In the United States, 1 of every 25 adults regularly uses prescription opioid medications.1 Accordingly, opioid use is higher in the United States than any other country in the world, and opioid-associated morbidity and mortality has been cited by the Centers for Disease Control and Prevention as a major public health concern.2,3 Nonetheless, opioid prescribing continues to accelerate.4 For example, from 1999 to 2011, consumption of hydrocodone doubled, and oxycodone consumption increased by 500% within the United States.1 As such, the proportion of individuals who are chronically using prescription opioid medications is expected to accelerate in the coming years.
Although opioids can provide effective pain relief for acute pain, chronic use is associated with numerous adverse dose-related effects. Chronic opioid exposure results in structural and functional central nervous system changes that mediate affect, impulse, reward, and motivation.5 Prior to surgery, these effects may lead to dependence, tolerance, and even reduction in pain thresholds with diffuse hyperalgesia, which may complicate pain management and patient rehabilitation following surgery.6,7 Previous studies have demonstrated that preoperative opioid use is linked to poorer clinical outcomes related to pain, morbidity, and mortality. Recently, we examined the relationship between self-reported preoperative opioid use and healthcare utilization and morbidity in the state of Michigan following abdominopelvic surgery, and observed that perioperative morbidity, readmission rates, and costs were higher among patients who reported preoperative opioid use.8 However, the threshold at which preoperative opioid influences postoperative healthcare utilization, cost, and morbidity has not been widely explored, and the extent to which patients who use higher doses of preoperative opioids are at greater risk for poor outcomes is not well understood.
In recent decades, much attention has been directed toward minimizing perioperative risk for patients with chronic conditions such as heart disease, diabetes, and tobacco use undergoing elective surgical procedures.9,10 Given the morbidity of chronic opioid use, identifying the potential opportunities to reduce costs and improve healthcare efficiency by addressing preoperative opioid use among patients and providers considering elective surgical procedures could provide considerable value. In this context, we sought to define the differences in healthcare utilization and associated costs among patients undergoing elective abdominal surgical procedures in the United States. We specifically examined 3 aspects of healthcare utilization: length of stay, discharge destination, and readmission within 30 days of the surgical procedure. Additionally, we examined postoperative costs at 90-, 180-, and 365-days following surgery. We hypothesized that patients who are chronically exposed to opioids undergoing surgery have higher healthcare utilization and associated costs compared with opioid naive patients, and that these outcomes increase with greater average daily opioid doses.
Data Source and Study Cohort
We analyzed the Truven Health MarketScan Research Databases, including the MarketScan Commercial Claims and Encounters Database and the Medicare Supplemental and Coordination of Benefits Database. These databases capture patient-level utilization of medical services, payment, and enrollment across inpatient, outpatient and prescription drugs, represent the healthcare utilization of approximately 50 million active employees, early retirees, Medicare-eligible retirees with employer-provided Medicare Supplemental plans, and their dependents each year. We specified that patients must be continually enrolled in healthcare plans captured by Marketscan with pharmaceutical coverage within 6 months prior to surgery and 1 year following surgery to capture comorbid conditions and opioid utilization before surgery and healthcare cost and utilization within 1 year after surgery.
We sought to examine the effects of preoperative opioid use on healthcare utilization among patients who underwent common elective abdominal surgical procedures. We examined the inpatient services claims from patients ages 18 and older who underwent one of the following major elective procedures requiring an inpatient stay between June 2009 and December 2012: hysterectomy, ventral hernia repair, anti-reflux procedures, and bariatric surgery procedures (see Figure 1, Supplemental Digital Content, http://links.lww.com/SLA/B159). These procedures were chosen as they represent abdominopelvic procedures commonly performed for both benign and malignant diseases in the United States. We identified these procedures using International Classification of Disease 9th revision codes (ICD-9 diagnosis codes) for the indication (primary diagnosis) and Current Procedural Terminology codes (CPT codes) for the procedure description (see Table 1, Supplemental Digital Content, http://links.lww.com/SLA/B159). As the majority of patients who undergo these procedures are admitted to the hospital for a short period of time, we opted to include patients whose claims were specifically billed as an inpatient stay. In this cohort, 23% of patients undergoing hysterectomy, 74% of patients undergoing ventral hernia repair, 38% of patients undergoing colorectal resection, and 29% of patients undergoing bariatric surgery were managed as an outpatient. In additional analyses, we observed that patients who underwent inpatient surgery were more likely to have a greater number of comorbid conditions, but were otherwise similar with respect to sociodemographic and clinical characteristics. We performed an additional sensitivity analysis including these patients in the cohort and separately, which did not significantly alter our findings. We have opted to present the results focused on patients who underwent inpatient procedures as we believe these patients represent a more homogenous cohort with respect to invasiveness of surgery.
