuse or overdose. The identification of opioid abuse or overdose was based on the presence of at least 1 inpatient or outpatient medical claim with an ICD-9 or ICD-10 diagnosis code for opioid, heroin, methadone, or synthetic opioid abuse, or intentional or accidental poisoning in the primary or secondary position (see Appendix Table).

Prescription claims for potential opioid abuse treatments (naloxone monotherapy, naloxone combination drugs, naltrexone [except morphine sulfate and naltrexone HCl ER], buprenorphine, or methadone) were not used to assess opioid abuse or overdose outcomes, because they can be prescribed for indications other than opioid abuse. Additional outcomes included the risk for opioid abuse or overdose–related hospitalization or emergency department visit, defined as having at least 1 inpatient hospital or emergency department medical claim that included a principal or a secondary diagnosis code for opioid abuse or overdose.

Hospitalization was also assessed for medical claims with ICD-9 or ICD-10 codes for opioid abuse or overdose in the primary position only. Opioid abuse or overdose–related medical expenditures were assessed by summing the standardized allowed payment amount from medical claims with ICD-9 or ICD-10 diagnosis codes for opioid abuse or overdose in the primary or secondary position during the measurement year.

Statistical Analysis

We performed descriptive analyses to summarize the members’ demographic and clinical characteristics and baseline HCRU and costs. Means and standard deviations were reported for the continuous variables, whereas frequency and percentages were reported for the categorical variables. The mean per-member per-month (PMPM) opioid abuse or overdose–related medical costs in the measurement year were calculated. Using the overall study cohort as the reference population, the adjusted rates or means were calculated for each study outcome using the LSMEANS statement as part of the SAS GENMOD procedure for estimating the adjusted relationships between formulary coverage of ADF opioids and the study outcomes.

Multivariable logistic regression models were used to compare the effect of varying levels of ADF formulary coverage on the binary study outcomes. To account for skewness and the dispersion of cost outcomes, negative binomial models were estimated to assess the association between formulary coverage and opioid abuse or overdose–related medical costs. All calculated outcomes (ie, estimated rates, odds ratios [OR], and cost ratios) were adjusted for member demographics (ie, age, region, payer type), comorbid conditions (ie, Charlson Comorbidity Index score, history of nonopioid substance abuse, history of mental disorders, and concomitant use of benzodiazepine), all-cause HCRU (ie, number of all-cause hospitalizations and number of emergency department visits), and costs (ie, monthly medical costs and monthly prescription drug costs) at baseline to mitigate confounding.

All statistical analyses were performed using SAS version 9.4 (SAS Institute; Cary, NC).

Results

A total of 1,350,607 members from 135 US health plans were included in the study (Table 1). A majority (43.4%) of the study members were aged 45 to 64 years, 61.9% were female, and 47.4% were enrolled in a managed Medicaid plan. Medical claims for 1 or more mental health disorders were present in 35.7% of the members, and 25.2% of the members had prescription claims for concomitant benzodiazepine use, which was defined as having ≥1 prescription claims for a benzodiazepine during the measurement year (Table 1).

Overall, 1.8% of the study population filled a prescription for an ADF opioid and 3.3% had medical claims for opioid abuse or overdose during the study period. Hospitalizations and emergency department visits associated with opioid abuse or overdose occurred in 0.6% and 0.5% of the study cohort, respectively. Hospitalizations with opioid abuse or overdose coded as the primary diagnosis occurred in 0.1% of the study cohort.

ADF Opioids Covered on Formulary (Definition A)

Most (77.1%) of the Definition A cohort was enrolled in plans that covered oxycodone HCl CR plus ≥1 ADF opioids on the formulary, 5.8% were in plans covering oxycodone HCl CR only, and the remaining 17.1% were in plans that had no ADF opioid coverage (Table 1). The members who were enrolled in Medicare Advantage plans were the most likely to have no ADF opioid access (Figure 2). A majority of members with commercial or managed Medicaid coverage had access to oxycodone HCl CR plus ≥1 ADF opioids.

