Healthcare utilization and costs increased in the 6 months after patients started opioid therapy for chronic pain; they then decreased but never reverted to baseline levels.
ABSTRACT Objectives: To evaluate treatment patterns, healthcare resource utilization, and costs among patients within a large managed care population chronically using opioids for non-cancer pain.
Study Design: Retrospective cohort study.
Methods: Patients aged ≥18 years with ≥1 prescription initiat-ing opioids between January 1, 2007, and December 31, 2011, who also had 12 months of continuous pre-index health plan enrollment, were identified. Patients with pre-index opioid use or cancer diagnosis were excluded. Opioid exposure was stratified by treatment duration—short-term (30-182 days) versus chronic (≥183 days)—and by index opioid type (weak vs strong).
Results: A total of 2.9 million patients initiating opioids were identified, of which 257,602 had at least 30 days of continuous use and were included in the study. The mean age was 51 years and 52% were female. Overall, 239,998 (93%) patients had short-term opioid use, and 17,604 (7%) had chronic use; 215,424 (84%) initiated treatment with a weak opioid, and 44,712 (17%) with a strong opioid. The specialty most associated with the use of less potent opioids was general/family practice (28%), and for more potent opioids it was surgery (22%). Large increases in health-care utilization were reported between the pre-index and first 6-month post initiation periods for chronic users. Utilization rates decreased after the first 6 months but never reverted to baseline levels. Costs mirrored utilization trends, more than doubling between baseline and the first 6 months of treatment for phar-macy ($2029 vs $4331) and all-cause medical ($11,430 vs $27,365). Costs declined after the first 6 months of opioid use but remained above pre-index levels.
Conclusions: These results demonstrated that healthcare resource utilization and costs increased during the first 6 months following clinical scenarios that necessitated opioid initiation and subse-quently declined, suggesting the need to monitor patients beyond the acute care period.
Am J Manag Care. 2015;21(3):e222-e234
The Institute of Medicine estimates that 100 million Americans are affected by chronic pain, with annual direct and indirect economic burdens, accompanied by high social costs, exceeding $600 billion.1 In March 2013, the American Academy of Pain Medicine acknowledged the absence of consensus on the treatment of chronic non-cancer pain, and suggested that opioid therapy is appropriate when conservative approaches are ineffective, as well as when treatment plans are designed to sidestep diversion, abuse and addiction, and serious side effects.2 Currently, opioids are prescribed for approximate-ly 90% of patients with chronic pain in the United States3; about 90% of patients presenting at pain management cen-ters already receive an opioid.4 Such high and escalating prescription rates are aimed at improving the management of nonmalignant pain, and an evaluation of the manage-ment of chronic and acute non-cancer pain in ambulatory and office-based settings in the United States showed that opioid prescribing almost doubled (from 11.3% to 19.6% of all pain-related visits) between 2000 and 2010.5
Prior studies have produced varied results regarding the effect of opioid therapy use for the treatment of non-can-cer pain. Most studies and clinical trials with opioids have yielded unfavorable results in the short term (eg, 12 weeks), demonstrated greater healthcare resource utilization (HRU), and indicated considerable risk-benefit concerns.6-12 Howev-er, there has been considerably less research regarding long-term outcomes of opioid use on HRU.
Continued opioid treatments have common side effects such as sedation, dizziness, nausea, vomiting, and respiratory distress,3 with the most prevalent and worst tolerated being constipation.3,13 These side effects interfere with treatment adherence, work productivity, and health-related quality of life, and are associated with escalated HRU and costs.3,13,14 Medication abuse15 and inadequate adherence and dosing limits3 stemming from side effects can un-dermine the analgesic goals of pain control, and also result in greater use of emergency department (ED) and outpatient services, along with substantial cost increases.13,16
A handful of studies have reported on the overall trends and characteristics of patients initiating opioid therapy, including long-term use14,17; dosing patterns and length of expo-sure18; the comparison of chronic, acute, and non-opioid users19; and prescribing trends based on differing pain types,20 among other areas. Over-all, there has been little systematic coordination of these disparate aspects in the literature; it is still not known how healthcare costs change over the duration of opioid therapy for pain management. The few studies that ex-amined opioid usage patterns and HRU focused on gas-trointestinal-related complications13; as of now, there has been no longitudinal analysis based on more than 1 year of continuous opioid therapy focused on enrollees in a national database of commercially insured patients.