Preoperative Opioid Use
Patients were classified as using preoperative opioid medications if they filled at least 1 opioid prescription within 30 days of their procedure and at least another prescription of opioid 30 to 90 days prior to the procedure to capture current and chronic use. We used generic drug names matched with National Drug Codes to identify opioid prescriptions from the insurance claims (see Table 2, Supplemental Digital Content, http://links.lww.com/SLA/B159), and obtain specific drug dose and type. For each prescription, we first converted the unit of the opioid component to milligrams, and then calculated average daily oral morphine equivalents (OMEs) for this prescription using the morphine equivalent conversion factor per milligram for this opioid medication.11,12 The total OME dosage for each opioid prescription was calculated as the daily OME exposure multiplied by the number of days of supply filled in this prescription. To quantify the average daily OME exposure for patients within 90 days prior to the surgery, we calculated total OME dosage in this period by adding up OME dosage of all the opioid prescriptions within this period. We then divided the total OME dosage within 90 days prior to the surgery by the total number of days supplied provided in these prescriptions to calculate the average daily OME exposure within 90 days prior to the surgery.
Our outcomes included measures of healthcare utilization and associated costs following surgery. We examined length of hospital stay in days, which was obtained from claims indicating the date of surgery until the date of discharge, as well as discharge destination (home vs. rehabilitation facility), and 30-day readmissions (for any cause) by examining associated claims in the postoperative period. We examined expenditures for the hospital stay, as well as 90-, 180-, and 365-day expenditures, excluding professional fees, including inpatient and outpatient financial data for procedural costs and facility fees for each patient. Expenditures were adjusted for inflation relative to the 2013 consumer price index.
We examined patient sociodemographic factors, specifically age (18–34, 35–44, 45–54, 55–64, and 65 yrs and older), sex, metropolitan statistical area (MSA), type of insurance plan, and geographic region of residence. The MSA describes geographical regions with a relatively high density of population at its core with adjacent communities with similar degree of social and economic factors as the core. We linked the MSA identifier for each patient to the 2010 census data and obtained median household income of each area where the patients resided. We then categorized regional median house income by quartiles for the patient cohort. We categorized insurance plan type into 5 groups: comprehensive health insurance, health maintenance organization plans, preferred provider organization plans, point of service plans, and other (eg, Medicare). Additionally, we included geographic region of the United States (northeast, north central, south, and west regions) into the analysis. Finally, we included the presence of comorbid conditions using standard risk-adjustment techniques. We identified comorbid conditions using their corresponding ICD-9-CM codes, and these were classified using the Elixhauser method.13
We generated descriptive statistics to describe the attributes of the study cohort. We created 2 independent variables describing the use of opioid before the procedure: (1) whether used opioids before surgery (dichotomous) and (2) average daily OME dosage within 90 days prior to the surgery. We categorized the latter variable into 7 groups (no preoperative opioid use, 0 < OME ≤ 10, 10 < OME ≤ 25, 25 < OME ≤ 50, 50 < OME ≤ 75, 75 < OME ≤ 100, and >100OME average per day). For each of the dependent variables (readmission within 30 d, discharge to home, length of hospital stay, cost of hospital stay, and healthcare cost within 90, 180, and 365 d following discharge), we constructed 2 regression models, each with 1 of the 2 variables describing opioid use before surgery as the primary predictor, and controlled for all other patient characteristics including procedure performed. Length of hospital stay was treated as count data, and we used negative binomial regression model for LOS; linear regression models were constructed to examine the relationship between preoperative opioid prescription fills and postoperative costs with log transformation. Using the adjusted odds ratios from logistic regression models, lambda from negative binomial regression models, and beta-coefficient from linear regression models, we back-calculated adjusted rate of readmission, adjusted rate of discharge to facilities, adjusted mean length of hospital stay and adjusted mean cost for hospital stay, and 90, 180, 365 days healthcare cost following surgery. P values <0.05 were considered significant in all final analyses and all statistical analyses were performed using SAS/STAT software (Cary, NC).