The adjusted mean utilization rate of ADF opioids was lowest in the no ADF opioid group (0.62%; 95% confidence interval [CI], 0.59%-0.65%), followed by 1.21% (95% CI, 1.15%-1.27%) in the oxycodone HCl CR–only group, and 1.15% (95% CI, 1.12%-1.17%) in the oxycodone HCl CR plus ≥1 ADF opioids group (Table 2). The adjusted rate of opioid abuse or overdose diagnosis was lowest for members in plans covering oxycodone HCl CR plus ≥1 ADF opioids (mean, 2.06%; 95% CI, 2.03%-2.10%), followed by plans covering oxycodone HCl CR only (mean, 2.67%; 95% CI, 2.56%-2.80%), and highest in plans covering no ADF opioid (mean, 2.95%; 95% CI, 2.85%-3.05%). Plans with the broadest coverage (ie, oxycodone HCl CR plus ≥1 ADF opioids) had the lowest adjusted opioid abuse or overdose–related PMPM medical costs ($30.65; standard error [SE], $0.57).

The results from multivariate regression models suggested that members enrolled in health plans with coverage of 1 or more ADF opioids were significantly more likely to fill a prescription for an ADF opioid than those enrolled in plans that did not cover ADF opioids (adjusted OR, 1.95 [95% CI, 1.85-2.06] for oxycodone HCl CR only vs no ADF opioid, and adjusted OR, 1.86 [95% CI, 1.76-1.96] for oxycodone HCl CR plus ≥1 ADF opioids vs no ADF opioid; both P <.0001; Table 3).

More important, the risk for opioid abuse or overdose diagnosis was lower among members enrolled in plans with broad ADF opioid coverage (adjusted OR, 0.91 [95% CI, 0.86-0.95] for oxycodone HCl CR only vs no ADF opioid, and adjusted OR, 0.70 [95% CI, 0.67-0.73] for oxycodone HCl CR plus ≥1 ADF opioids vs no ADF opioid; both P <.0001).

Furthermore, the members of plans that covered oxycodone HCl CR plus ≥1 ADF opioids had a significantly lower risk for an opioid abuse or overdose diagnosis than the members in plans that covered oxycodone HCl CR only (adjusted OR, 0.77; 95% CI, 0.73-0.81; P <.0001). There was a 16% lower risk for opioid abuse or overdose–related hospitalizations and a 26% reduction in opioid abuse or overdose–related medical costs among members in plans covering oxycodone HCl CR plus ≥1 ADF opioids compared with their counterparts in plans covering oxycodone HCl CR only on the drug formulary (P <.01 and P <.0001, respectively).

ADF Opioids Covered on Preferred Tier with No Restrictions (Definition B)

Among the Definition B cohort, 69.6% of members were enrolled in health plans that had no ADF opioids on preferred tier without restrictions, 26.2% in plans that covered oxycodone HCl CR only on preferred tier without restrictions, and 3.2% in plans that covered oxycodone HCl CR plus ≥1 ADF opioids on preferred tier without restrictions (Table 1). The members in managed Medicaid plans (94%) and Medicare Advantage plans (77%) were most likely to have no ADF opioid covered, whereas commercial plans were most likely to have oxycodone HCl CR only covered (68%; Figure 2).

The adjusted rates for ADF opioid utilization were lowest among plans covering no ADF opioid (0.82%; 95% CI, 0.80%-0.84%), and highest for plans covering oxycodone HCl CR only (1.67%; 95% CI, 1.62%-1.73%). The plans with the broadest coverage (oxycodone HCl CR plus ≥1 ADF opioids) had the lowest rates of opioid abuse or overdose (1.55%; 95% CI, 1.45%-1.65%) and the lowest mean opioid abuse or overdose–related PMPM medical costs ($27.90; SE, $2.24) compared with plans that covered either no ADF opioid or oxycodone HCl CR only (Table 2).