The objective of this study was to describe patient char-acteristics, treatment patterns, HRU, and costs among patients newly initiating opioid therapy. Subgroups were examined separately: those patients with at least 6 months of opioid exposure (chronic users) and those with 30 to 182 days of opioid use (intermediate-term users), combined with indexing on weak or strong opioids.
METHODS
Data Source and Study Design
This retrospective cohort study, which utilized an in-ception cohort (new users) design,21 queried the Health-Core Integrated Research Environment (HIRE) to identify patients with ≥1 prescription (first fill defined as index date) for opioids during the period January 1, 2007, to December 31, 2011. A repository of more than 40 mil-lion researchable lives, the HIRE contains medical, phar-macy, and other administrative claims data originating in 14 geographically dispersed commercial health plans, yielding coverage across the continental United States. This nonexperimental study, which did not require inves-tigational review board review, complied with all appli-cable provisions of the Health Insurance Portability and Accountability Act. Patient confidentiality was preserved throughout and data remained anonymous; researchers only had access to relevant data sets, from which indi-vidual patient identifiers were purged.
Inclusion/Exclusion Criteria
To be included in the study, patients were required to have ≥1 prescription fill for an opioid during the patient identification period, and ≥30 days of continuous opioid use starting at index, defined as the first prescription fill for an opioid medication. In addition, patients were re-quired to be 18 years or older on the index date and have at least 12 months of pre-index health plan enrollment. Patients with any prescription fill for opioids during the 12-month pre-index period were excluded from the study; also excluded were patients with a diagnosis of cancer during the 12-month pre-index period. A cancer diag-nosis (International Classification of Diseases, Ninth Re-vision, Clinical Modification [ICD-9-CM] diagnosis codes 140.xx-209.3x, 230.xx-234.xx) was based on at least 2 claims on distinctly different dates for the same type of cancer (identified with 3-digit ICD-9-CM diagnosis codes) occurring within 60 days of each other.
Study Measures
Chronic opioid use and patient follow-up. Opioid us-ers were separated into subgroups based on their duration of continuous opioid treatment (30-182 days, considered intermediate-term use; and 183+ days, considered chronic use) and analyzed separately. Opioid treatment was consid-ered continuous if an opioid prescription was filled within 30 days after the prescription date plus the days supply of the previous prescription fill. The duration of therapy was calculated as the number of days from opioid initiation to the date of the last fill considered to be continuous use, plus the number of days supply of that fill. All patients were required to have at least 30 days of opioid use as a means to exclude acute opioid use, while 6 months was used to separate chronic opioid use from that of interme-diate-term ailments such as surgery and fractures, and to provide sufficient time for measuring healthcare utilization and costs. The 30-day and 6-month thresholds have also been used in prior studies of opioid use.13,18
To investigate how healthcare utilization and costs changed for chronic opioid users during an extended peri-od of continuous use (in excess of 6 months) after the initia-tion of opioid therapy, the subgroup of chronic use patients were analyzed in greater detail before and after the 183rd day of opioid treatment. The rationale for this extended analysis is based on the hypothesis that considerable costs could accrue shortly after the initiation of opioids to ad-dress the condition that warranted the opioid use, but may decrease over time. All comorbid conditions were identi-fied via ICD-9-CM diagnosis codes, while pharmacy and medical records were used to determine medication use.
The observation period of all patients included the 12 months prior to opioid initiation through the last day of continuous opioid therapy.
Comorbidities and prior medication use. Use of medi-cation other than opioids was identified during the 12 months prior to opioid initiation. The proportion of pa-tients with at least 1 prescription fill for each of the classes of interest was captured. Comorbid conditions were also identified in the 12 months prior to starting opioid ther-apy, and were based on the presence of at least 1 medical claim including an ICD-9-CM diagnosis code of the con-dition of interest.
HRU and costs. Evaluations of HRU included office, outpatient, ED, and inpatient encounters, as well as the length of stay of inpatient hospitalizations. Costs were calculated using the common health economics measure, per patient per year,22,23 which reflected the total accumu-lated costs divided by number of patient-years of obser-vation. Included cost categories were total healthcare, including inpatient, outpatient, office visit, ED, and phar-macy. Costs were reported by plan paid, patient paid, and total (plan paid + patient paid).
Proportion of days covered (PDC). PDC was used to measure opioid use of medications, and was calculated as the ratio of the number of days covered by any opioid prescription filled during the post index period divided by the number of days from index date to end of follow-up. Days covered included the date an opioid prescription was filled plus the days supply on that prescription minus 1. If 1 day was covered by multiple opioid medications it was only counted once. PDC could range from >0 to 1.0 for patients not taking any medication, and to 1 for pa-tients who had all post index days covered. Only patients with at least 1 post index opioid fill were included in this calculation.