We identified 200,005 patients who underwent elective abdominal procedures during the study period: 55% underwent hysterectomy, 28% bariatric surgery, 6% reflux surgery, and 11% ventral hernia repair (Table 1). The majority of hysterectomy and bariatric patients were between the ages of 35 and 54 (74% and 59%, respectively) and the majority of reflux and hernia patients were 45 and older (68% and 81%, respectively). More than half of the participants receiving a procedure were female [64%–100% (hysterectomy)] and the majority of participants had a median annual household income of $40 to $60,000 (64%–67%). The greatest proportion of patients had a provider preferred organization insurance plan (56%–58%), followed by health maintenance organization plans (12%–15%). Although the majority of patients had multiple comorbidities with Elixhauser scores of 3 or more, concomitant psychiatric diagnoses were less common. Overall, approximately 1 in 10 patients had filled an opioid prescription prior to surgery (7%–13%).
Preoperative Opioid Prescription Fills
For the entire cohort, 182,428 (91%) patients did not fill opioid prescriptions preoperatively, 964 (0.5%) patients filled prescriptions corresponding to an average of 10 OMEs or less per day, 4577 (2.3%) patients filled prescriptions corresponding to an average between 10 and 25 OMEs per day, 7533 (3.8%) patients filled prescriptions corresponding to an average between 25 and 50 OMEs per day, 2305 (1.2%) patients filled prescriptions corresponding to an average between 50 and 75 OMEs per day, 899 (0.5%) patients filled prescriptions corresponding to an average between 75 and 100 OMEs per day, and 1299 (0.6%) patients filled prescriptions that corresponded to more than 100 OMEs per day.
Table 2 details the differences in measures of healthcare utilization and expenditures between preoperative opioid users and opioid naive patients across all procedures, adjusted for all patient demographic and clinical characteristics. On average, preoperative opioid users remained in the hospital longer than patients who did not use opioids preoperatively across all procedures (2.9 d vs. 2.5 d, P <0.001). Preoperative opioid users were also more likely to be readmitted to the hospital within 30 days of their respective surgical procedure, and discharged to a rehabilitation facility in comparison to opioid naive patients (4.5% vs. 3.6%, P <0.001; 3.6% vs. 2.5%, P <0.001, respectively). Of those readmitted, 21.4% of patients were due to infection, 14.5% for gastrointestinal reasons (ex. bowel obstruction, cholecystitis, nausea, vomiting), 8.6% for respiratory reasons (ex. pneumonia, respiratory failure), and 6.7% for failure to thrive (ex. dehydration, electrolyte abnormalities, fatigue, anorexia). However, we did not observe statistically significant differences in the indications for readmission by preoperative opioid use.
Figure 1 displays the differences in healthcare utilization postoperatively, stratified by the average daily preoperative opioid dose during the preceding 90 days. Preoperative opioid users, with the exception of those who had a 90-day preoperative dosage of 0 < OME ≤ 10, had significantly longer lengths of stay than their opioid naive counterparts (Fig. 1A, P <0.01 for all). Similarly, patients who filled prescriptions for opioids prior to surgery, with the exception of those who had a 90-day preoperative dosage of 0 < OME ≤ 10, were more likely to be readmitted to the hospital within 30 days of their procedure (Fig. 1B, P <0.05 for all). Additionally, patients who filled preoperative opioid prescriptions, with the exception of those who had a 90-day preoperative dosage of 0 < OME ≤ 10 and 75 < OME ≤ 100, were more likely to be discharged to a rehabilitation facility after their procedure (Fig. 1C, P < 0.05 for all).
Across all procedures, patients who used opioids preoperatively accrued higher total postoperative costs at 90 ($12036.60 vs. $3863.40, P <0.001), 180 ($16973.70 vs. $6790.60, P <0.001), and 365 ($25495.70 vs. $12113.80, P <0.001) days after their procedure than their opioid naive counterparts (Table 2).
Figure 2 illustrates measures of healthcare expenditure by category of oral morphine equivalent. Regardless of their 90-day preoperative daily dosage, all preoperative opioid users had significantly higher postoperative costs 90, 180, and 365 days after surgery than their opioid naive counterparts (P <0.001 for all). Although there were minimal differences in cost for the hospital stay related to the procedure, costs were substantially higher with increasing preoperative oral morphine dose in the later postoperative period at 90, 180, and 365 days following surgery.
