Bernacki, Edward J. MD, MPH; Yuspeh, Larry BA; Lavin, Robert MD, MS; Tao, Xuguang (Grant) MD, PhD
The therapeutic use of opioids has increased dramatically in the United States, as evidenced by the 127% rise in retail sales of opioids between 1997 and 2007.1 This increase in opioid use is related to the significant expansion of opioids to treat chronic non–cancer-related musculoskeletal pain.2–8 Data from the US National Ambulatory Medical Care Survey indicated that the frequency of opioid prescriptions to treat chronic musculoskeletal pain doubled from 8% of visits in 1980 to 16% of visits in 2000, whereas the use of opioids to treat acute pain increased 38% (8% to 11% of visits) over this time frame.4 Other studies confirm an increase in the proportion of individuals placed on opioids to treat chronic pain.4,5 Along with this increase was an increase in costs and a growing reliance on the part of medical providers to use stronger opioids as well as long-acting (LA) opioids in their treatment plans.1,7,8
The use of opioids to treat acute and chronic pain associated with work-related conditions is related to the high prevalence of musculoskeletal injuries.9 A Workers Compensation Research Institute study of 16 states found that 26% of the lost-time (LT) claims had at least one opioid prescription associated with it.10 A similar proportion was observed among California workers' compensation claimants with back conditions occurring between January 2002 and November 2005.11 A study in Washington state found that 42% of workers' compensation claimants suffering from LT back injuries were prescribed an opioid within a year of their injury.12
Using National Council on Compensation Insurance data, Lipton and colleagues9 found that the proportion of claimants prescribed opioids for pain for treatment within 12 months of injury increased 75% between 1999 and 2004. However, there was no increase in the proportion that used opioids to treat injuries between the 12th and 36th month among claimants.9 Franklin and colleagues13 reported that prescriptions for schedule II opioids as a percentage of all scheduled opioid prescriptions increased from 19.3% in 1996 to 37.2% in 2002. The average daily morphine-equivalent dose during this time frame increased to 132 mg (50%) per day.13 In contrast, Gross and colleagues14 studying Canadian workers found that opioid prescriptions within the first year of a claim decreased from 11.4% of claimants to 8.3%.
The proportionate and absolute use of opioids in the treatment of work-related injuries varies considerably from state to state.10 The average annual cumulative dose of opioids prescribed for nonsurgical claims with more than 7 days of LT was the second highest in Louisiana (equivalent to 3513 mg of morphine per claim), second only to New York, with 4040 mg per claim.10 By contrast, the annual cumulative dose of opioids per claim in most study states was between 1000 and 2000 mg of morphine equivalence or less.
The annual cumulative dose and cost of opioids used to treat non–work-related conditions for both acute and chronic pain have increased considerably, and trends in opioids prescribed to treat pain associated with occupational injuries have been studied as a new focus.10,14 We were interested in investigating the annual cumulative dose and cost of opioids per claim prescribed to treat work-related injuries in the state of Louisiana for both acute and chronic pain. Furthermore, we wanted to determine the use and cost of short-acting (SA) (immediate-release) and LA (methadone or controlled-release) opioid medications. To study these questions, we utilized workers' compensation claims information from the state of Louisiana paid by the Louisiana Workers' Compensation Corporation (LWCC) over an 11-year period, 1999 to 2009. LWCC is a private mutual insurance company writing workers' compensation insurance for approximately 30% of the fully insured market in the state of Louisiana. Several papers have been published by these authors utilizing the same population to study various workers' compensation cost–related topics.15–17
MATERIALS AND METHODS
As indicated earlier, this investigation utilized data from the LWCC. Information on all workers' compensation claims administered by the LWCC resides in the LWCC Claims Payment Database (CPD). All LWCC claims filed from 1992 to 2009 and opened between 1999 and 2009 were used as the population for this study. Information on prescription drugs was obtained from LWCC's Pharmacy Benefit Manager (PBM). A file, termed the PBM Database (PBMD), was constructed using prescription information. This information was linked to claims of all injury years that were open at some point between 1999 and 2009. The CPD information included demographic data (age, sex, etc) and injury data (date of injury, body part, International Classification of Diseases, Ninth Revision code, etc). In addition, the CPD file contained information on whether or not the claimant lost time from work (an LT claim) or only received medical care, but did not lose time from work (a medical only [MO] claim), as well as claim costs and the claim closing dates. The PBMD included the date of the prescription, National Drug Code for each prescription, number of prescriptions, and the number of pills per prescription between 1999 and 2009.
