Back pain is the most common reason for a healthcare visit in the United States and the most disabling disorder worldwide.1 In a 2015 survey of US adults, nearly one in three reported chronic back pain in the previous year and the total expenditures for treating spine conditions between 2012 and 2014 were estimated at $315 billion.2 Over the last several decades, opioid therapy has become a common approach for treating chronic back pain, despite evidence demonstrating no benefit to long-term opioid therapy when compared with nonopioid medications.3–6 Still, a national survey found back pain was reported in 59% of long-term opioid users, and a 2016 review of over 34,000 surgical patients found 57% of patients presenting for spine surgery used opioids preoperatively.7,8
The dangers of opioids are well established. According to the Centers for Disease Control, opioids led to over 42,000 deaths in 2016, with 40% of these involving a prescription opioid.9 In addition to the risks of abuse and overdose, higher preoperative opioid dosages have been associated with worse patient-reported outcomes after spine surgery based on self-reported data.10–12 Chronic preoperative opioid therapy has been associated with postoperative complications, postoperative opioid use, and increased cost after spine surgery,13,14 but published studies thus far have been unable to accurately assess the impact of both preoperative chronic opioid therapy and preoperative opioid dosage on 1-year patient-reported outcomes after spine surgery. Understanding the impact of both preoperative chronic opioid therapy and preoperative opioid dosage on long-term outcomes after spine surgery is essential to accurately inform patients and guide the development of opioid weaning protocols.
To examine how preoperative chronic opioid therapy and opioid dosage relates to long-term outcomes after spine surgery, we used our state's Prescription Drug Monitoring Program (PDMP) to collect daily Morphine Milligram Equivalents (MME) from 9 months preoperatively until 1 year postoperatively for patients undergoing elective spine surgery at a single academic medical center. Our primary outcomes were clinically meaningful improvements in 1-year patient-reported outcomes for pain, function, and quality of life. Additional outcomes measured included satisfaction with surgery at 1 year, return to work, complications within 90 days of surgery, and postoperative chronic opioid use. A clear understanding of the associations between preoperative opioid use and negative clinical outcomes will allow for improved patient counseling and with regard to perioperative opioid use.
MATERIALS AND METHODS
Study Design and Patient Population
Using data from our prospective clinical spine registry linked with opioid prescription data from the state PDMP database, we conducted a longitudinal cohort study of patients undergoing elective lumbar or cervical spine surgery at a single academic medical center between January 2011 and February 2017. Patients undergoing spine surgery for degenerative spine disease at our institution are enrolled in a longitudinal registry if they are English speaking, age > 18, and willing to participate. For this study, patients were excluded if they lacked 1-year follow-up data, lived in a different state or had no identifiable record in the state PDMP, had pathologic spine disease (tumor or infection), or presented with less than 3 months of axial and/or extremity pain.
Tracking Dispensed Opioids
The state PDMP was queried to collect the dosage (MME) and duration of dispensed opioid prescriptions, and calculate the daily MME for each patient from 9 months preoperatively until 1 year after surgery (Figure 1). The main exposure variables examined were preoperative chronic opioid therapy based on the common definition of most days in 90 days,4 and high-preoperative opioid dosage, defined as a daily MME > 30 based on previously published thesholds.11,15 To minimize misclassification, we specified preoperative chronic opioid therapy as having an active prescription most days (>50%) in each month for 3 consecutive months prior to surgery, and we determined a patient's preoperative dosage using an average daily MME of the 2 weeks prior to surgery.
Primary outcomes were clinically meaningful improvements in pain, function, and quality of life at 1 year after surgery. Patient-reported outcomes were collected preoperatively, 3 months postoperatively, and 1 year postoperatively through telephone or web-based interviews.16 Axial and extremity pain were assessed using the 11-point numeric rating scale for neck pain, back pain, leg pain, and arm pain.17 Function was assessed using the Oswestry Disability Index (ODI) or Neck Disability Index (NDI) for lumbar or cervical spine patients, respectively.18,19 Quality of life was assessed using the Euro-Qol 5D (EQ-5D).20 A “clinically meaningful improvement” was defined as a 30% improvement from preoperative scores.21 Additional outcomes of interest included 1-year satisfaction with surgery (using The North American Spine Society Satisfaction Scale22), return to work status, 90-day complications, and postoperative chronic opioid use at 1 year. Complications tracked included pneumonia, myocardial infarction, deep vein thrombosis, urinary tract infection, surgical site infection or hematoma, or new neurological deficit.
