Background:
Multiple state, federal, and private initiatives have decreased the number of individuals treated with opioids and number of prescriptions filled. However, the decline in prescriptions was greater than the decline in total opioid volume (3), and the decline’s magnitude varied by patient age, sex, and prescriber’s specialty (1, 2).
Objective:
To examine changes in total opioids filled at retail pharmacies by patient, prescriber, and county characteristics over the decade including oxycontin reformulation in 2010 and the prescribing peak in 2011.
Methods
Using IQVIA Prescription data (4) capturing approximately 90% of prescriptions at U.S. retail pharmacies, we used days’ supply and total daily opioid dose to calculate per-capita morphine milligram equivalents (MME) for opioid prescriptions filled in 2008-2009 and 2017-2018 overall and by patient age, sex, payer and characteristics of the prescriber’s county, including: race/ethnicity composition; fatal overdose rate; and county urbanicity. For prescribers, we calculated MME per practicing clinicians in each specialty. We used SAS 9.4 for the analyses.
Findings
Total opioid MME volume per-capita filled in retail pharmacies decreased 21.2%, (202.2 MME) from 951.4 MME in 2008-2009 to 749.3 MME in 2017-2018 (Table 1). The greatest percentage decrease was in prescriptions paid by commercial insurance (263.9 MME, 41.5% decrease. Individuals age 46-55 had the greatest decline in MME per-capita (664.6 MME, 33.7% decrease); individuals 18-25 years had the greatest percentage decline (140.2 MME, 66.6% decrease). MME per-capita among individuals age 56-65 was essentially unchanged (1.0 MME; 0.1% decrease); prescriptions paid for by Medicare had the largest MME per-capita decrease (485.3 MME; 17.5% decrease). Per-capita MME declined the most in metropolitan counties (219.8 MME; 22.6%), and in counties in the quartile with the highest fatal overdose rate (473.8 MME; 34.6%). There was substantial variation both within and across states (Figure 1): In some states MME per-capita increased in multiple counties; in many other states, there were counties with both substantial increases and substantial decreases. There was no clear pattern in change of total opioid MME per-capita based on the percentage of non-White county residents (Table 1).
Table 1.
Changes in opioid MME per-capita and per practicing clinician in filled prescription in 2008-2009 and 2017-2018
| 2008-2009 | 2017-2018 | Decrease in MME per-capita from 2008-2009 to 2017- 2018 |
% Decrease in MME per-capita from 2008-2009 to 2017- 2018 |
|
|---|---|---|---|---|
| MME per-capita | MME per-capita | |||
| Total a | 951.4 | 749.3 | 202.2 | 21.2 |
| Sex a | ||||
| Male | 857.9 | 682.4 | 175.5 | 20.5 |
| Female | 1042.6 | 781.0 | 261.6 | 25.1 |
| Age Group a | ||||
| 12-17 | 41.5 | 20.4 | 21.2 | 51.0 |
| 18-25 | 210.7 | 70.4 | 140.2 | 66.6 |
| 26-35 | 645.0 | 277.6 | 367.5 | 57.0 |
| 36-45 | 1188.0 | 714.5 | 473.5 | 39.9 |
| 46-55 | 1970.7 | 1306.2 | 664.6 | 33.7 |
| 56-65 | 1865.5 | 1864.5 | 1.0 | 0.1 |
| 66+ | 1586.8 | 1397.0 | 189.9 | 12.0 |
| Primary payer type a | ||||
| Commercial | 636.5 | 372.6 | 263.9 | 41.5 |
| Medicaid | 646.8 | 467.7 | 179.1 | 27.7 |
| Medicare | 2780.2 | 2294.9 | 485.3 | 17.5 |
| County characteristics | ||||
| % black residents b | ||||
| 1st quintile (lowest) | 666.24 | 593.42 | 72.8 | 10.9 |
| 2nd quintile | 928.39 | 767.34 | 161.0 | 17.3 |
| 3rd quintile | 937.12 | 721.82 | 215.3 | 23.0 |