The selection of analysis cohort of claims and claim duration range combination are shown in Table 1. To control for the possible bias due to claim aging, we restricted the observation to only claims with a claim age of 7 years or less because 96% of LWCC claims close during the first 7 years after injury. The observed prescription period was 1999 to 2009. Using this strategy, as shown in Table 1, we would always have a comparable mixture of claim durations (0 to 7 years) when we examined the usage over the calendar period from 1999 to 2009. The data for each prescription year is a snapshot of the claims inventory at LWCC for any open claims that are 7 years or less in duration.
We further split the claims into two groups: (1) those claims for which opioids we prescribed during the year of the accident (0 year), representing treatment for acute pain and (2) those for which opioids were prescribed during the period after the first accident year up to 7 years after the accident (1 to 7 years), representing treatment for chronic pain.
Opioids were defined as morphine-like medications that are naturally occurring, semisynthetic, or wholly synthetic substances utilized to control pain. All National Drug Codes that fit this definition were abstracted from the PBMD and were consolidated into two groups: SA (immediate release) and LA (methadone or controlled release). Because all opioids do not have the same analgesic effect, we converted the annual cumulative dose of the individual opioids into an equianalgesic dose utilizing morphine as the index. This was termed the morphine milligram equivalent (MME). Except as otherwise indicated, the opioid conversions were based on the GlobalRPh.com, an Internet opioid converter.18 The fentanyl transdermal patch conversion was based on the Duragesic package insert with the mid-dose for the 25 μg/hour fentanyl patch of 100 mg (60 to 134 mg) morphine equivalents chosen for the conversion.19–21 Conversions for tapentadol, sublingual buprenorphine, fentanyl lozenge, propoxyphene, pentazocine, merperidine, and methodone were based on Fudin and Perkins.19,21 The book by Cousins and colleagues was used to determine the relative potency of tramadol.22 For the purposes of this article, the term dose refers to the annual cumulative dose of opioids prescribed per claim.
The MME per prescription was calculated on the basis of the following formula: D = c × d × r, where D is the MME, c is the number of pills in the prescription, d is the dose (mg) of an opioid drug in a pill, and r is the conversion ratio of morphine for the specific opioid medication. Example 1: to calculate MME for 40 pills of acetaminophen/codeine 300/30 mg, where c = 40, d = 30 mg codeine, and r = 0.15 mg morphine; then MME would be: D = 40 × 30 × 0.15 = 180 mg. Example 2: to calculate the cumulative MME for 10 patches of transdermal fentanyl at 50 μg/hour, where c = 10, d = 0.05 mg/hour, and r = 12,000; then MME would be: D = 10 × 0.05 × 12,000 = 6000 mg. Each patch is supposed to be changed every 3 days; therefore, 10 patches will last 30 days. This is equivalent to a person using 200 mg of morphine equivalent daily, multiplied by 30 days = 6000 mg in a month.
Analyses describe the MME per claim per calendar year by claim age (0 year, 1 to 7 years) for prescription years 1999 through 2009. A linear analysis method was used to estimate the average annual change of MME and average cost. In the model, average MME annual cumulative dose or the annual cost per claim is the dependent variable, whereas the independent variable is the year after 1999 with 1999 as the initial year. The reason for using a linear regression instead of nonlinear regression was to obtain an annual change in the variables for the study period rather than strictly fit the trend lines that may have random fluctuations.23 The regression and significance testing were performed using Microsoft Excel 2007 (Microsoft Corporation, Redmond, WA) data analysis tools. We then analyzed the MME and opioid medication costs per claim per calendar year by the type of opioid utilized in a claim and by claim age (0 year, 1 to 7 years). The claims defined as LA contained LA opioids with or without SA opioids, which frequently are prescribed for breakthrough or activity-related pain for patients already receiving LA opioids. However, the claims defined as SA only contained exclusively SA (immediate-release) opioids. Cost per MME was analyzed by type of opioid claims, LA or SA, for the period from 1999 to 2009. The annual change trends for the latter analysis also were simulated using linear regression.