Baseline patient and surgical characteristics were prospectively recorded at the time of registry enrollment through interview or chart review. Data included age, sex, race, smoking history, history of arthritis, anxiety or depression (collected from EQ-5D questionnaire), insurance payer, duration of symptoms, American Society of Anesthesiologist (ASA) class, cervical versus lumbar surgery, revision surgery, number of surgical levels, preoperative pain scores (extremity and axial), and preoperative disability (ODI and NDI scores). All data were recorded and managed using REDCap electronic data capture tools.23
Descriptive statistics were calculated for the study population and stratified based on exposure to preoperative chronic opioid therapy. Covariates were compared using chi-square test for categorical variables (expressed as frequencies and percentages), and Mann–Whitney test for continuous variables (expressed as medians with interquartile ranges [IQR]). The average daily MME was computed for each patient and plotted over time (Figure 1). The unadjusted rates of each outcome were compared between patients with and without preoperative chronic opioid therapy, and unadjusted odds ratios were calculated.
A series of multivariable logistic regression analyses were used to assess for associations between preoperative chronic opioid therapy and each outcome. Covariates were chosen a priori and all covariates listed previously were included in all models. Adjusted odds ratios (aOR) with 95% confidence intervals (95% CI) were computed for the odds of a negative clinical outcome associated with preoperative chronic opioid therapy. Additional multivariable regression analyses were used to assess the odds of a negative clinical outcome associated with high-preoperative opioid dosage, while adjusting for preoperative chronic opioid therapy and other covariates. Lastly, a subgroup analysis was performed to determine if high-preoperative opioid dosage was associated with negative clinical outcomes only among patients without preoperative chronic opioid therapy.
A sensitivity analysis was conducted by modifying the definition of “high-preoperative opioid dosage” into multiple categories: low- (<30 MME/d), intermediate- (30–90 MME/d), and high-preoperative (>90 MME/d) opioid dosage. The proportion of patients presenting with preoperative chronic opioid therapy each year were compared to assess for bias related to a change in prescribing patterns over time. A P value less than 0.05 was used to indicate statistical significance. All analyses were performed using SPSS version 25 (IBM Inc, Chicago, IL).
Of 2128 patients included in this study, 78% (n = 1666) had at least one opioid prescription dispensed in the 9 months prior to surgery, and 21% overall (n = 445) had preoperative chronic opioid therapy. A high-preoperative opioid dosage (>30 MME/d) was identified in 73% of patients with preoperative chronic opioid therapy (n = 327) and 8% of patients without preoperative chronic opioids (n = 131). When compared with those without, patients with preoperative chronic opioid therapy had several significant differences in baseline characteristics, including worse baseline scores for pain, function, and quality of life (P < 0.001) (Table 1). During the study period, our 1-year follow-up rate was 78.2% and < 1% of data was missing.
Preoperative Chronic Opioid Therapy
Compared with those without and before adjusting for covariates, patients with preoperative chronic opioid therapy had significantly less meaningful improvements at 1 year in extremity pain, axial pain, function, and quality of life; lower satisfaction at 1 year; fewer had returned to work; more had 90-day complications; and more had postoperative chronic opioid use at 1 year (Figure 2).
After adjusting for covariates, preoperative chronic opioid therapy was associated with significantly higher odds of not achieving a clinically meaningful improvement at 1 year in extremity pain (aOR, 1.55; 95% CI, 1.21–1.99; P = 0.001), axial pain (aOR, 1.73; 95% CI, 1.37–2.18; P < 0.001), function (aOR, 1.72; 95% CI, 1.36–2.16; P < 0.001), and quality of life (aOR, 1.48; 95% CI, 1.18–1.87; P = 0.001); not being satisfied with surgery (aOR, 1.73; 95% CI, 1.33–2.27; P < 0.001), having a 90-day complication (aOR, 2.87; 95% CI, 1.68–4.89; P < 0.001), and having postoperative chronic opioid use at 1 year (aOR, 15; 95% CI, 11.42–19.71; P < 0.001) (Figure 3).
