Opioids are the most frequently prescribed medications for the treatment of severe pain in developed countries, with up to 90% of American patients who present to specialized pain centers being treated with opioids.1 The frequency of opioid use is increasing, with one study reporting that opioid prescriptions for chronic abdominal pain in the United States more than doubled between 1997 and 2008.2 In addition to this rise in opioid use, the incidence of opioid-related adverse events has increased as well, with opioid-induced constipation (OIC) being one of the most common and persistent events.1
In a systematic review of randomized trials of the use of oral opioids for chronic noncancer pain, the reported incidence of OIC from individual trials was as high as 71%.3 OIC has been associated with significant clinical and economic burdens, and may have a negative impact on patients’ quality of life.4 In addition, persistent constipation may lead to serious medical sequelae, such as bowel obstruction and fecal impaction,5 resulting in the increased use of medical services and decreased productivity loss.6
To the best of our knowledge, this is the first study using real-world administrative data to have quantified the economic burden of OIC in patients who are receiving a variety of opioids specifically for noncancer pain. Previous studies have shown that OIC is associated with significant economic burden; however, these studies have included only patients with cancer-related pain7 or have included cancer and noncancer pain without differentiating between them.8
Because of the exponential growth in opioid use for noncancer pain9,10 and the lack of real-world data on its gastrointestinal-related risks, it is important to quantify the economic impact of OIC in this population. One study concluded that the economic burden of gastrointestinal events among opioid users for noncancer pain is substantial, but that study included only patients receiving immediate-release oxycodone or hydrocodone.11 In addition, that study followed patients only for a relatively short period (90 days), which may underestimate the burden of OIC, a chronic condition in many cases.
Furthermore, previous literature has focused on commercially insured and relatively young populations; therefore, the generalizability of their results to an older population is questionable. It is widely reported in the literature that elderly patients and long-term care residents are more likely to develop OIC than younger patients.12,13 However, little is known about the economic burden associated with OIC in these 2 patient populations. The paucity and limitations of the existing evidence underscore the need for new studies to evaluate the economic burden for long-term opioid users who have noncancer pain.
The objective of our study was to estimate the healthcare resource utilization and costs associated with OIC among patients who are receiving an opioid long term for noncancer pain. To better understand the economic burden among various populations, 3 subpopulations were involved in our analysis, including nonelderly patients, elderly patients, and patients who received opioids in long-term care facilities.
This study was a retrospective claims analysis using data from the Truven Health MarketScan Commercial Claims and Encounters Database and Truven Health MarketScan Medicare Supplemental and Coordination of Benefits Database from January 1, 2007, through December 31, 2011. The index event window, preindex window, and postindex follow-up window are defined in Figure 1. The initial opioid prescription during the index period (January 2008-December 2010) was assigned as the study index date. Patients had continuous plan insurance coverage for 1 year before (baseline period) and after the index date.
Patients were included if they had at least 1 claim for an opioid prescription during the study period. To be included as long-term opioid users, patients had to have at least a 90-day supply of opioids11 with ≤15 days between prescription refills. Patients were excluded if they were aged <18 years or were diagnosed with cancer (International Classification of Diseases, Ninth Edition [ICD-9] code, 140.x-239.x), or had drug abuse (ICD-9 code, 305.x) or drug dependence (ICD-9 code, 304.x) during the study period. Patients who had claims for opioids during the 12-month preindex period were excluded from the study. In addition, patients were excluded if they had a primary or secondary diagnosis of constipation (ICD-9 code, 564.0) or bowel obstruction (ICD-9 code, 560.x) within the 3 months before the index date.14
This study identified 3 study cohorts—nonelderly patients (aged 18-64 years); elderly patients (aged ≥65 years); and patients who used opioids at long-term care facilities, who could be either elderly or nonelderly. Each study cohort was divided into a constipation group, which included long-term opioid users who developed constipation, and a group without constipation, which included long-term opioid users who did not develop constipation.