The study file contained 80,159 unique claims that were open or still open at some point during the 11-year study period from 1999 to 2009. The total opioid prescriptions included for all claims with claim age at 7 years or less are 210,413: 67.7% men with a median birth year in 1964 and 32.3% women with a median birth year in 1962. Opioid prescriptions during the year of the accident represent prescriptions for acute pain, whereas opioid prescriptions after the accident year through 7 years represent prescriptions for chronic pain. Table 2 indicates that the mean claim duration for the short-duration LT and MO opioid claims was 99 and 59 days duration, respectively. These claim durations and, therefore, the prescriptions written during this time frame fit a common definition for the treatment of acute pain (<6 months or 180 days).24,25 The mean claim durations for the long-duration LT and MO opioid claims were 1164 and 1028 days, respectively, which conform to the definition of treatment times for chronic pain (≥6 months or 180 days). Because the difference in claim duration varied little between LT and MO claims for both claim categories (0 year, 1 to 7 years), we merged MO and LT claims to increase the size of the study cohorts.
As seen in Figure 1, the cumulative MME per claim per calendar year significantly increased over the study period for claimants treated for acute pain (0-year claims: P = 0.0025) and chronic pain (1- to 7-year claims: P = 0.0058). The cumulative MME increased approximately 55 mg per claim per calendar year for acute pain, whereas for chronic pain it increased 461 mg per claim per calendar year. As seen in Figure 2, the cost of opioid medications per claim per calendar year increased significantly for individuals treated for chronic pain (approximately $23 per claim per calendar year) (P = 0.0398) but remained relatively flat for individuals treated for acute pain (approximately $100 per claim per calendar year throughout the study period). The cost change for acute pain was not significant (P = 0.6783).
Figure 3 presents changes in the annual cumulative dose of opioids prescribed per claim per calendar year for two claim categories: one group for which a LA opioid was prescribed with or without SA opioids and another group for which only SA opioids were prescribed. The cumulative MME increased significantly for acute pain claimants (36 mg per claim per calendar year; P = 0.0084) and chronic pain claimants (233 mg per claim per calendar year; P = 0.0045) taking only SA opioids. Cumulative MMEs increased for chronic pain claimants prescribed LA opioids with or without SA opioids from 22,386 mg per claim per calendar year in 1999 to 54,396 mg per claim per calendar year in 2004, dropping down to 38,397 mg per claim per calendar year in 2009 (P = 0.1069). Claimants treated for acute pain with predominantly LA opioids increased approximately 431 mg per calendar year, which was not a significant increase (P = 0.2020). However, similar to the claimants with chronic pain treated with LA opioids, the claimants treated for acute pain with LA opioids increased 190% in 2003 and then dropped 57% the following year and leveled off thereafter.
As shown in Figure 4, the monies expended on claimants with chronic pain who were prescribed predominantly LA opioids increased an average of $112 per claim per calendar year (P = 0.070). The annual dose of opioids prescribed and the cost of opioids for claimants with chronic pain increased 164% during the period from 1999 to 2004 and then decreased 25% over the next 5 years. The opioid medication expenditures for claimants with acute pain treated with LA opioids increased $17 per calendar year over the study period. This increase was not significant (P = 0.5026). Nevertheless, similar to the claimants with chronic pain treated with LA opioids, claimants treated for acute pain with LA opioids increased 176% from 1999 to 2003, then dropped 132% in 2004 and leveled off thereafter (P = 0.4210). There were no significant increases in the costs for claimants treated with only SA opioids for acute or chronic pain over the study period (P = 0.7395).
Figure 5, shows that over the study period, the cost per MME decreased significantly for SA opioids, from $0.7 per milligram to approximately $0.6 per milligram (P = 0.0007), whereas the cost per MME of LA opioids remained close to $0.6 per MME until 2009, when it increased to a little less than $0.7 per MME in the last year of the study. The trend was not significant (P = 0.3665).