High-preoperative Opioid Dosage
After adjusting for preoperative chronic opioid therapy and other covariates, high-preoperative opioid dosage was significantly associated with postoperative chronic opioid use at 1 year (aOR, 4.1; 95% CI, 2.9–5.7; P < 0.001) but not with other outcomes (Figure 3). In a subgroup analysis of patients without preoperative chronic opioid therapy, high-preoperative opioid dosage was significantly associated with postoperative chronic opioid use (aOR, 4.93; 95% CI, 3.08–7.89; P < 0.001) but not with other outcomes (Figure 4).
In a sensitivity analysis with low, intermediate, and high-dose exposure categories; intermediate- and high-preoperative opioid dosages were significantly associated with postoperative chronic opioid use, but not with other outcomes (Supplemental Table 1, http://links.lww.com/BRS/B402). The proportion of patients presenting with preoperative chronic opioid therapy each year was not significantly different (Supplemental Table 2, http://links.lww.com/BRS/B402).
We analyzed the relationship between preoperative chronic opioid therapy and opioid dosage with long-term outcomes in 2128 patients undergoing elective spine surgery between January 2011 and February 2017 at a single academic center. After adjusting for preoperative pain and disability scores, depression, anxiety, and additional covariates, patients treated with preoperative chronic opioid therapy were at substantially higher risk of not achieving meaningful improvements in pain, function, or quality of life at 1 year; higher risk of dissatisfaction with surgery at 1 year; higher risk of 90-day complications; and higher risk of continued postoperative chronic opioid use. In contrast to prior studies, we found high-preoperative opioid dosage (>30 MME/d) to be significantly associated with postoperative chronic opioid use, but not with other outcomes.
Our results add to a rapidly growing body of literature surrounding the detrimental effects of opioids. In spine surgery specifically, the literature reports an association between opioids and poor outcomes but has thus far been limited to insurance or administrative claims data, small cohorts, or self-reported opioid dosages. For example, Lee et al10 collected self-reported preoperative opioid dosages and found each 10 mg increase to be associated with a corresponding decrease in patient-reported outcomes at 1 year after surgery. In another study using patient-reported preoperative opioid dosage, Wick et al11 found a daily MME of 29 as a threshold for achieving Minimal Clinically Important Difference for functional patient-reported outcomes after spine surgery. While these studies demonstrate an association with preoperative opioid dosage and outcomes, they utilize self-reported data and do not account for chronic opioid use. Two studies by Jain et al used commercial insurance data and reported that preoperative chronic opioid therapy was associated with cost and increased risk of postoperative complications after spinal fusion, but they were unable to report patient-reported outcomes.13,14 Furthermore, an analysis of opioid prescriptions dispensed in the United States in 2008 identified 4.8 of 19 million prescriptions were paid in cash, which will be missed by insurance database.24 Thus, a comprehensive analysis on the relationship of both preoperative chronic opioid therapy and preoperative opioid dosage with long-term patient-reported outcomes after spine surgery has yet to be published.
In a review of the literature, McAnally12 discussed the rationale and approach to preoperative opioid weaning as a component of preoperative optimization. Early reports of preoperative opioid weaning demonstrate promise. In a 2015 retrospective study of total joint arthroplasty patients, 41 chronic opioid users that successfully weaned their preoperative daily opioid dosage by 50% had improved outcomes compared with a matched cohort. But this study was relatively small and did not account for several potential confounders. Given the prevalence of opioid use among spine patients, opioid weaning has the potential for substantial improvements in clinical outcomes. Proper patient counseling and the development of effective weaning protocols rely on a detailed understanding of the effects of preoperative opioid dosage and duration of use.
The present study addresses a gap in the literature surrounding the association of preoperative chronic opioid therapy and dosage with long-term patient-reported outcomes after spine surgery. In a risk-adjusted analysis, we found preoperative chronic opioid therapy, but not preoperative opioid dosage, to be negatively associated with outcomes for pain, function, quality of life, and satisfaction at 1 year; and complications within 90 days. Both preoperative chronic opioid therapy and high-preoperative opioid dosage were independently associated with risk of chronic postoperative opioid use. These findings have important implications, as previous publications reporting a preoperative opioid dosage threshold for predicting outcomes were likely confounded by chronic opioid use. While our observational study does not establish a causal relationship, our results indicate that patients treated with chronic opioids in the 3 months prior to surgery have a substantially increased risk of poor clinical outcome, regardless of dosage.