Measurement and Outcomes
OIC was defined as at least 1 claim with a primary or secondary ICD-9 diagnosis code for constipation (564.0) or intestinal obstruction without hernia (560.x) in the 12-month postindex period after the initiation of long-term opioid therapy.
Healthcare resource utilization and healthcare costs were measured for a 12-month period after the initiation of opioids. All costs were presented as annual costs per patient and were adjusted to 2011 values based on the current Consumer Price Index information provided by the US Bureau of Labor Statistics.15 The analysis of the healthcare resource utilization included the inpatient admission rate, inpatient length of stay, emergency department visit rate, the number of emergency department visits, office visit rate, and the number of physician visits during the follow-up period. The total healthcare costs included costs for inpatient, pharmacy, outpatient, emergency department, long-term care facility, and other costs.
The demographics, comorbidities, opioid use patterns, and resource utilization were compared between patients who developed OIC and those who did not using a bivariate approach. The comparisons of continuous variables were performed using t-tests and categorical variables using Pearson chi-square tests. The cost data were compared between the groups with and without OIC using a nonparametric test (Wilcoxon rank sum test), because cost data are often heavily skewed.16 Patients with OIC were matched by propensity score to patients without OIC with a 1:1 ratio using the nearest-neighbor matching algorithm. All the baseline demographic and clinical variables were used for the matching method.
The differences in healthcare resource utilization, total costs, and cost components were calculated between the patients with OIC and their propensity score–matched group without OIC during the 12-month follow-up period. A generalized linear model with gamma distribution and log link was performed to estimate the impact of OIC on total cost, which was adjusted for covariates (age, sex, insurance plan, region, dose and duration of opioid use, baseline hospitalization, Charlson comorbidity score, nausea/vomiting, and constipation-related disease conditions). The analyses were performed using Statistical Analysis Software version 9.2.3 (SAS Institute; Cary, NC). P <.05 was considered to indicate statistical significance.
Figure 2 outlines the selection of study cohorts. More than 24 million patients who started therapy with an opioid during the study period were identified. Of the 16,766 long-term opioid users who met the study criteria, 13,808 patients were classified as nonelderly and 2958 patients were classified as elderly; among these, 566 patients received an opioid medication in a long-term care facility.
Baseline characteristics of the study cohorts are summarized in Table 1. The mean patient ages were 48.6 years (standard deviation [SD], 10.4); 78.7 years (SD, 8.1); and 80.7 years (SD, 11.6) for the nonelderly, elderly, and long-term care cohorts, respectively. A total of 50%, 70%, and 77% of the patients, respectively, were female.
Among the nonelderly cohort, a higher proportion of patients with OIC were female (63.1%) compared with patients without OIC (49.8%). Patients with OIC had significantly higher comorbidity scores, and more patients experienced nausea or vomiting compared with the patients without OIC across all 3 study cohorts.
In the nonelderly and elderly cohorts, depression and Parkinson disease were more prevalent among patients with OIC than among those without OIC. No significant differences in any constipation-associated conditions were found between those with and without OIC in the long-term care cohort.
Healthcare Resource Utilization
After propensity score matching, the nonelderly and elderly patients with OIC had more hospital admissions and a longer inpatient length of stay than the group without OIC. Among the nonelderly patients, the group with OIC had more patients with physician office visits and a higher annual number of office visits than the group without OIC. Among the elderly patients, more patients with OIC had emergency department visits, and the group with OIC had a higher mean annual number of emergency department visits than the patients without OIC. In the long-term care cohort, there was no significant difference in healthcare resource utilization between the groups with and without OIC (Table 2).
After matching by key covariates, patients with OIC had significantly higher total healthcare costs than patients without OIC in all 3 cohorts, including the nonelderly population ($23,631 ± $67,209 vs $12,652 ± $19,717, respectively; P <.001 [Figure 3]), the elderly population ($16,923 ± $38,191 vs $11,117 ±$19,525, respectively; P = .009 [Figure 4]), and the long-term care population ($16,000 ± $22,897 vs $14,437 ± $25,690, respectively; P = .049).