In the United States, prescription opioids are widely used for chronic pain, and have grown increasingly popular in recent years. In this cohort of adults undergoing elective abdominal procedures, we observed that preoperative opioid use is an independent risk factor for longer length of stay, higher 30-day readmission rates, lower probability of being discharged home, and greater health expenditures in the postoperative period. Approximately 10% of patients regularly fill prescriptions for opioids medications in the preoperative period, with doses ranging from less than 10 OMEs (or the equivalent of 2 tablets of 5 mg hydrocodone) to over 100 OMEs daily (or the equivalent of 20 tablets of 5 mg hydrocodone). In this cohort, we did not observe a clear dose response relationship with respect to average cost of the hospital stay or length of stay. Although there are significant differences in length of stay by groups, the absolute differences are small and not likely to be meaningful differences. However, we observed dose-effect findings with respect to readmission rates, discharge destination, and late healthcare expenditures, which we hypothesize may reflect differences in pain control, recovery of activities of daily living, and rehabilitation between opioid users and nonusers.
Previous studies have examined the relationship between preoperative opioid use and postoperative outcomes, and have documented poorer outcomes among patients who use opioid preoperatively. For example, following total knee arthroplasty, chronic opioid users have poorer self-reported outcomes, greater stiffness, and more revision procedures, and chronic opioid users have a higher risk of in-hospital postoperative morbidity and mortality following elective orthopedic surgical procedures.14,15 Similarly, preoperative opioid use has been shown to increase the risk of revision surgery for pain-related conditions, such as compression neuropathy.16 Taken together, these findings suggest that identifying strategies to minimize opioid consumption prior to surgery, particularly elective surgery, is an opportunity to reduce patient risk, improve outcomes, and lower healthcare costs.
Although we are not able to discern the reason for which opioids were prescribed preoperatively, we observed that patients who filled prescriptions for opioid medications prior to surgery were more likely to have a greater number of comorbid medical, pain conditions, and psychological conditions (depression, substance abuse, anxiety) in the preceding 6 months of surgery. The interplay between comorbid medical, psychological, and pain conditions is complex, and many patients may take opioid pain medication for symptoms beyond pain. For example, patients often continue to use opioids after knee or hip arthroplasty despite pain relief.17 Goesling et al17 examined opioid consumption, mood, function, and pain longitudinally across hip and knee arthroplasty patients and observed that greater body pain, greater affected joint pain, and greater catastrophizing correlated with chronic use. Interestingly, however, improvement in joint pain following surgery did not predict opioid use. In addition, chronic opioid users more frequently report depressive symptoms.18 Opioid use is not correlated with pain severity or function among patients with depressive symptoms, suggesting that patients may use opioids to treat depressed mood, and depression may be critical risk factor for chronic opioid use.17,19,20 Future prospective studies focused on the extent to which these relationships mediate the effect of prescription opioid use on postoperative outcomes, cost, and healthcare utilization will inform strategies to mitigate the risk of chronic opioid use on surgical outcomes.
Our study has several notable limitations. First, our analysis focuses on insurance claims data, which is sensitive to detect perioperative events and medication fills, but lacks sufficient granularity to define opioid consumption and patient-reported pain. To define our study cohort, we used CPT codes for each procedure, but cannot discern with certainty that the indications for surgery were elective outside ICD-9 diagnosis codes. Additionally, this data is gathered from individuals with employer-based insurance and their beneficiaries, and may not be generalizable to the uninsured, underinsured, and those who rely solely on state and federal healthcare coverage. We chose to study 4 procedures that are commonly performed in the United States, but our findings may not be applicable to other types of procedures. Finally, we do not have access to the extent to which provider-level variation explains differences in the outcomes we observed.