Opioid strength. The mechanism of action of the opi-oid class of compounds relies on the ability to bind to the opioid receptors in the brain, spinal column, and sur-rounding tissues. The strength of opioids, strong or weak, is based on their affinity for binding with the μ, κ, and δ receptors.9 Opioid users were analyzed according to the strength of the index opioid: weak (codeine, dihydroco-deine, hydrocodone, propoxyphene, tapentadol, trama-dol) versus strong or potent (fentanyl, hydromorphone, levorphanol tartrate, meperidine, methadone, morphine, oxycodone, oxymorphone).
Statistical Analysis
Analyses were conducted on new opioid users based on the first observed prescription fill of opioid medication and were stratified based on duration of use (intermediate-term and chronic use). Additionally, chronic opioid users were examined in more detail. Descriptive statistics, in-cluding proportions and means, are reported throughout. Unadjusted bivariate statistical tests (χ2 and t tests) were used to compare patient demographics and opioid use between those initiating strong versus weak opioids and intermediate-term users versus chronic users.
Costs per patient-year, allowing for unequal follow-up times, were reported for chronic users across 3 time points: during the 12-month pre-index period, the first 6 months of therapy, and the period after the first 6 months of therapy. To obtain generalized estimating equations to compare costs across each time point, we used repeated measures gamma regression analyses (to account for with-in-patient correlation and the skewed nature of cost data), weighted for follow-up times.
RESULTS
Data Attrition and Sample Size
From the 21.8 million identified patients, 2.9 million were included in the analysis (21.8%). Of those, a total of 239,998 patients with no prior malignancy had intermedi-ate-term opioid use, and 17,604 patients had chronic use identified. A total of 215,424 patients initiated treatment with a weak opioid, while 44,712 commenced treatment with a strong opioid (2534 initiated a weak and strong opioid simultaneously and were counted within each subgroup).
Patient Demographics and Index Opioid Exposure
More than 250,000 patients had intermediate-term or chronic opioid exposure: 239,998 had intermediate-term use (2.6 ± 1.0 months on average; median = 2.2 months), and 17,604 had chronic opioid use (15.8 ± 11.0 months on average; median = 11.3 months). The duration of follow-up did not differ between weak and strong opioid users (mean = 3.5 vs 3.6 months; median = 2.3 vs 2.4 months, respectively), as shown in Table 1. The overall population was 51 years old on average and had slightly more females (52%). Patients initiating weak opioids were slightly older (mean = 54 years; median = 53 years) with a smaller pro-portion of females (49%), and those initiating a strong opi-oid were slightly younger (mean = 50 years; median = 50 years) with a lower percentage of females (48%). Among all opioid users, 83.6% were indexed on a weak opioid and 17.4% on a strong opioid (1% indexed on both)—propor-tions that remained consistent regardless of the duration of exposure.
Hydrocodone and oxycodone were the most com-monly prescribed weak and strong opioids, respectively, at index in both duration strata. Among all other opioids prescribed at index, the largest difference in the propor-tions between intermediate-term and chronic users was observed for tramadol (18.5% for intermediate-term users vs 24.5% for chronic users). The most common specialties for index prescribers of weak opioids was family medicine (FM) / general practice (GP) (28.3%), while for strong opi-oids, it was surgery (21.6%); evidently, patients were more likely to receive strong opioids following surgery than during a routine office visit. Greater proportions of pa-tients with chronic opioid use were prescribed their index medication by an FM/GP or internal medicine physician compared with short-term users (52.5% vs 39.6%, respec-tively). Almost twice the proportion of intermediate-term users received the prescription for their index medicatio from a surgeon compared with chronic users (15% vs 8.2%, respectively). This may reflect the nature of conditions be-ing treated by the differing specialties.