Patients with OIC had significantly higher inpatient and emergency department costs among the elderly and nonelderly patients compared with their matched patients without OIC. In addition, nonelderly patients with OIC had significantly higher outpatient costs and other costs; however, the same trend was not found for the pharmacy cost.
Among elderly patients, those with OIC had significantly higher long-term care costs compared with their matched counterparts without OIC. Among patients initiating opioids in long-term care facilities, patients with OIC had higher total costs than patients without OIC.
The results from the multivariate regression (generalized linear model) suggest that elderly patients with OIC had 89% higher costs than elderly patients without OIC ($19,963 vs $10,556, respectively; P <.001), whereas costs were 52% higher in nonelderly patients with OIC than in those without OIC ($15,737 vs $10,332; P <.001). In the long-term care cohort, the healthcare costs were not significantly different between the patients with and without OIC (P = .42).
Among the nonelderly and elderly patients, the common cost drivers included baseline hospitalizations, having nausea or vomiting during follow-up, and the higher Charlson comorbidity score. In addition, for the nonelderly patients, there are several significant cost drivers, including comorbid conditions, such as obesity, depression, multiple sclerosis, and Parkinson’s disease (Table 3).
Our study is the first to examine the economic burden associated with constipation among subpopulations (nonelderly, elderly, and long-term care cohorts) of long-term opioid users for noncancer pain using data from a real-world setting. We found that for patients treated with opioids for ≥90 days, constipation after opioid therapy significantly increased the healthcare resource utilization and the total healthcare costs among all 3 subpopulations.
In particular, OIC is associated with significantly increased inpatient, outpatient, and emergency department costs among the nonelderly population. Among the elderly population, OIC is associated with significantly increased inpatient and emergency department costs. In addition, our study identified significant and clinically important cost drivers, such as certain comorbid conditions, having vomiting or nausea, and having baseline hospitalization.
This finding is consistent with previous studies that evaluated the healthcare resource utilization of groups with OIC versus groups without OIC among opioid users.7,8,11 The studies by Iyer and colleagues and Candrilli and colleagues assessed the economic burden of OIC among patients receiving opioid therapy.7,8 However, these 2 studies did not differentiate between patients with cancer-related pain and those without cancer.7,8 Kwong and colleagues identified the substantial economic burden of gastrointestinal events among users of opioids for noncancer pain, but the target population was limited to patients receiving immediate-release oxycodone or hydrocodone, and the costs were evaluated for only 90 days after the initiation of an opioid.11
Outside of the United States, there were few published studies assessing the economic burden of OIC.17,18 For example, a Swedish study examined the indirect and direct medical costs associated with OIC using survey data, and the researchers concluded that OIC imposed substantial costs to society, especially for patients with severe OIC.18 However, these studies did not differentiate among subpopulations (eg, the elderly and patients residing in long-term care facilities) with different levels of risk for OIC that may lead to varied levels of resource utilization.
The burden of OIC reported in literature varies widely across studies, and can rise to approximately $40,000 as a result of the difference in service type, study perspectives (ie, societal, payer, provider), inclusion of patients with cancer, or severity of the disease. Our study found a $10,979 incremental cost of OIC in the nonelderly population, which is higher than the finding by Kwong and colleagues.11 The differences between both studies may result from our focus on long-term opioid users only, whereas their study included patients who used opioids for a short time only (ie, 90 days).
We also found that the elderly population and patients residing in long-term care facilities had a higher rate of identified cases of OIC. It is well-documented in the literature that older adults are more likely to develop OIC than younger patients.12 The reasons may include age-related physiologic changes in the gastrointestinal tract; comorbidities (eg, Parkinson’s disease); concomitant use of multiple medications, especially other medications that cause constipation; difficulty reaching the bathroom; low fluid and fiber intake; and inadequate time and privacy to defecate.12,13 Such factors may be especially prevalent in the elderly and in patients residing in long-term care facilities. Therefore, it is particularly important to study the elderly population and the population residing in long-term care facilities, which are disproportionately affected by OIC.