The average annual opioid cost for claims involving an LA opioid to treat both acute and chronic pain was approximately 8 times higher than claims involving only SA opioids (Table 3). Thus, although the cost of opioids per MME varied little between LA and SA opioids, the opioid cost per claim per calendar year of claims involving LA opioids was much greater than claims where only SA opioids were used. Although the data are not presented, approximately 85% of claims involving LA opioids to treat both acute and chronic pain also had SA opioids prescribed. Hydrocodone accounted for 60% of the dosage (in MMEs) of SA opioids, followed by oxycodone (16%) and propoxyphene (16%). Sustained-release oxycodone accounted for approximately 51% of the dosage of LA opioids followed by transdermal fentanyl (33%) and methadone (10%).
Chronic pain commonly is defined as pain that lasts longer than 3 to 6 months and/or pain that persists beyond the normal time for tissue healing.24,25 The mean claim duration for claims involving opioids in the year of the injury for both MO and LT claims was less than 3 to 6 months (MO, 59 days; LT, 99 days). For claims that were open past the year of injury to 7 years, the average claim duration was 1164 days for 1 to 7 years LT and 1028 for 1 to 7 years MO. Therefore, we feel that studying prescription use of short-duration claims approximates prescription use for acute pain, and prescription practices for long-duration claims, chronic pain. A somewhat similar strategy, but of differing time periods, has been used by other authors studying opioid-prescribing practices in workers' compensation.11,26,27
Our study indicated that the annual cumulative dose of opioids prescribed for both acute and chronic pain associated with workplace injuries in the state of Louisiana increased significantly. In each of the study years, the annual cumulative dose of opioids prescribed to treat chronic pain was approximately four to seven times greater than the annual cumulative dose used to treat acute pain. This was related to increasing doses of opioids prescribed for claims defined as LA opioid claims (with or without SA opioids) in the treatment of chronic pain. In our investigation, we found that controlled-release oxycodone and transdermal fentanyl accounted for 84% of cumulative MMEs involving LA opioids. These findings are consistent with other studies from a comparable time period that report greater increases in opioids utilized in LA preparations than used in SA or immediate-release opioid medications.1,7,8 It also may be reflective of the growing perception among Louisiana physicians that LA opioids have advantages, such as consistent dosing schedules and sustained analgesia, over immediate-release preparations. What we found interesting was that the cost per MME for both LA and SA opioids was approximately the same ($0.6 to $0.7). Nevertheless, the expense for claims utilizing LA medications was eight times higher for individuals treated for chronic pain than claims involving only SA medications in treating what we defined as chronic pain. This suggests that once a decision has been made to utilize LA medications, the annual cumulative dose of opioids prescribed increases dramatically. It was noted that hydrocodone accounted for the majority of SA opioids prescribed, probably due to ease of prescription because it is a schedule III opioid.
Systemic reviews of opioid use for chronic back pain and chronic noncancer pain provides little or no evidence of the effectiveness of utilizing opioids on a long-term basis to reduce pain and improve functional status.1,6–8 The increases in the annual cumulative dose of opioids used and the reliance on LA opioids to treat chronic pain takes place despite the evidence that their use does not decrease pain nor increase function for injured workers. In fact, in workers' compensation claimants prescribed opioids, there is an increased risk of delayed return to work.11,26
There were 1642 claimants who lost time and were treated for chronic pain and 691 claimants who lost time and were treated for acute pain in our study. Louisiana is one of the few US states that uses the wage-loss method in calculating indemnity benefits. Temporary/total benefits are paid until the injured employee returns to work, at which time the claim is closed, except for the few claimants who are paid supplemental benefits after they returned to work (if they return to a lesser-paying job).28 Therefore, an LT claimant in our study remained out of work for the entire duration of the claim. This differs from studies performed in non–wage-loss states where a claim is classified as an LT claim if the individual loses enough time from work to meet the definition of eligibility for temporary/total benefits. This difference in the way temporary/total is calculated in Louisiana versus the majority of other states may partially explain the relatively high annual cumulative dose of opioid use in the state of Louisiana versus other states.10 It may also explain some of the differences in temporal trends between our study and studies by others in non–wage-loss states.9,13
In December 2004, the LWCC adopted a preferred drug list (PDL) to guide health care providers to use efficacious and cost-effective opioids and other medications for their patients.29 The PDL listed 3 tiers of medication. Tier 1 represented medications deemed to be first-line medications that did not require prior authorization. Tiers 2 and 3 required the prescriber to complete a prior authorization form before the medication could be dispensed. Tier 1 opioids included codeine, hydrocodone, morphine, oxycodone, propoxyphene, tramadol, and hydromorphone. The imposition of the PDL seems to have had an effect on the dose and types of opioids prescribed, as well as the cost. The PDL undoubtedly constrained LA opioid use as evidenced by a sharp reduction in cumulative MMEs of LA opioids between 2004 and 2006. Morden and colleagues30 also observed a significant decrease in controlled-release oxycodone use after the imposition of prior authorization in a Medicaid population. We performed a Joinpoints analysis (SEER Surveillance Epidemiology and End Point, National Cancer Institute, Bethesda, MD) on the opioid costs of 1- to 7-year claims and found that the slope was +72 between 1999 and 2004 and was −26 from 2004 to 2009.31 This lends support to the notion that the PDL did alter the rate of growth of medications and their cost in the treatment of chronic pain and perhaps acute pain. After the PDL was adopted, there was a significant decrease in the amount paid per MME for SA opioids, presumably because more generic medications were used after its adoption. Tier 1 medications did not include any brand-name formulations, which would have the effect of decreasing the unit cost of the SA opioids.