Baseline Comorbidities and Medication Use by Patient Groups
Figure 1
Figure 2
Common pre-specified comorbid conditions and medication classes were found with greater frequency among chronic opioid users than among intermediate-term us-ers; notable differences were observed for patients with low back pain (27.7% vs 18.6%), cardiovascular disease (51.2% vs 45.9%; and especially noticeable for hyperten-sion), and anxiety (11% vs 7.5%), respectively, as shown in . Small differences in medications use were ob-served between the 2 duration-of-therapy groups; chronic use patients had slightly greater usage of almost every pre-specified medication type (eg, antidepressants, antianxiety agents, anticonvulsants, muscle relaxants), as shown in . The prevalence of comorbid conditions was sim-ilar for patients who initiated a strong opioid versus those initiating weak opioids; however, a substantial difference in proportion was observed for nontraumatic fracture Small differences in medications use were ob-served between the 2 duration-of-therapy groups; chronic use patients had slightly greater usage of almost every pre-specified medication type (eg, antidepressants, antianxiety agents, anticonvulsants, muscle relaxants), as shown in Figure 2. The prevalence of comorbid conditions was sim-ilar for patients who initiated a strong opioid versus those initiating weak opioids; however, a substantial difference in proportion was observed for nontraumatic fracture (13.9 vs 5.2%) and substance abuse (10.1% vs 6.7%). Pa-tients initiating strong opioids, however, had fewer fills for every major medication class during the pre-index pe-riod, possibly indicating an undertreated population for their comorbid conditions.
Post Index Treatment Patterns
For both intermediate-term and chronic users, hydro-codone and tramadol were the most utilized weak opi-oids during follow-up; oxycodone was the most utilized strong opioid. Interestingly, despite low use of all strong opioids at index (excluding oxycodone), the proportion of patients exposed to fentanyl, hydromorphone, and morphine increased during follow-up. This increase was at least 10-fold for chronic users, as shown in Table 2. Of the patients indexing on a weak opioid, 17% and 9.4% were exposed to strong opioids, oxycodone and fentanyl, respectively, during the follow-up. Conversely, of those indexing on a strong opioid, 42% and 9.9% were exposed to weak opioids: hydrocodone and tramadol, respectively, during follow-up. On average, chronic users filled prescriptions for 2.3 different types of opioids (median = 2.0) and had opioid medication on hand for 76% of the follow-up period, while intermediate-term users filled pre-scriptions for 1.7 different types of opioids (median = 1.0) and had opioid medication on hand for just 42% of the follow-up period. The low PDC within the short-term pa-tients’ follow-up likely reflects the acute nature of their treatment. No difference in PDC was seen between users of weak and strong opioids, although initiators of strong opioids filled more types during follow-up (mean = 2.0 vs 1.6 and median = 2.0 vs 1.0, respectively).
HRU Pre- and Post Index
Figure 3
Patients indexing on a strong opioid had more than twice the number of mean prior inpatient hospitalizations compared with patients who indexed on weak opioids (0.48 vs 0.21 visits in the pre-index period, respectively), as shown in. Prior utilization of ED, office, and out-patient visits was also higher. Chronic opioid users had similar HRU prior to initiating treatment compared with intermediate-term users; the only major difference was in mean inpatient length of stay (11.6 days for chronic use patients vs 7.3 days for intermediate-term use patients; medians = 5.0 vs 3.0, respectively).
Post index, HRU increased substantially relative to the pre-index period for the study population—number of inpatient visits: 0.25 pre-index to 0.77 per patient year post index; ED visits: 0.22 to 0.78; office visits: 6.48 to 13.5; and other outpatient visits: 9.70 to 31.9. Additionally, patients indexing on a strong opioid continued to have higher rates of hospitalization, ED, and outpatient visit utilization during the post index period relative to those initiating weak opioids. Chronic opioid users had slightly lower utilization rates during the post index period compared with short-term users; however, the rates were still higher than those observed pre-index. The greater num-ber of visits per patient per year of treatment for interme-diate-term users suggests high utilization rates in the first 6 months while they were exposed to opioids. This hypoth-esis was investigated further using a subgroup analysis of the chronic users.
Subgroup Analysis: Chronic Opioid Users
Figure 4
Chronic users were examined separately to see how HRU and costs change over time while being managed with opioid treatment. HRU during the pre-index period, the first 6 months post index, and the period after the first 6 months post index are shown in . Large increas-es were observed from the pre-index period to the first 6 months after initiation of therapy using opioids, especial-ly outpatient visits. After 6 months, utilization declined but it did not reach pre-index levels.