Despite these limitations, our findings have important implications for patients anticipating surgery, and clinicians and surgeons seeking to optimize patient safety and postoperative outcomes. Achieving acceptable pain control following surgery is critical, and the under-treatment of pain has garnered national attention as an indicator of poor surgical quality.21–24 In recent years, there is a keen awareness of pain as a component of health status, and pain has been termed the “5th vital sign.”25 Inadequately controlled postoperative pain delays important recovery milestones, such as ambulation or attending to activities of daily living, and results in longer hospital stays and chronic postsurgical pain.26–30 Poorly managed pain is also a common reason for readmission, and is correlated with poor patient satisfaction and experience.31,32It is possible that for patients taking very high doses of prescription opioids prior to surgery, postoperative pain control is exceedingly challenging in routine clinical postoperative care, further limiting recovery. It is also known that chronic opioid use adversely affects other organ systems, such as the immune, endocrine, and respiratory systems, which may lead to additional morbidity and expenditures.33 In addition, although some patients benefit from chronic opioid therapy, most do not, and morbidity and abuse outweigh any benefits derived. 34–37 Furthermore, the effectiveness of opioids is not uniform across individuals, and varies by genetic, environmental, and biopsychosocial factors.38 Opioids may also be less effective among individuals with mood disturbance, including depression, anxiety, and catastrophic thinking, as well as patients who suffer from fibromyalgia, neuropathic pain, and psychiatric disease.18,39–41
Preoperative opioid use complicates perioperative pain management and postoperative rehabilitation, and can result in including respiratory depression, sedation, postoperative nausea and vomiting, urinary retention, and ileus. Even controlling for complications and comorbid risk factors, preoperative opioid use results in longer length of stay, higher readmission rates, and greater surgery-related expenditures.42–44 In this context, elective surgical procedures offer an opportunity for surgeons to pause, and identify strategies to optimize patient safety and minimize risk in the preoperative period. It also provides an opportunity to engage primary care providers at the time of surgical decision-making and preoperative workup to streamline transitions of care for those patients who may continue to require opioids for an extended period of time following surgery. Finally, our findings suggest that highlighting opioid alternatives for pain control when possible may reduce postoperative costs and healthcare utilization.
In conclusion, chronic opioid use complicates management following surgery, and increases postoperative healthcare utilization and costs independent of other risk factors. Therefore, developing preoperative interventions that focus on opioid cessation and alternative prior to elective surgery may improve the quality of surgical care delivered in the United States.
1. Kenan K, Mack K, Paulozzi L. Trends in prescriptions for oxycodone and other commonly used opioids in the United States, 2000–2010. Open Med
2. Paulozzi LJ, Jones CM, Mack KA, et al. Vital signs: overdoses of prescription opioid pain relievers—United States, 1999–2008. MMWR Morb Mortal Wkly Rep
3. Paulozzi LJ, Mack KA, Hockenberry JM. Vital signs: variation among states in prescribing of opioid pain relievers and benzodiazepines—United States, 2012. MMWR Morb Mortal Wkly Rep
4. Levy B, Paulozzi L, Mack KA, et al. Trends in opioid analgesic-prescribing rates by specialty, U.S., 2007–2012. Am J Prev Med
5. Younger JW, Chu LF, D’Arcy NT, et al. Prescription opioid analgesics rapidly change the human brain. Pain
6. Hina N, Fletcher D, Poindessous-Jazat F, et al. Hyperalgesia induced by low-dose opioid treatment before orthopaedic surgery: an observational case-control study. Eur J Anaesthesiol
7. Trang T, Al-Hasani R, Salvemini D, et al. Pain and poppies: the good, the bad, and the ugly of opioid analgesics. J Neurosci
8. Cron DC, Englesbe MJ, Bolton CJ, et al. Preoperative opioid use is independently associated with increased costs and worse outcomes after major abdominal surgery. Ann Surg
2016; [Epub ahead of print].
9. Bottle A, Mozid A, Grocott HP, et al. Preoperative risk factors in 10 418 patients with prior myocardial infarction and 5241 patients with prior unstable angina undergoing elective coronary artery bypass graft surgery. Br J Anaesth
10. Godoy DA, Di Napoli M, Biestro A, et al. Perioperative glucose control in neurosurgical patients. Anesthesiol Res Pract
11. Sullivan MD, Edlund MJ, Fan MY, et al. Trends in use of opioids for non-cancer pain conditions 2000–2005 in commercial and Medicaid insurance plans: the TROUP study. Pain
12. Logan J, Liu Y, Paulozzi L, et al. Opioid prescribing in emergency departments: the prevalence of potentially inappropriate prescribing and misuse. Med Care
13. Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care
14. Menendez ME, Ring D, Bateman BT. Preoperative opioid misuse is associated with increased morbidity and mortality after elective orthopaedic surgery. Clin Orthop Relat Res
15. Zywiel MG, Stroh DA, Lee SY, et al. Chronic opioid use prior to total knee arthroplasty. J Bone Joint Surg Am
16. Gaspar MP, Jacoby SM, Osterman AL, et al. Risk factors predicting revision surgery after medial epicondylectomy for primary cubital tunnel syndrome. J Shoulder Elbow Surg
17. Goesling J, Moser SE, Zaidi B, et al. Trends and predictors of opioid use following total knee and total hip arthroplasty. Pain
18. Goesling J, Henry MJ, Moser SE, et al. Symptoms of depression are associated with opioid use regardless of pain severity and physical functioning among treatment-seeking patients with chronic pain. J Pain
19. Braden JB, Sullivan MD, Ray GT, et al. Trends in long-term opioid therapy for noncancer pain among persons with a history of depression. Gen Hosp Psychiatry
20. Sullivan MD, Edlund MJ, Zhang L, et al. Association between mental health disorders, problem drug use, and regular prescription opioid use. Arch Intern Med
21. Porter J, Jick H. Addiction rare in patients treated with narcotics. N Engl J Med
22. Acute pain management in adults: operative procedures. Agency for Health Care Policy and Research. Clin Pract Guidel Quick Ref Guide Clin
23. Acute pain management: operative or medical procedures and trauma, Part 1. Agency for Health Care Policy and Research. Clin Pharm
24. Cheatle MD. Prescription opioid misuse, abuse, morbidity, and mortality: balancing effective pain management and safety. Pain Med
2015; 16 (suppl 1):S3–S8.