The significant impact of OIC may not be well-recognized by healthcare professionals. Most symptoms of OIC persist for the length of treatment with an opioid, which has a substantially negative impact on patients’ quality of life and could reduce treatment adherence with pain management.19 For example, one survey-based study found that more than 33% of the population had missed, decreased, or stopped using opioids to reduce their OIC.20
The trade-off between optimal opioid use for pain therapy and the risk for the discontinuation of opioids as a result of OIC poses a challenge in reaching the pain management goal. Thus, the alleviation of constipation may help optimize treatment with an opioid, help with pain control, and reduce economic burden. OIC is a persistent condition.21 Although over-the-counter laxatives can address some of the symptoms of OIC, alternative therapies are needed to relieve constipation and its related symptoms for patients receiving long-term treatment with an opioid.
The results of our study highlight the importance of providing patients with OIC with effective and available treatment options, as well as enhancing patient–physician communication regarding the management of OIC. Such efforts may reduce the economic burden of OIC in the population of long-term opioid users. Future research should focus on identifying the risk factors for and measuring the impact of OIC on pain management.
Our study is subject to several limitations that are inherent to the use of a claims analysis, which might have influenced our estimates of the economic burden of OIC. First, the patients we identified as having OIC may represent more severe constipation cases as a result of our reliance on ICD-9 codes and not on patient-reported measures. Therefore, our estimated economic burden may reflect more severe cases of OIC.
Second, nonreimbursable items, such as over-the-counter medications, were not captured in the commercial insurance claims. As a result, the estimated pharmaceutical costs associated with OIC may be conservative.
Third, healthcare resource utilization and cost data should be interpreted carefully, because these were not limited to visits and costs resulting only from OIC. In other words, it is not possible to establish the causality relationship between the incremental economic burden and OIC given the nature of retrospective study design.
Fourth, OIC is also likely to be associated with productivity loss, which cannot be assessed through claims database analyses.
Fifth, patients with OIC may have had more serious or severe medical comorbidities, and may have used more healthcare resources and medical services that cannot be attributed to OIC. Propensity score matching cannot correct bias from unmeasured confounders, and this may tend to overestimate the magnitude of the difference in economic costs between patients who did and did not experience OIC.
Notwithstanding these limitations, results from this study are consistent with previous findings that OIC is associated with significant economic burden.5 Our study adds to the current knowledge by assessing elderly patients, nonelderly patients, and residents in long-term care facilities who receive long-term opioid therapy for the management of noncancer pain. Our findings show that OIC is associated with increased healthcare costs in all 3 subpopulations that were included in the study. The economic burden of OIC should be considered when evaluating the cost-effectiveness of treatments for the management of pain. The use of effective therapies may reduce the economic burden that is associated with OIC. Future research using alternative sources, such as patient surveys or chart reviews, is needed to better determine the burden of OIC on pain management from a societal perspective, especially in high-risk subpopulations, such as the elderly.
The authors would like to thank Shawn Yu of Takeda Pharmaceuticals International, Inc, for his contribution in reviewing the manuscript and commenting on the statistical methods before submission.
This study was funded by Takeda Pharmaceuticals International, Inc.
Author Disclosure Statement
Ms Wan, Dr Corman, Dr Gao, and Mr Liu are employees of Pharmerit International, which received research funding from Takeda Pharmaceuticals International, Inc. Dr Patel is an employee of Immensity Consulting, and a contractor for Takeda Pharmaceuticals International, Inc; Dr Mody is an employee of Takeda Pharmaceuticals International, Inc.
1. Trescot A, Glaser SE, Hansen H. Effectiveness of opioids in the treatment of chronic non-cancer pain. Pain Physician. 2008;11:S181-S200.