In our study, we found that the annual cumulative dose of opioids to treat acute pain averaged around 14 MMEs, whereas the use of LA opioids to treat chronic pain averaged approximately 110 MMEs. It seems that special vigilance should be directed at these individuals because the potential for overdose is significant among individuals at these high dosage levels, with an 8.9 odds ratio of overdose for individuals prescribed 100 mg opioids or more daily.10,13,32
There are a number of limitations in this study. We could study only the annual cumulative dose of medication prescribed per claim rather than the actual daily dosage used by the claimant. It is conceivable that all of the medication that was prescribed may not have been used by the claimants and the dose prescribed not representative of the dose used. Our definition of acute and chronic pain may be challenged. However, we attempted to utilize the common time frames that are employed to define acute and chronic pain. We feel that this approach is justified in performing an epidemiologic study of the type we engaged in here. As indicated, Louisiana is a wage-loss state and, because of this, the magnitude of our findings may differ from non–wage-loss states.
Our study indicates the dose of opioid medication prescribed for chronic and acute pain increased significantly over the study period. However, the primary driver of these increases was related to LA opioids prescribed for chronic pain. Corresponding cost increases were associated with the increase in volume, as the price per MME remained rather constant throughout the study period. What we observed in Louisiana seems to correspond to the increase in opioid use to treat chronic pain in North America.1,2,4,7,8,32 This investigation leads to a number of questions about the use of opioids in workers' compensation. How does the use of opioids change over the duration of a claim? Does the early use of LA opioids influence claim duration? How do opioid prescriptions affect overall claim cost and duration of disability? Does a physician's specialty affect the dose and type of opioids prescribed? We plan to study these and other questions utilizing the same data set we utilized in this study.
1. Manchikanti L, Singh A. Therapeutic opioids: a ten-year perspective on the complexities and complications of the escalating use, abuse, and nonmedical use of opioids. Pain Physician. 2008;11:S63–S88.
2. Centers for Disease Control and Prevention. Adult use of prescription opioid pain medications – Utah, 2008. MMWR Morb Mortal Wkly Rep. 2010;59:153–157.
3. Wilson RD. Analgesic prescribing for musculoskeletal complaints in the ambulatory care setting after the introduction and withdrawal of cyclooxygenase-2 inhibitors. Arch Phys Med Rehabil. 2009;90:1147–1151.
4. Caudill-Slosberg MA, Schwartz LM, Woloshin S. Office visits and analgesic prescriptions for musculoskeletal pain in US: 1980 vs. 2000. Pain. 2004;109:514–519.
5. Olsen Y, Daumit GL, Ford DE. Opioid prescriptions by U.S. primary care physicians from 1992 to 2001. J Pain. 2006;7:225–235.
6. Martell BA, O'Connor PG, Kerns RD, et al. Systematic review: opioid treatment for chronic back pain: prevalence, efficacy, and association with addiction. Ann Intern Med. 2007;146:116–127.