25. Lorenz KA, Sherbourne CD, Shugarman LR, et al. How reliable is pain as the fifth vital sign? J Am Board Fam Pract
26. Lawton J, Waugh N, Barnard KD, et al. Challenges of optimizing glycaemic control in children with Type 1 diabetes: a qualitative study of parents’ experiences and views. Diabet Med
27. Apfelbaum JL, Ashburn MA, Connis RT, et al. American Society of Anesthesiologists Task Force on Acute Pain M. Practice guidelines for acute pain management in the perioperative setting: an updated report by the American Society of Anesthesiologists Task Force on Acute Pain Management. Anesthesiology
28. Alvarez MP, Foley KE, Zebley DM, et al. Comprehensive enhanced recovery pathway significantly reduces postoperative length of stay and opioid usage in elective laparoscopic colectomy. Surg Endosc
29. Kehlet H, Jensen TS, Woolf CJ. Persistent postsurgical pain: risk factors and prevention. Lancet
30. Huang A, Katz J, Clarke H. Ensuring safe prescribing of controlled substances for pain following surgery by developing a transitional pain service. Pain Manag
31. Curtin CM, Hernandez-Boussard T. Readmissions after treatment of distal radius fractures. J Hand Surg Am
32. Chang CB, Cho WS. Pain management protocols, peri-operative pain and patient satisfaction after total knee replacement: a multicentre study. J Bone Joint Surg Br
33. Ray WA, Chung CP, Murray KT, et al. Prescription of long-acting opioids and mortality in patients with chronic noncancer pain. JAMA
34. Reuben DB, Alvanzo AA, Ashikaga T, et al. National Institutes of Health Pathways to Prevention Workshop: the role of opioids in the treatment of chronic pain. Ann Intern Med
35. Chou R, Turner JA, Devine EB, et al. The effectiveness and risks of long-term opioid therapy for chronic pain: a systematic review for a National Institutes of Health Pathways to Prevention Workshop. Ann Intern Med
36. Deyo RA, Von Korff M, Duhrkoop D. Opioids for low back pain. BMJ
37. Campbell JN. The fifth vital sign revisited. Pain
38. Ren ZY, Xu XQ, Bao YP, et al. The impact of genetic variation on sensitivity to opioid analgesics in patients with postoperative pain: a systematic review and meta-analysis. Pain Physician
39. Helmerhorst GT, Vranceanu AM, Vrahas M, et al. Risk factors for continued opioid use one to two months after surgery for musculoskeletal trauma. J Bone Joint Surg Am
40. Wasserman RA, Brummett CM, Goesling J, et al. Characteristics of chronic pain patients who take opioids and persistently report high pain intensity. Reg Anesth Pain Med
41. Hooten WM, Shi Y, Gazelka HM, et al. The effects of depression and smoking on pain severity and opioid use in patients with chronic pain. Pain
42. Oderda GM, Evans RS, Lloyd J, et al. Cost of opioid-related adverse drug events in surgical patients. J Pain Symptom Manage
43. Oderda GM, Gan TJ, Johnson BH, et al. Effect of opioid-related adverse events on outcomes in selected surgical patients. J Pain Palliat Care Pharmacother
44. Raebel MA, Newcomer SR, Reifler LM, et al. Chronic use of opioid medications before and after bariatric surgery. JAMA