A recent national survey found that 57% of patients presenting for spine surgery had some preoperative opioid exposure.8 Hence, preoperative opioid use is perhaps the most prevalent modifiable risk factor that spine practitioners encounter. As of 2018, 49 states have active PDMPs and thus identifying patients with consistent opioid use, regardless of dosage or type of opioid, can be achieved quickly in the clinic through a PDMP query. If short-term opioids are initiated for preoperative pain control, our results indicate that the dosage should be minimized to avoid persistent postoperative opioid use.
Our results should be interpreted in the context of certain limitations. First, using prescription data from our state's PDMP required exclusion of patients living outside the state, making external validity unclear. Additionally, this method provided opioid dispensing data rather than actual opioid use. However, we utilized the state PDMP rather than self-reported data, due to the availability, reproducibility, and because the state PDMP captures all opioid prescriptions dispensed regardless of payer, thus minimizing selection bias associated with insurance claims data. Next, we categorized continuous exposure and outcome variables for ease of interpretation, which may introduce bias and loss of information. While the true impact of duration of opioid use may lie on a continuum, we defined chronic opioid therapy using the common duration of 3 months for simplicity and clinical application. We categorized patients as “low-” or “high-dosage” based on previously published values,11 which risks misclassification bias. For this reason, we performed a sensitivity analysis with multiple opioid dosage categories to verify our results. Third, while we chose validated instruments for pain, function, and quality of life after spine surgery, there are no well-established improvement scores for what defines clinical success. Rather than using minimal clinically important difference (which represents the lowest clinically relevant change), we defined a positive clinical outcome using the recommended definition for “clinically meaningful improvement” of > 30% improvement for pain, function, and quality of life scores.21,25–28 Furthermore, percent change rather than change in raw-score better adjusts for baseline patient-reported outcome scores.29 Finally, we were unable to adjust for chronic pain syndromes, which risks unmeasured confounding. To minimize this bias, we used surrogate covariates for chronic pain including duration of symptoms, preoperative pain scores, anxiety, depression, and NDI scores.30,31
The present study clearly illustrates the association between preoperative chronic opioid therapy and poor 1-year clinical outcomes, 90-day complications, and postoperative chronic opioid use after elective spine surgery. Additionally, we found a preoperative opioid dosage > 30 MME per day to be associated with postoperative chronic opioid use, but not with other outcomes. We define a significant modifiable risk factor that can be identified rapidly through a state PDMP, and we provide robust outcome data that can be utilized for patient education. Given the prevalence of spine disorders and opioid use among this population, there is a need to develop effective strategies for weaning patients on chronic opioid therapy presenting for spine surgery. When considered in conjunction with the known dangers of opioids and strong evidence demonstrating no benefit to long-term opioids, our results should substantially narrow the indications for chronic opioid therapy in treating spine pathology.
1. Hoy D, March L, Brooks P, et al. The global burden of low back pain: estimates from the Global Burden of Disease 2010 study. Ann Rheum Dis
2. United States Bone and Joint Initiative: The Burden of Musculoskeletal Diseases in the United States (BMUS)], Fourth edition, forthcoming. Available at: http://www.boneandjointburden.org
. Accessed July 15, 2018.
3. 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
4. Dowell D, Haegerich TM, Chou R. CDC guideline for prescribing opioids
for chronic pain—United States, 2016. MMWR Recomm Rep
5. 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
6. Krebs EE, Gravely A, Nugent S, et al. Effect of opioid vs nonopioid medications on pain-related function
in patients with chronic back pain or hip or knee osteoarthritis pain: the SPACE Randomized Clinical Trial. JAMA
7. Hudson TJ, Edlund MJ, Steffick DE, et al. Epidemiology of regular prescribed opioid use: results from a national, population-based survey. J Pain Symptom Manage
8. Hilliard PE, Waljee J, Moser S, et al. Prevalence of preoperative opioid
use and characteristics associated with opioid use among patients presenting for surgery. JAMA Surg
9. Hedegaard H, Warner M, Minino AM. Drug overdose deaths in the United States, 1999-2016. NCHS Data Brief, no 294. Hyattsville, MD: National Center for Health Statistics. 2017/CDC.], 2016. Available at: http://wonder.cdc.gov
. Accessed July 15, 2018.