2. Dorn SD, Meek PD, Shah ND. Increasing frequency of opioid prescriptions for chronic abdominal pain in US outpatient clinics. Clin Gastroenterol Hepatol. 2011;9:1078-1085.e1. Erratum in: Clin Gastroenterol Hepatol. 2012;10:332.
3. Moore RA, McQuay HJ. Prevalence of opioid adverse events in chronic non-malignant pain: systematic review of randomised trials of oral opioids. Arthritis Res Ther. 2005;7:R1046-R1051.
4. Bell T, Milanova T, Grove G, et al. OBD symptoms impair quality of life and daily activities, regardless of frequency and duration of opioid treatment: results of a US patient survey (PROBE survey). J Pain. 2007;8(4 suppl 1). Abstract 882.
5. Singh G, Kahler K, Bharathi V, et al. Constipation in adults: complications and comorbidities. Gastroenterology. 2005;128(4 suppl 2). Abstract S960.
6. Bell T, Annunziata K, Freedman D, et al. Opioid-induced constipation increases healthcare resource use and impairs work productivity: comparison with other patient groups with and without constipation. J Pain. 2007;8(4 suppl 1). Abstract 897.
7. Candrilli SD, Davis KL, Iyer S. Impact of constipation on opioid use patterns, health care resource utilization, and costs in cancer patients on opioid therapy. J Pain Palliat Care Pharmacother. 2009;23:231-241.
8. 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;19:44-51.
9. 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.
10. Sullivan MD, Edlund MJ, Fan M-Y, et al. Trends in use of opioids for non-cancer pain conditions 2000-2005 in commercial and Medicaid insurance plans: the TROUP study. Pain. 2008;138:440-449.
11. Kwong WJ, Diels J, Kavanagh S. Costs of gastrointestinal events after outpatient opioid treatment for non-cancer pain. Ann Pharmacother. 2010;44:630-640.
12. Max EK, Hernandez JJ, Sturpe DA, Zuckerman IH. Prophylaxis for opioid-induced constipation in elderly long-term care residents: a cross-sectional study of Medicare beneficiaries. Am J Geriatr Pharmacother. 2007;5:129-136.
13. McMillan SC. Assessing and managing opiate-induced constipation in adults with cancer. Cancer Control. 2004;11(3 suppl 1):3-9.
14. Staats PS, Markowitz J, Schein J. Incidence of constipation associated with long-acting opioid therapy: a comparative study. South Med J. 2004;97:129-134.
15. National Center for Health Statistics. Health, United States, 2010: with special feature on death and dying. February 2011. www.cdc.gov/nchs/data/hus/hus10.pdf#specialfeature. Accessed March 10, 2015.
16. Kessler ER, Shah M, Gruschkus SK, Raju A. Cost and quality implications of opioid-based postsurgical pain control using administrative claims data from a large health system: opioid-related adverse events and their impact on clinical and economic outcomes. Pharmacotherapy. 2013;33:383-391.
17. Takemoto MLS, Fernandes RA, Almeida GR, et al. Health care resource use and costs in opioid-treated patients with and without constipation in Brazil. Value Health. 2011;14(5 suppl 1):S78-S81.
18. Hjalte F, Berggren A-C, Bergendahl H, Hjortsberg C. The direct and indirect costs of opioid-induced constipation. J Pain Symptom Manage. 2010;40:696-703.
19. Goodheart CR, Leavitt SB. Managing opioid-induced constipation in ambulatory-care patients. Pain Treatment Topics. August 2006. http://paincommu
nity.org/blog/wp-content/uploads/Managing_Opioid-Induced_Constipation.pdf. Accessed March 2, 2015.
20. Bell TJ, Panchal SJ, Miaskowski C, et al. The prevalence, severity, and impact of opioid-induced bowel dysfunction: results of a US and European Patient Survey (PROBE 1). Pain Med. 2009;10:35-42.
21. Panchal SJ, Müller-Schwefe P, Wurzelmann JI. Opioid-induced bowel dysfunction: prevalence, pathophysiology and burden. Int J Clin Pract. 2007;61:1181-1187.