Table 3 shows the post index treatment patterns with opioids for the first 6 months of treatment versus the pe-riod after the first 6 months. The proportion of patients with each type of opioid was lower during the period immediately after the first 6 months of treatment, even though follow-up was longer (patients were followed for an additional 9.8 months on average [median = 5.3] after the first 6 months). The mean number of opioid prescrip-tion fills was slightly higher in the follow-up period after the first 6 months (14.5 vs 11.2 in the first 6 months); however, this was likely due to the longer follow-up pe-riod and outliers, as the median number of fills was higher during the first 6 months (9.0 vs 6.0 in the period after). More unique types of opioids were filled during the first 6 months after treatment initiation (mean = 1.9; median = 2.0) than in the period after the first 6 months (mean = 1.5; median = 1.0). The PDC was higher during the first 6 months after initiation compared with the follow-up pe-riod beyond 6 months (mean 0.78 vs 0.67; median 0.82 vs 0.73).
The healthcare costs per patient-year prior to, 6 months after, and >6 months after initiation of opioid treatment are shown in Table 4. Not unexpectedly, mean all-cause pharmacy costs doubled from the 12-month period prior to treatment initiation ($2029 ± $3970 per patient per year) to the first 6 months of opioid treatment ($4331 ± $7347) and beyond the first 6 months of therapy ($4890 ± $7937). All-cause medical costs per patient-year also more than doubled from the pre-index period ($11,430) to the first 6 months of treatment ($27,365); costs declined after the first 6 months of treatment ($15,815) but did not return to pre-index levels. While inpatient hospital costs almost doubled from the pre-treatment period ($7911) to the first 6 months of treatment ($12,895), they returned to below pre-treat-ment levels after the first 6 months of treatment ($7500).
DISCUSSION
The results of this study are consistent with prior find-ings in the literature on patient attributes, treatment pat-terns, and HRU and costs for patients initiating opioids for chronic pain.3,13,18,20 Addressing an important gap in the literature, however, we investigated how new initia-tors of opioids accumulated expenses immediately after the start of treatment and how those costs changed over time, which to our knowledge has not been investigated and reported in any prior published study. It was hypoth-esized that costs were likely to decline once the target condition was controlled or managed. By examining sub-groups for extended durations, our findings provide new insights into changes in HRU and costs over time for pa-tients chronically using opioids in a real-world managed care setting.
As expected, comorbid conditions were more common in patients who continued opioid treatments for at least 6 months. We found that among the pain categories, ar-thritis and low back pain were the most common, with low back pain more prevalent in patients chronically us-ing opioids. Overall, cardiovascular disease was the most commonly occurring comorbidity, followed by psychiat-ric disorders; both were most prevalent among patients chronically using opioids.
Opioid use was more substantial in the first 6 months of treatment than after the first 6 months (eg, greater number of different prescriptions filled, total number of prescriptions). At treatment initiation, the use of a strong opioid was less common; 84% of chronic opioid users in-dexed on a weak opioid. However, during the post index period, there was a large increase in use, with more than one-half of the patients filling a prescription for at least 1 strong opioid. Such changes in therapeutic strategy could be a reflection of clinicians’ efforts to better manage pain, the building of opioid tolerance, and adverse effects in tandem. Opioid switching or rotation has been credited with reduced side effects while controlling pain.24,25
Initiators of strong opioids had higher rates of inpa-tient hospitalization during the pre-index period com-pared with those initiating weak opioids, the higher inpatient hospitalization rates possibly being driven by the frequent causes of the administration of strong opi-oids: acute traumatic conditions (eg, severe injury, sur-gery, fracture) rather than chronic conditions. Available evidence suggests that a history of substance abuse was more common among patients initiating strong opioids relative to those taking weak opioids at the inception of treatment: more than 10% of patients initiating strong opioids had a history of substance abuse. Physicians are less likely to prescribe strong opioids to known substance abusers; however, in a number of cases they may have been unaware of certain aspects of their patients’ medi-cal history, including any drug abuse. This problem may be exacerbated by patients’ “doctor shopping,” whereby patients approach multiple physicians, one or several of whom may be persuaded to prescribe opioids for a fabri-cated condition, commonly back pain.26-29 Such patients typically pursue strong opioids for the increased psycho-tropic effects that are generally less obtainable from a weak opioid. Doctor shoppers, however, account for only a small portion of opioid users—reported as less than 1% of oxycodone users in a prior retrospective database study30—and opioid abuse does not always result in doc-tor shopping.31,32
The data used in this study were primarily drawn from prescription fills within employer-provided health plans and did not include data related to cash transactions or illegally purchased drugs. As a result, it was not possible to estimate the full prevalence of opioid abuse in the study population. A prior study, which utilized administrative claims data and expert panels, developed an algorithm and found that potential opioid misuse was present among 0.001% to 0.25% of plan members,33 while a meta-analysis found opioid abuse/addiction to be present among 3.27% of chronic opioid users.34 In the current study, a prior his-tory of drug abuse could not necessarily be interpreted to mean that the patient was abusing or misusing the currently prescribed opioid therapy. Nonetheless, a his-tory of abuse and the risk of addiction are important fac-tors to consider, and they underscore the importance of physician monitoring of patients and their opioid use. Although it was beyond the scope of the current study, additional research may be warranted to explore the pos-sible impact of opioid abuse, supported by the high rates of prior substance abuse and the high prevalence of other commonly abused medication classes—antidepressants and antianxiety medications35,36—found in this study.
