Secondary Logo

Journal Logo


Opioid Use in Adults With Low Back or Lower Extremity Pain Who Undergo Spine Surgical Treatment Within 1 Year of Diagnosis

Fatemi, Parastou MD; Zhang, Yi BA; Ho, Allen MD; Lama, Roberto; Jin, Michael BS; Veeravagu, Anand MD; Desai, Atman MD; Ratliff, John K. MD

Author Information
doi: 10.1097/BRS.0000000000003663

The United States remains in the midst of an opioid crisis. In 2017, 1.7 million Americans abused prescription opioids and 47,000 died from opioid overdose.1,2 The 2016 Centers for Disease Control guidelines on prescribing opioids for chronic pain included 12 recommendations for primary care physicians to increase safe treatment of pain and reduce risk of long-term opioid therapy.3 Opioid therapy may be initiated in individuals with low back or lower extremity pain (LBP/LEP). When these patients undergo spine surgery, their preoperative exposure to opioids may influence postoperative opioid requirement.

The American Pain Society and the American Society of Anesthesiologists have concluded that, concerning post-surgical pain management after hospital discharge, “research on methods and outcomes of discharge planning and follow-up are scarce and insufficient to provide strong guidance on optimal methods.”4 Data are sparse regarding opioid dose-tapering for surgical patients.5 Some recommend a 20% to 25% dose reduction every 1 to 2 days for those on opioids for more than 1 to 2 weeks postoperatively.4

Spine patients are often prescribed opioids for postoperative pain control; according to the Dartmouth Atlas, spine surgery rates are steadily increasing.6 Neurosurgical and orthopedic surgeries are associated with the highest prevalence of chronic opioid use across all surgical specialties,7 and studies have identified several risk factors and negative consequences of long-term postoperative opioid use in spine patients.8–13

Using a nationally representative employer-based insurance claims database, we retrospectively followed a cohort of patients without prescription opioids for 6 months before newly diagnosed LBP/LEP who underwent surgery within 1 year after diagnosis. We then investigated patterns of opioid use in these patients before and after surgery. Surgeries included decompression procedures (e.g., laminotomy, laminectomy, osteotomy, or discectomy), in addition to arthrodesis with or without decompression. We further investigated health care utilization of different opioid use groups following surgery.


Study Design, Setting, Size, and Participants

For this retrospective cohort study, we used the MarketScan (Truven Health) inpatient, outpatient, and pharmaceutical claims databases. MarketScan captures approximately 50% of the US population with employer-sponsored insurance and 40% of retirees. MarketScan pharmaceutical data capture all covered prescriptions filled by a patient regardless of the prescriber or location of fill. MarketScan data are de-identified and received Institutional Review Board approval at our institution.

Enrolled patients were 18 years and older diagnosed with LBP/LEP between January 1, 2008 and December 31, 2015. Patients had no previous diagnosis of LBP/LEP for 1 year before index LBP/LEP diagnosis, were not on prescription opioids for at least 6 months before diagnosis, did not undergo thoracic or lumbosacral surgery within 1 year before diagnosis, and underwent thoracic or lumbosacral surgery within 1 year following diagnosis. Patients with any health care encounter for a red flag diagnosis (including cancer, infection, incontinence, foot drop, and cauda equina syndrome) in the 12 months before or after diagnosis were excluded.14 These criteria were selected to maximize the likelihood of capturing a homogenous cohort of surgical patients whose disease started with primarily pain-related symptoms, without neurologic compromise and unrelated to cancer or infection, who may have then received opioid prescription(s) after symptom onset Supplement Figure 1:

To limit surveillance bias, all patients had 12 months of continuous follow-up before LBP/LEP diagnosis, and after index surgery. Figure 1 demonstrates the study flow diagram. These patients comprise a subset of our previously published cohort.14

Figure 1:
Flow diagram of cohort. LBP indicates low back pain; LEP, lower extremity pain.

LBP/LEP was defined using International Classification of Diseases, Clinical Modification (ICD-9-CM, ICD-10-CM) (Supplementary Table 1a, Opioid prescription was defined as Drug Enforcement Administration's (DEA) controlled substances in category II or III. Specific opioid medications are listed in Supplementary Table 1b, Surgical procedures were defined by Current Procedural Terminology (CPT) codes (Supplementary Table 1c, Index surgery was defined as the earliest post-diagnosis claim for a service containing one of these procedure codes.