7. Trescot AM, Glaser SE, Hansen H, Benyamin R, Patel S, Manchikanti L. Effectiveness of opioids in the treatment of chronic non-cancer pain. Pain Physician. 2008;11:S181–S200.
8. Trescot AM, Helm S, Hansen H, et al. Opioids in the management of chronic non-cancer pain: an update of American Society of the Interventional Pain Physicians' (ASIPP) Guidelines. Pain Physician. 2008;11:S5–S62.
10. Wang D, Muellwe K, Hashimoto D. Interstate Variations in Use of Narcotics. Cambridge, MA: WCRI; 2011.
11. Swedlow A, Gardner LB, Ireland J, Genovese E. Pain Management and the Use of Opioids in the Treatment of Back Conditions in the California Workers' Compensation System, CWCI report to the industry. San Francisco, CA: CWCI; June 2008.
12. Franklin GM, Rahman EA, Turner JA, Daniell WE, Fulton-Kehoe D. Opioid use for chronic low back pain: a prospective, population-based study among injured workers in Washington state, 2002–2005. Clin J Pain. 2009;25:743–751.
13. Franklin GM, Mai J, Wickizer T, Turner JA, Fulton-Kehoe D, Grant L. Opioid dosing trends and mortality in Washington State workers' compensation, 1996–2002. Am J Ind Med. 2005;48:91–99.
14. Gross DP, Stephens B, Bhambhani Y, Haykowsky M, Bostick GP, Rashiq S. Opioid prescriptions in Canadian workers' compensation claimants: prescription trends and associations between early prescription and future recovery. Spine (Phila Pa 1976). 2009;34:525–531.
15. Bernacki EJ, Tao XG. The relationship between attorney involvement, claim duration, and workers' compensation costs. J Occup Environ Med. 2008;50:1013–1018.
16. Bernacki EJ, Tao XG, Yuspeh L. An investigation of the effects of a healthcare provider network on costs and lost time in workers' compensation. J Occup Environ Med. 2006;48:873–882.
17. Bernacki EJ, Yuspeh L, Tao X. Determinants of escalating costs in low-risk workers' compensation claims. J Occup Environ Med. 2007;49:780–790.
20. PriCara. Duragesic. 2009. Available at: http://www.duragesic.com/prescribing-information.html
. Accessed January 12, 2012.
22. Cousins MJ, Bridenbaugh PO, Carr DB, eds. Neural Blockade In Clinical Anesthesia and Pain Medicne, 4th ed. Philadelphia, PA: Lippincott, Wilkins and Williams; 2009:1175.
23. Gordis L. Epidemiology. Philadelphia, PA: Saunders Book Company; 2008.
24. Bonica J. The Management of Pain. Philadelphia, PA: Lea & Febinger; 1953.
25. Merskey H, Bogduk N. Classification of Chronic Pain. Seattle, WA: IASP Press; 1994.
26. Franklin GM, Stover BD, Turner JA, Fulton-Kehoe D, Wickizer TM. Early opioid prescription and subsequent disability among workers with back injuries: the Disability Risk Identification Study Cohort. Spine (Phila Pa 1976). 2008;33:199–204.
27. Webster BS, Verma SK, Gatchel RJ. Relationship between early opioid prescribing for acute occupational low back pain and disability duration, medical costs, subsequent surgery, and late opioid use. Spine (Phila Pa 1976). 2007;32:2127–2132.
28. Barth PS. Compensating workers for permanent partial disability. Soc Secur Bull. 2003;65:16–23.
29. LWCC. Louisiana Workers' Compensation Commission Preferred Drug List. Baton Rouge, LA: LWCC; 2004.
30. Morden NE, Zerzan JT, Rue TC, et al. Medicaid prior authorization and controlled-release oxycodone. Med Care. 2008;45:573–580.
31. Kim HJ, Fay MP, Feuer EJ, Midthune DN. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med. 2000;19:335–351. correction: 2001;20:655.
32. Dunn KM, Saunders KW, Rutter CM, et al. Opioid prescriptions for chronic pain and overdose: a cohort study. Ann Intern Med. 2010;152:85–92.
This article has been cited 1 time(s).
Current Medical Research and OpinionUtilization of duloxetine and celecoxib in osteoarthritis patientsCurrent Medical Research and Opinion
©2012The American College of Occupational and Environmental Medicine