Multivariate regression modeling results were consistent with formulary Definition A outcomes in that there was an overall trend where broader health plan coverage (covered on a preferred tier with no restrictions) of ADF opioids was associated with improved member outcomes compared with plans with restrictions (Table 3).

Compared with members in plans with no ADF opioid coverage, those enrolled in plans that covered oxycodone HCl CR plus ≥1 ADF opioids (adjusted OR, 1.83; 95% CI, 1.73-1.93) or oxycodone HCl CR only (adjusted OR, 2.03; 95% CI, 1.95-2.12) were significantly more likely to fill a prescription for an ADF opioid (both P <.0001).

The risk for a diagnosis of opioid abuse or overdose was significantly lower among members enrolled in plans with broader ADF opioid coverage (adjusted OR, 0.72; 95% CI, 0.66-0.77) for oxycodone HCl CR plus ≥1 ADF opioids versus oxycodone HCl CR only, and for oxycodone HCl CR plus ≥1 ADF opioids plans versus no ADF opioid (adjusted OR, 0.67; 95% CI, 0.63-0.72; both P <.0001).

The risk for having an opioid abuse or overdose–related hospitalization was 28% lower for members in plans that offered oxycodone HCl CR plus ≥1 ADF opioids compared with no ADF opioid (adjusted OR, 0.72; 95% CI, 0.62-0.84; P <.0001), and the risk was 23% lower compared with members from plans that limited choice to oxycodone HCl CR only.

A statistically significant reduction in opioid abuse or overdose–related medical expenditures was seen among members in plans covering oxycodone HCl CR plus ≥1 ADF opioids compared with those in plans with no ADF opioid (adjusted OR, 0.85; 95% CI, 0.72-0.99; P <.05). A similar magnitude of reduction was observed among members in plans covering oxycodone HCl CR plus ≥1 ADF opioids versus members in plans covering oxycodone HCl CR only (adjusted OR, 0.86; 95% CI, 0.72-1.02); however, the reduction was not statistically significant.

Discussion

Recent studies have evaluated the budget impact of switching patients from non-ADF opioids to ADF opioids. In 2017, the Institute for Clinical and Economic Review (ICER) released the findings of an analysis that considered the health and economic benefits of a policy that required 100% conversion from non-ADF opioid to ADF opioid prescriptions in the state of Massachusetts.¹⁶ ICER’s estimates suggest that such a policy would prevent 850 new cases of opioid abuse over 1 year in a 100,000-member plan at a cost of $599,000 per new case prevented. Net cost-savings were not possible because the lower abuse-related costs of ADF opioids compared with non-ADF opioids were outweighed by higher prescription costs of the ADF drugs.¹⁶

Unlike the ICER assessment, other studies have assessed the budget impact of ADF opioids based on partial formulary replacement with varying results.^17-19 In 2017, Descoteaux and colleagues assumed 5%, 7%, and 10% replacement rates of long-acting non-ADF opioids with their ADF counterparts in years 1, 2, and 3, respectively.¹⁷ Based on 1 million lives, the results suggested that the addition of ADF opioids to a drug formulary would result in an average increase of $0.16 PMPM spending.¹⁷

In 2009, White and colleagues analyzed the formulary budget impact of introducing a theoretical ADF opioid to a commercial drug formulary.¹⁸ Depending on the degree of formulary uptake, potential cost-savings to plan were estimated to be $0.6 billion to $1.6 billion annually.¹⁸ A 2019 study by Yenikomshian and colleagues reported that the difference in ADF opioid technology was associated with varying degrees of the medical cost-savings used in budget calculations to arrive at the total health plan cost.¹⁹

All the studies cited above have concluded that meaningful reductions in opioid misuse or abuse can be observed by shifting patients from traditional non-ADF drugs to ADF drugs. The ability to achieve budget neutrality or cost-savings within a health plan, however, is influenced by drug cost, formulary uptake, and the abuse reduction potential of a given ADF opioid. To our knowledge, this is the first research study to examine how health plan formulary coverage of ADF opioids affects their utilization; the risk for opioid abuse or overdose diagnosis among commercial, Medicare Advantage, and managed Medicaid health plan members; and the associated downstream economic consequences.