This study demonstrated substantive increases in HRU during the first 6 months after opioid initiation, which subsequently declined but never returned to base-line. Healthcare costs followed a similar pattern, spiking early during treatment, then decreasing after 6 months but remaining higher than pre-index levels. Opioid users displayed high morbidity and HRU with some variabil-ity by strata based on type of index opioid and duration of opioid exposure. It could not be determined in this study whether the increase in HRU and resulting costs were due to the opioid treatment itself, the underlying ailment being treated with opioids, or an unrelated con-dition. However, previous studies have suggested that opioid therapy in itself was costly compared with other treatment options. A pair of studies that focused on the cost-effectiveness of various prescription therapies for the treatment of 2 of the most common conditions in our study population, osteoarthritis and low back pain, found that opioids had significantly poorer cost-effective-ness compared with antidepressants and nonsteroidal anti-inflammatory drugs.37,38 Thus, it is likely that the sus-tained increase in costs is not due solely to the underlying condition being treated but also to the opioid therapy— both directly through prescription costs, and indirectly through complications, lack of effectiveness, and other issues related to the treatment.
These findings have implications for the manage-ment of patients receiving opioid therapy and under-score the need for close monitoring beyond the acute care period of the opioid treatment regimen and side ef-fects, as well as the underlying condition being targeted. These insights have clinical and economic implications for stakeholders including patients, clinician providers, and payers.
Limitations
Analyses in this study were unadjusted for potential covariates; further, associations cannot distinguish be-tween cause and effect and results should be interpreted with care. Our primary data—administrative claims— did not provide instructive demographic factors such as ethnicity, education level, family history, smoking rates, alcohol use, and laboratory values. Patients were identified via index prescription, with the pharmacy fill date considered the start date of medication usage, but actual usage could have commenced later. In this study, continuous opioid use required medications to be filled no more than 30 days after the prescribed days supply. However, if patients were taking their medication on an as-needed basis and/or were pill-splitting, a single prescription could last more than 30 days beyond the in-tended days supply, with patients erroneously classified as discontinuing therapy. Also, any generalizing of the findings from this study would need to be limited to simi-lar commercially insured populations.
CONCLUSIONS
Non-acute opioid users displayed high morbidity and HRU and costs with some variability between subgroups based on the type of index opioid and duration of opioid use. Because the clinical and economic impacts of opioid therapy for the management of pain are evident beyond the acute treatment phase, patient management could benefit from a longer time horizon.
Acknowledgments: Bernard B. Tulsi, MSc, provided writing and other editorial support for this manuscript.
Author Affiliations: HealthCore, Inc (DMK, SZ, OT, JS), Wilmington, DE; AstraZeneca Pharmaceuticals LP (SC, MS, RJL), Wilmington, DE.
Source of Funding: AstraZeneca LP sponsored this study. The re-searchers had complete access to the de-identified data set and formu-lated the protocol, study design, and statistical analysis. The researchers had full authority over the administration of the study and over the decision to publish their findings. Researchers from both AstraZeneca and HealthCore were involved in the interpretation of study results and preparation and review of the manuscript prior to submission.
Author Disclosures: Drs LoCasale and Sostek and Mr Chavoshi are employed by AstraZeneca LP and also own stock in the company. Mr Kern and Drs Zhou, Tunceli, and Singer are employed by HealthCore, Inc, which received funding from AstraZeneca LP for this study.
Authorship Information: Concept and design (DMK, SZ, SC, OT, MS, JS, RJL); acquisition of data (DMK, SZ, OT); analysis and interpreta-tion of data (DMK, SZ, SC, OT, MS, JS, RJL); drafting of the manuscript (DMK, MS, JS, RJL); critical revision of the manuscript for important in-tellectual content (DMK, SZ, OT, MS, JS, RJL); statistical analysis (DMK, SZ, OT, RJL); obtaining funding (SC, OT, RJL); administrative, technical, or logistic support (DMK, SC, JS); supervision (DMK, OT, MS, JS, RJL).