| 4th quintile | 964.13 | 735.97 | 228.2 | 23.7 |
| 5th quintile (highest) | 976.76 | 798.15 | 178.6 | 18.3 |
| % Hispanic residents b | ||||
| 1st quintile (lowest) | 895.22 | 729.23 | 166.0 | 18.5 |
| 2nd quintile | 1,014.32 | 817.32 | 197.0 | 19.4 |
| 3rd quintile | 1,090.93 | 904.60 | 186.3 | 17.1 |
| 4th quintile | 1,037.74 | 856.98 | 180.8 | 17.4 |
| 5th quintile (highest) | 860.17 | 641.79 | 218.4 | 25.4 |
| % other minority residents b | ||||
| 1st quintile (lowest) | 581.47 | 510.86 | 70.6 | 12.1 |
| 2nd quintile | 803.69 | 649.63 | 154.1 | 19.2 |
| 3rd quintile | 912.69 | 846.33 | 66.4 | 7.3 |
| 4th quintile | 1,140.60 | 882.09 | 258.5 | 22.7 |
| 5th quintile (highest) | 926.80 | 711.95 | 214.8 | 23.2 |
| County fatal overdose rate c | ||||
| 1st quartile (lowest) | 505.70 | 529.22 | −23.5 | −4.7 |
| 2nd quartile | 712.10 | 578.01 | 134.1 | 18.8 |
| 3rd quartile | 1,012.24 | 810.62 | 201.6 | 19.9 |
| 4th quartile (highest) | 1,369.74 | 895.89 | 473.8 | 34.6 |
| Urbanicity d | ||||
| Metropolitan | 971.66 | 751.83 | 219.8 | 22.6 |
| Non-Metro Adjacent | 778.44 | 686.16 | 92.3 | 11.9 |
| Non-Metro Non-Adjacent | 958.39 | 832.17 | 126.2 | 13.2 |
| MME per practicing - clinician |
MME per practicing clinician |
Decrease in MME per-clinician from 2008-2009 to 2017- 2018 |
% Decrease in MME per-clinician from 2008-2009 to 2017- 2018 |
|
| Prescriber specialty e | ||||
| Pain specialist | 1,020,808.4 | 863,140.7 | 157,667.7 | 15.4 |
| Adult PCP | 651,489.4 | 390,841.0 | 260,648.4 | 40.0 |
| Surgeon | 220,764.6 | 111,904.4 | 108,860.3 | 49.3 |
| APP (physician assistant/nurse practitioner) | 112,873.9 | 138,459.3 | −25,585.5 | −22.7 |
| Emergency physician | 99,254.5 | 29,234.3 | 70,020.1 | 70.5 |
| Oncologist | 51,731.2 | 20,941.4 | 30,789.8 | 59.5 |
| Psychiatrists | 50,464.3 | 16,533.0 | 33,931.3 | 67.2 |
| Dentist | 22,345.3 | 13,126.1 | 9,219.1 | 41.3 |
| Other | 14,835.3 | 5830.8 | 9,004.5 | 60.7 |
Per-capita numbers for total population were calculated from the U.S. Census Bureau, Population Division, The Census Bureau’s Population Estimates Program (PEP), County Population Estimates. https://www.census.gov/programs-surveys/popest.html Per-capita numbers for sex and age cohorts were calculated from the U.S. Census Bureau, the Population Projections Program, the 2008 National Population Projections. https://www.census.gov/data/datasets/2008/demo/popproj/2008-popproj.html. Per-capita numbers for commercially insured individuals and Medicaid enrollees were calculated using 2008, 2009, 2017 and 2018 insurance population information from Kaiser Family Foundation, State Health Facts, Health Insurance Coverage of the Total Population. https://www.kff.org/other/state-indicator/total-population/ Per-capita numbers for Medicare Part D beneficiaries were calculated using 2008, 2009, 2017 and 2018 information from CMS Program Statistics, Medicare Enrollment Section, Medicare Part D Enrollment.
Percentage of non-Hispanic Black, percentage Hispanic, and percentage non-Hispanic non-Black minority residents were calculated from the Area Health Resources Files (AHRF).
We calculated county overdose rates as the per-capita rate of overdose deaths due to any drug using data from the Centers for Disease Control and Prevention (5).
County urbanicity was categorized as metropolitan using Rural-Urban Continuum Codes (RUCC)) 1, 2, or 3, “non-metropolitan adjacent” (RUCC 4, 6, or 8), or “non-metropolitan non-adjacent” (RUCC 5, 7, or 9).