Among the 135 managed care plans included in the study, plan formulary coverage of ADF opioids was generally limited during 2015 and 2016. Only 8% of Medicare Advantage, 2% of commercially insured, and 2% of managed Medicaid members in our study were enrolled in plans that made oxycodone HCl CR plus ≥1 ADF drugs available without restrictions or higher patient copays.

Formulary restrictions might have contributed to the low utilization of ADF opioids among the health plan members included in this study. Comparisons across formulary management policies suggest that patients in plans with no coverage of ADF opioids or those that imposed restrictions, such as prior authorization, step therapy, or tiering, are less likely to use ADF opioids.

Members in plans covering oxycodone HCl CR plus ≥1 ADF opioids on a preferred tier without restrictions were almost twice as likely to fill at least 1 prescription for an ADF opioid compared with their counterparts in plans that either did not cover ADF opioids or those that used restrictions, such as utilization management, and/or had high patient copays.

Contrary to our hypothesis that access to more options of different ADF drugs would result in the higher overall utilization of ADF opioids, the utilization rate of ADF opioids in our study was slightly higher in plans covering oxycodone HCl CR only than among members in plans covering oxycodone HCl CR plus 1 or more ADF drugs. This may be a result of other plan-level factors, physician prescribing habits, and/or member characteristics favoring the use of oxycodone HCl CR that were not accounted for in this study. It is also plausible that the number of ADF drugs available only influences the choice of a specific ADF opioid, but it does not affect the overall utilization of ADF opioids.

More important, the members in plans that provided broader access to ADF opioids had a significantly lower risk for being diagnosed with opioid abuse or overdose compared with plans that have more narrow coverage. There was a trend toward a reduced risk for opioid abuse or overdose diagnosis when more ADF drugs (morphine sulfate and naltrexone HCl ER and/or hydrocodone bitartrate ER) were offered on formulary versus the coverage of oxycodone HCl CR only. It is unclear whether higher utilization of ADF opioids or differences in abuse-deterrent properties of the ADF opioids have contributed to this finding. Additional studies are needed to further examine the possible reasons for this study finding.

In general, the reduced incidence of opioid abuse or overdose diagnosis also translated into lower rates of opioid abuse or overdose–related HCRU and direct medical expenditures. Compared with the no ADF group, the oxycodone HCl CR plus ≥1 ADF opioids group consistently had a lower risk for opioid abuse or overdose–related hospitalization and lower direct medical expenditures. However, this trend was not observed for the study outcome on the risk for opioid abuse or overdose–related emergency department visits. The oxycodone HCl CR–only group had a lower risk for opioid abuse or overdose–related hospitalizations but a higher risk for opioid abuse or overdose–related emergency department visits and similar opioid abuse or overdose–related direct medical expenditures compared with the no-ADF opioid group.

Considering that opioid abuse or overdose is a rare event from an administrative claims database perspective, the lack of stability or reliability in the estimates on opioid abuse or overdose–related HCRU and medical expenditures may result from insufficient statistical power. The true societal cost of opioid abuse or overdose is likely much higher than what is captured in our study.

According to the National Center for Health Statistics, the number of reported drug overdose deaths by natural or semisynthetic opioids (most prescription opioids are semisynthetic opioids) remains high, exceeding 12,151 deaths in the 12-month period ending in March 2019.20 Although ADF opioids are not resistant to abuse or addiction, they may present an opportunity for incremental progress toward reducing the potential or risk for opioid abuse or overdose.^14,21

Increasing real-world evidence from observational research demonstrates that ADF opioids are associated with decreased rates of abuse and diversion (ie, transfer from lawful to unlawful opioid use) of opioids in the United States.^16-18 Furthermore, our research suggests that health plan coverage of multiple ADF opioids (ie, oxycodone HCl CR plus ≥1 ADF opioids) is associated with a reduction in the risk for opioid abuse or overdose. Yet, among the 1.35 million health plan members included in this study who were prescribed Schedule II opioids in 2015 and 2016, fewer than 2% received an ADF opioid during the study period. Expanding access to ADF opioids on drug formulary could be an important strategy to improve the health of the insured population and reduce the clinical and economic burdens imposed by the opioid crisis.