Address correspondence to: David M. Kern, MS, HealthCore, Inc, 800 Delaware Ave, 5th Fl, Wilmington, DE 19801-1366. E-mail: dkern@ healthcore.com.
REFERENCES
1. Institute of Medicine. Relieving Pain in America: A Blueprint for Transforming Prevention, Care, Education and Research. Washington, DC: The National Academies Press; 2011.
2. The American Academy of Pain Medicine. Use of opioids for the treatment of chronic pain: a statement from the American Academy of Pain Medicine. http://www.painmed.org/files/use-of-opioids-for-the-treatment-of-chronic-pain.pdf. Published March 2013. Accessed March 17, 2015.
3. Benyamin R, Trescot AM, Datta S, et al. Opioid complications and side effects. Pain Physician. 2008;11(2 supp):S105-S120.
4. Trescot AM, Helm S, Hansen H, et al. Opioids in the management of chronic non-cancer pain: an update of American Society of the Inter-ventional Pain Physicians’ (ASIPP) Guidelines. Pain Physician. 2008; 11(2 suppl):S5-S62.
5. Daubresse M, Chang HY, Yu Y, et al. Ambulatory diagnosis and treat-ment of nonmalignant pain in the United States, 2000-2010. Med Care. 2013;51(10):870-878.
6. Eisenberg E, McNicol E, Carr DB. Opioids for neuropathic pain. Cochrane Database Syst Rev. 2006;(3);CD006146.
7. Eriksen J, Sjøgren P, Bruera E, Ekholm O, Rasmussen NK. Critical is-sues on opioids in chronic non-cancer pain: an epidemiological study. Pain. 2006;125(1-2):172-179.
8. Finnerup NB, Sindrup SH, Jensen TS. The evidence for pharmaco-logical treatment of neuropathic pain. Pain. 2010;150(3):573-581.
9. Freynhagen R, Geisslinger G, Schug SA. Opioids for chronic non-cancer pain. BMJ. 2013;346:f2937.
10. Manchikanti L, Ailinani H, Koyyalagunta D, et al. A systematic re-view of randomized trials of long-term opioid management for chronic non-cancer pain. Pain Physician. 2011;14(2):91-121.
11. Noble M, Treadwell JR, Tregear SJ, et al. Long-term opioid manage-ment for chronic noncancer pain. Cochrane Database Syst Rev. 2010; (1);CD006605.
12. Nüesch E, Rutjes AW, Husni E, Welch V, Jüni P. Oral or transdermal opioids for osteoarthritis of the knee or hip. Cochrane Database Syst Rev. 2009;(4);CD003115.
13. Iyer S, Davis KL, Candrilli S. Opioid use patterns and health care resource utilization in patients prescribed opioid therapy with and without constipation. Manag Care. 2010;19(3):44-51.
14. Boudreau D, Von Korff M, Rutter CM, et al. Trends in long-term opioid therapy for chronic non-cancer pain. Pharmacoepidemiol Drug Saf. 2009;18(12):1166-1175.
15. Manchikanti L, Manchukonda R, Pampati V, et al. Does random urine drug testing reduce illicit drug use in chronic pain patients receiving opioids? Pain Physician. 2006;9(2):123-129.
16. Kwong WJ, Diels J, Kavanagh S. Costs of gastrointestinal events after outpatient opioid treatment for non-cancer pain. Ann Pharmaco-ther. 2010;44(4):630-640.
17. Von Korff M, Kolodny A, Deyo RA, Chou R. Long-term opioid therapy reconsidered. Ann Intern Med. 2011;155(5):325-328.
18. Cepeda MS, Etropolski M, Weinstein R, Fife D, Boston R, Matcho A. Dose patterns in commercially insured subjects chronically exposed to opioids: a large cohort study in the United States. BMC Palliat Care. 2010;9:14.
19. Cicero TJ, Wong G, Tian Y, Lynskey M, Todorov A, Isenberg K. Co-morbidity and utilization of medical services by pain patients receiving opioid medications: data from an insurance claims database. Pain. 2009;144(1-2):20-27.
20. Braden JB, Fan MY, Edlund MJ, Martin BC, DeVries A, Sullivan MD.
Trends in use of opioids by noncancer pain type 2000-2005 among Arkansas Medicaid and HealthCore enrollees: results from the TROUP study. J Pain. 2008;9(11):1026-1035.