10. Lee D, Armaghani S, Archer KR, et al. Preoperative opioid
use as a predictor of adverse postoperative self-reported outcomes
in patients undergoing spine surgery
. J Bone Joint Surg Am
11. Wick JB, Sivaganesan A, Chotai S, et al. Is there a preoperative morphine equianalgesic dose that predicts ability to achieve a clinically meaningful improvement following spine surgery
12. McAnally H. Rationale for and approach to preoperative opioid
weaning: a preoperative optimization protocol. Perioper Med (Lond)
13. Jain N, Phillips FM, Weaver T, et al. Pre-operative chronic opioid therapy: a risk factor for complications
, readmission, continued opioid use and increased costs after one- and two-level posterior lumbar fusion. Spine (Phila Pa 1976)
14. Jain N, Brock JL, Phillips FM, et al. Chronic preoperative opioid
use is a risk factor for increased complications
, resource use, and costs after cervical fusion. Spine J
2018; [Epub ahead of print].
15. Kidner CL, Gatchel RJ, Mayer TG. MMPI disability
profile is associated with degree of opioid use in chronic work-related musculoskeletal disorders. Clin J Pain
16. Adogwa O, Elsamadicy AA, Han JL, et al. Do measures of surgical effectiveness at 1 year after lumbar spine surgery
accurately predict 2-year outcomes
? J Neurosurg Spine
17. Tosteson AN. Preference-based health outcome measures in low back pain. Spine (Phila Pa 1976)
18. Vernon H. The Neck Disability
Index: state-of-the-art, 1991-2008. J Manipulative Physiol Ther
19. Fairbank JC, Pynsent PB. The Oswestry Disability
Index. Spine (Phila Pa 1976)
20. Solberg TK, Olsen JA, Ingebrigtsen T, et al. Health-related quality of life assessment by the EuroQol-5D can provide cost-utility data in the field of low-back surgery. Eur Spine J
21. Ostelo RW, Deyo RA, Stratford P, et al. Interpreting change scores for pain and functional status in low back pain: towards international consensus regarding minimal important change. Spine (Phila Pa 1976)
22. Daltroy LH, Cats-Baril WL, Katz JN, et al. The North American spine society lumbar spine outcome assessment Instrument: reliability and validity tests. Spine (Phila Pa 1976)
23. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform
24. McDonald DC, Carlson KE. Estimating the prevalence of opioid diversion by “doctor shoppers” in the United States. PLoS One
25. Zannikos S, Lee L, Smith HE. Minimum clinically important difference and substantial clinical benefit: does one size fit all diagnoses and patients? Semin Spine Surg
26. Glassman SD, Copay AG, Berven SH, et al. Defining substantial clinical benefit following lumbar spine arthrodesis. J Bone Joint Surg Am
27. Carreon LY, Glassman SD, Campbell MJ, et al. Neck Disability
Index, short form-36 physical component summary, and pain scales for neck and arm pain: the minimum clinically important difference and substantial clinical benefit after cervical spine fusion. Spine J
28. Copay AG, Glassman SD, Subach BR, et al. Minimum clinically important difference in lumbar spine surgery
patients: a choice of methods using the Oswestry Disability
Index, Medical Outcomes
Study questionnaire Short Form 36, and pain scales. Spine J
29. Farrar JT, Pritchett YL, Robinson M, et al. The clinical importance of changes in the 0 to 10 numeric rating scale for worst, least, and average pain intensity: analyses of data from clinical trials of duloxetine in pain disorders. J Pain
30. van Hecke O, Torrance N, Smith BH. Chronic pain epidemiology and its clinical relevance. Br J Anaesth
31. Ris I, Barbero M, Falla D, et al. Pain extent is more strongly associated with disability
, psychological factors, and neck muscle function
in people with non-traumatic versus traumatic chronic neck pain: a cross sectional study. Eur J Phys Rehabil Med
2018; [Epub ahead of print].