Limitations

This study has several limitations. Instances of opioid abuse or overdose were assessed based on diagnosis codes in medical claims, which may not capture less severe cases or cases that did not result in a medical claim (eg, overdose resulting in death outside of a medical facility setting) and may underestimate the true rate of opioid abuse or overdose, HCRU, and the cost of care.

Similar to other observational analyses using administrative data, although we controlled for known member characteristics when assessing outcomes, this was an observational study and may therefore be subject to unmeasurable confounding. For example, patient socioeconomic status is not available in claims data and could not be controlled in the statistical models.

The study sample was drawn from 135 managed care plans and may not represent all people who are prescribed opioids in the United States, particularly those who are uninsured.

Furthermore, we did not measure the association between the use of ADF opioids and changes in other health behaviors, such as the use of illegal drugs. In 2019, Powell and colleagues showed that states with higher-than-average rates of opioid use or misuse that adopted the use of abuse-deterrent oxycodone HCl CR had increases in the use of heroin and in the rate of hepatitis C.²²

Conclusion

The study findings suggest that broader plan formulary coverage of ADF opioids is associated with lower rates of opioid abuse or overdose, opioid abuse or overdose-related HCRU, and the associated medical costs. Health plan administrators and policymakers may consider improving access to ADF opioids as an important strategy in fighting the opioid crisis while continuing to provide appropriate access to necessary pain medications.

The adoption of ADF drugs into US health plan formularies has been conservative to date. Large, population-based prospective and retrospective studies of real­world patient populations are needed to measure the full effectiveness, safety, and public health considerations for ADF drugs.

Acknowledgments
We thank Barton Jones, of Avalere Health, for the creation of analytic data files and analysis support, Manasi Suryavanshi for analysis support, and Anna Newton, of Avalere Health, for project management support.

Funding Source
This study was funded by Daiichi Sankyo, Inc.

Author Disclosure Statement
Ms Petrilla is an employee of Avalere Health; Ms Marrett is an employee of Daiichi Sankyo; Dr Shen has received research/grant support from Daiichi Sankyo; and Dr Kwong is an employee of Daiichi Sankyo. Dr Pezalla has no conflicts of interest to report.