MME per practicing physician numbers were calculated using 2007 and 2017 physician numbers from the Association of American Medical Colleges (AAMC) Physician Specialty Data Report; MME per practicing APP numbers were calculated using 2007 physician assistant numbers from He, National trends in the United States of America physician assistant workforce from 1980 to 2007. Hum Resour Health. 2009;7:86., 2017 physician assistant numbers from the National Commission on Certification of Physician Assistants (NCCPA), 2017 Statistical Profile of Certified Physician Assistants Annual Report, and nurse practitioners numbers in 2007 and 2018 from American Association of Nurse Practitioners (AANP) Infographic: https://storage.aanp.org/www/documents/NP-Infographic.pdf ; MME per practicing dentists were calculated using 2008, 2009, 2017, and 2018 dentist numbers from the American Dental Association (ADA), Health Policy Institute analysis of ADA masterfile.
Figure 1:
Change in County Per-Capita MME from 2008-2009 to 2017-2018
Opioid MME volume per practicing clinician declined across most specialties. The greatest decline in MME per practicing clinician was among adult PCPs (260,648 MME, 40.0%) and pain specialists (157,667 MME, 15.4%); the greatest percentage decline (70,020.1 MME; 70.5%) was among emergency physicians. Advanced practice providers (APPs) were the exception: the MME per-APP increased by 22.7% (25,585.5 MME).
Discussion
Using national retail pharmacy records, we found that substantial declines in opioid MME per-capita and per-clinician filled at retail pharmacies included large variations by patient, prescriber, and county characteristics. The greatest decrease in MME per practicing clinician was among adult PCPs and pain specialists—clinicians with the highest MME per clinician in 2008-2009. The greatest percentage decrease was among emergency physicians – clinicians likely prescribing opioids predominantly to individuals experiencing acute pain in acute care settings. APP opioid prescribing increased, potentially reflecting their expanded scope of practice in many states; however, APP MME per-clinician remained less than half that of adult PCPs in 2017-2018.
There has been a substantial decline in MME per-capita over the last decade, corresponding with clinical and policymaker efforts to decrease clinically unnecessary opioid prescribing. However, we find substantial variation in changes in MME per-capita by patient, county, and prescriber characteristics, as well as within states. Counties experiencing substantial decreases in per-capita MME often lay adjacent to counties with substantial per-capita increases. These results suggest that the effects of clinician and policymaker efforts to reduce opioid prescribing may have differentially affected different populations, and future efforts to enhance clinically appropriate opioid prescribing may need to be more clinically nuanced and targeted for specific populations.
Acknowledgements
The authors thank Hilary Peterson, B.A. of the RAND Corporation for her feedback and editorial assistance on earlier versions of the manuscript. No additional compensation was received beyond their RAND salary. Dr. Stein and Ms. Sheng had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
This research was supported by grants from the National Institutes of Health R01DA045055, and P50DA046351. The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The authors report no conflicts of interest.
Contributor Information
Bradley D. Stein, RAND Corporation, Pittsburgh PA.
Erin A. Taylor, RAND Corporation, Santa Monica, CA.
Flora Sheng, RAND Corporation, Arlington, VA.
Andrew W. Dick, RAND Corporation, Boston, MA.
Mary Vaiana, RAND Corporation, Santa Monica, CA.
Mark Sorbero, RAND Corporation, Pittsburgh PA.
References
- 1.Tucker J, Salas J, Zhang Z, Grucza R, Scherrer JF. Provider specialty and odds of a new codeine, hydrocodone, oxycodone and tramadol prescription before and after the CDC opioid prescribing guideline publication. Prev Med. 2021;146:106466. [DOI] [PubMed] [Google Scholar]
- 2.Schieber LZ, Guy GP Jr., Seth P, Losby JL. Variation in Adult Outpatient Opioid Prescription Dispensing by Age and Sex - United States, 2008-2018. MMWR Morb Mortal Wkly Rep. 2020;69(11):298–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Goldman AH, Johnson DD, Griffis CE, Land V, Balazs GC. Opioid prescribing in the U.S Military Health System, 2014 to 2018: fewer prescriptions, fewer pills, and shorter treatment duration. Pain. In press. [DOI] [PubMed] [Google Scholar]
- 4.IQVIA: Real World Data and Insights. https://www.iqvia.com/solutions/real-world-evidence/real-world-data-and-insights on July 13, 2020.
- 5.Centers for Disease Control Prevention. Drug Overdose Deaths. 2020; https://www.cdc.gov/drugoverdose/data/statedeaths.html on May 20, 2020. [Google Scholar]