21. Schneeweiss S. A basic study design for expedited safety signal evaluation based on electronic healthcare data. Pharmacoepidemiol Drug Saf. 2010;19(8):858-868.
22. Begley CE, Annegers JF, Lairson DR, Reynolds TF. Methodological issues in estimating the cost of epilepsy. Epilepsy Res. 1999;33(1): 39-55.
23. Braman SS. The global burden of asthma. Chest. 2006;130(1 suppl): 4S-12S.
24. Gatti A, Reale C, Occhioni R, et al. Standard therapy with opioids in chronic pain management: ORTIBER study. Clin Drug Investig. 2009; 29(suppl 1):17-23.
25. Gatti A, Reale C, Luzi M, et al. Effects of opioid rotation in chronic pain patients: ORTIBARN study. Clin Drug Investig. 2010;30(suppl 2): 39-47.
26. Inciardi JA, Surratt HL, Cicero TJ, Beard RA. Prescription opioid abuse and diversion in an urban community: the results of an ultrarap-id assessment. Pain Med. 2009;10(3):537-548.
27. Inciardi JA, Surratt HL, Stivers Y, Cicero TJ. FDA approvals of generic drugs: impact on the diversion of opioid analgesics with a potential for abuse. J Opioid Manag. 2009;5(2):81-87.
28. Kuehn BM. Scientists probe ways to curb opioid abuse without hindering pain treatment. JAMA. 2007;297(18):1965-1967.
29. Kuehn BM. Opioid prescriptions soar: increase in legitimate use as well as abuse. JAMA. 2007;297(3):249-251.
30. Cepeda MS, Fife D, Vo L, Mastrogiovanni G, Yuan Y. Comparison of opioid doctor shopping for tapentadol and oxycodone: a cohort study. J Pain. 2013;14(2):158-164.
31. Cepeda MS, Fife D, Berlin JA, Mastrogiovanni G, Yuan Y. Character-istics of prescribers whose patients shop for opioids: results from a cohort study. J Opioid Manag. 2012;8(5):285-291.
32. Cepeda MS, Fife D, Chow W, Mastrogiovanni G, Henderson SC. As-sessing opioid shopping behaviour: a large cohort study from a medi-cation dispensing database in the US. Drug Saf. 2012;35(4):325-334.
33. Parente ST, Kim SS, Finch MD, et al. Identifying controlled sub-stance patterns of utilization requiring evaluation using administrative claims data. Am J Manag Care. 2004;10(11, pt 1):783-790.
34. Fishbain DA, Cole B, Lewis J, Rosomoff HL, Rosomoff RS. What percentage of chronic nonmalignant pain patients exposed to chronic opioid analgesic therapy develop abuse/addiction and/or aberrant drug-related behaviors? a structured evidence-based review. Pain Med. 2008;9(4):444-459.
35. National Institute on Drug Abuse. Research Report Series: Pre-scription Drug Abuse. 2014:16 (NIH Publication; no. 15-4881). NIH website. http://www.drugabuse.gov/sites/default/files/prescriptiondru-grrs_11_14.pdf. Updated November 2014. Accessed March 17, 2015.
36. Licata SC, Rowlett JK. Abuse and dependence liability of benzo-diazepine-type drugs: GABA(A) receptor modulation and beyond. Pharmacol Biochem Behav. 2008;90(1):74-89.
37. Wielage RC, Myers JA, Klein RW, Happich M. Cost-effectiveness analyses of osteoarthritis oral therapies: a systematic review. Appl Health Econ Health Policy. 2013;11(6):593-618.
38. Wielage RC, Patel AJ, Bansal M, Lee S, Klein RW, Happich M. Cost effectiveness of duloxetine for osteoarthritis: a Quebec societal per-spective. Arthritis Care Res (Hoboken). 2014;66(5):702-708.
Overhauling Quality Measurement in the US: Measure What Matters
October 30th 2024As the US charts its course through the next political era, it is crucial that we boldly allocate resources and prioritize what truly impacts patients. When faced with complexity, feasibility concerns, or entrenched norms, we must proclaim: “It’s the outcomes, stupid.”
Read More
No Free Lunch: The Misaligned Incentives of the American Health Care System
October 30th 2024The author highlights reasons why we have not seen substantial cost savings in the health care industry and why future efforts are likely to continue to see forceful pushback, as well as offers potential solutions.
Read More