References

1. Chou R, Fanciullo GJ, Fine PG, et al; for the American Pain Society–American Academy of Pain Medicine Opioids Guidelines Panel. Clinical guidelines for the use of chronic opioid therapy in chronic noncancer pain. J Pain. 2009;10:113-130.e22.
2. Dowell D, Haegerich TM, Chou R. CDC guideline for prescribing opioids for chronic pain–United States, 2016. JAMA. 2016;315:1624-1645.
3. Hanlon JT, Backonja M, Weiner D, Argoff C. Evolving pharmacological management of persistent pain in older persons. Pain Med. 2009;10:959-961.
4. Centers for Disease Control & Prevention. 2018 Annual Surveillance Report of Drug-Related Risks and Outcomes: United States. August 31, 2018. www.cdc.gov/drugoverdose/pdf/pubs/2018-cdc-drug-surveillance-report.pdf. Accessed November 7, 2019.
5. Substance Abuse and Mental Health Services Administration. National Survey on Drug Use and Health (NSDUH-2006). www.datafiles.samhsa.gov/study/national-survey-drug-use-and-health-nsduh-2006-nid13554. Accessed November 7, 2019.
6. Ahrnsbrak R, Bose J, Hedden SL, et al; for the Substance Abuse and Mental Health Services. Key Substance Use and Mental Health Indicators in the United States: Results from the 2016 National Survey on Drug Use and Health. HHS Publication No SMA 17-5044, NSDUH Series H-52. September 2017. www.samhsa.gov/data/sites/default/files/NSDUH-FFR1-2016/NSDUH-FFR1-2016.htm#illicit4. Accessed November 7, 2019.
7. Centers for Disease Control & Prevention. Opioid overdose. Prescription opioids: overdose death maps. www.cdc.gov/drugoverdose/data/prescribing/overdose-death-maps.html. Accessed January 14, 2020.
8. Scholl L, Seth P, Kariisa M, et al. Drug and opioid-involved overdose deaths—United States, 2013–2017. MMWR Morb Mortal Wkly Rep. 2019; 67:1419-1427.
9. Weiss AJ, Elixhauser A, Barrett ML, et al. Opioid-related inpatient stays and emergency department visits by state, 2009-2014. HCUP Statistical Brief #219. December 2016. Agency for Healthcare Research and Quality, Rockville, MD. https://hcup-us.ahrq.gov/reports/statbriefs/sb219-Opioid-Hospital-Stays-ED-Visits-by-State.jsp. Accessed January 15, 2020.
10. Vivolo-Kantor AM, Seth P, Gladden RM, et al. Vital Signs: trends in emergency department visits for suspected opioid overdoses—United States, July 2016–September 2017. MMWR Morb Mortal Wkly Rep. 2018;67:279-285.
11. Califf RM, Woodcock J, Ostroff S. A proactive response to prescription opioid abuse. N Engl J Med. 2016;374:1480-1485.
12. US Food and Drug Administration. Abuse-deterrent opioids—evaluation and labeling: guidance for industry. April 2015. www.fda.gov/regulatory-infor mation/search-fda-guidance-documents/abuse-deterrent-opioids-evaluation-­and-labeling. Accessed November 7, 2019.
13. US Food and Drug Administration. FDA opioids action plan. April 26, 2018. www.fda.gov/drugs/information-drug-class/fda-opioids-action-plan. Accessed November 7, 2019.
14. Pergolizzi JV Jr, Raffa RB, Taylor R Jr, Vacalis S. Abuse-deterrent opioids: an update on current approaches and considerations. Curr Med Res Opin. 2018;34:711-723.
15. US Food and Drug Administration. FDA analysis of long-term trends in prescription opioid analgesic products: quantity, sales, and price trends. March 1, 2018. www.fda.gov/media/111695/download. Accessed November 7, 2019.
16. Institute for Clinical and Economic Review. Abuse-deterrent formulations of opioids: effectiveness and value: final evidence report. August 8, 2017. https://icer-review.org/wp-content/uploads/2016/08/NECEPAC_ADF_Final_Report_08_08_17.pdf. Accessed November 7, 2019.
17. Descoteaux A, Borrelli E, Kogut S. A budget impact analysis of abuse deterrent opioid formulation. Value Health. 2017;20:A295. Abstract PMH22.
18. White AG, Birnbaum HG, Rothman DB, Katz N. Development of a budget-impact model to quantify potential cost savings from prescription opioids designed to deter abuse or ease of extraction. Appl Health Econ Health Policy. 2009;7:61-70.
19. Yenikomshian MA, White AG, Carson ME, et al. Modeling the potential impact of abuse-deterrent opioids on medical resource utilization. J Med Econ. 2019;22:1073-1079.
20. Ahmad FB, Escobedo LA, Rossen LM, et al. Provisional drug overdose death counts. National Center for Health Statistics; 2019. www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm. Accessed November 7, 2019.
21. US Food and Drug Administration. Abuse-deterrent opioid analgesics. June 11, 2019. www.fda.gov/drugs/postmarket-drug-safety-information-patients-and-­providers/abuse-deterrent-opioid-analgesics. Accessed November 7, 2019.
22. Powell D, Alpert A, Pacula RL. A transitioning epidemic: how the opioid crisis is driving the rise in hepatitis C. Health Aff (Millwood). 2019;38:287-294.