The primary outcome was postoperative high-frequency opioid prescription, defined as six or more opioid script fills in 12 months following index surgery, without consideration to strength or number of doses in each script. Although multiple definitions of opioid use patterns exist in the literature, six or more opioid script fills in a 1-year period has been associated with long-term opioid use and increase in hospitalization with opioid-related adverse events when compared with fewer than six scripts filled in a 1-year period.15,16 Individuals not meeting this criterion are referred to as low-frequency postoperative patients.

Covariates included age, sex, depression, comorbidities assessed using the Elixhauser index,17 state of residence, surgical type, and preoperative opioid use frequency. Preoperative opioid prescription frequency was defined similarly to postoperative opioid frequency (average of six or more opioid prescriptions fills in a 12-month period), but scaled to the time period between LBP/LEP diagnosis and index surgery. During 6 months before the LBP/LEP diagnosis, all patients had no opioid prescriptions.

Secondary outcomes included total healthcare cost within 12 months following index surgery, defined as total eligible charges including health plan and patient components. We analyzed 12-month postoperative outpatient and inpatient health care encounters across both cohorts (Supplementary Table 4,

Duration of postoperative opioid use was defined as the number of days between index surgery and last opioid prescription, plus the number of days covered by that last prescription. The last opioid prescription was defined as the final prescription patients received in the 1 year postoperative period. All patients had 1 year of follow-up after surgery and analysis of postoperative opioid use was capped at the 1-year mark. Patients who remained on opioids through the end of the 1-year mark were considered to have not terminated opioid use. Discontinuation of opioids was defined as occurring in the 1-year postoperative period if the final opioid prescription and the days supplied by it terminated before 1 year after surgery.

Morphine milligram equivalents (MME) were calculated using CDC dose conversions.18 Total preoperative and postoperative MMEs were calculated and divided by the number of coverage days for each time period to provide direct comparison of average daily MMEs before and after surgery.

Statistical Methods

χ2 test with continuity correction was used to compare binary variables, whereas student t test was performed to compare continuous variables. Association between various variables and postoperative prescription frequency was assessed using a generalized linear model. Results are presented as odds ratios (OR) and 95% confidence intervals (CI).

A multivariate linear regression model adjusting for the binary variables of age (<45 vs. ≥45) years, sex, and surgical type (instrumentation vs. decompression) was generated to assess the effect of high-frequency opioid use on total costs within 12 months following surgery. For the purposes of additional cost analysis, a cohort of all 2952 high-frequency opioid users matched to 2952 low-frequency opioid users was selected using propensity score matching to balance cohorts for age, sex, surgical type, depression, and common comorbidities (hypertension, diabetes, hypothyroidism, chronic pulmonary disorder, cardiac arrhythmia, and obesity).

Multivariate Cox regression was performed to assess time to discontinuation of opioid prescription among patients in high-frequency and low-frequency opioid cohorts, adjusting for age, sex, and surgical type. Results are presented as hazard ratios and 95% CI.

All analyses were completed in R.



A total of 25,506 patients were diagnosed with LBP/LEP and underwent surgery within 1 year of diagnosis. After LBP/LEP diagnosis, 18,219 (71.4%) were prescribed opioids, whereas 7287 (28.6%) were not. After surgery, 2952 (11.6%) were prescribed opioids with high frequency and 22,554 (88.4%) with low frequency. Tables 1 and 2 demonstrate the medical and demographic characteristics of the groups.

TABLE 1 - Baseline Characteristics of Patients Newly Diagnosed With Low Back or Lower Extremity Pain (LBP or LEP) Who Undergo Thoracic or Lumbosacral Spine Surgery Within 1 Year Following Diagnosis. Grouped by Initiation of Opioids After LBP or LEP Diagnosis and Before Surgery
No Preoperative Opioid Use (N = 7287) Low-frequency Preoperative Opioid Use (N = 5789) High-frequency Preoperative Opioid Use (N = 12,430) P
Age group in years, no. (%) <0.001
 18–34 703 (9.6%) 783 (13.5%) 1535 (12.3%)
 35–44 1183 (16.2%) 1217 (21.0%) 2955 (23.8%)
 45–54 1753 (24.1%) 1485 (25.7%) 3490 (28.1%)
 55–64 2212 (30.4%) 1593 (27.5%) 3305 (26.6%)
 ≥65 1436 (19.7%) 711 (12.3%) 1145 (9.2%)
Median age, y (interquartile range) 53.6 ± 14.3 50.0 ± 13.9 49.3 ± 12.7 <0.001
Female sex, no. (%) 3069 (42.1%) 2508 (43.3%) 4793 (38.6%) <0.001
Depression, no. (%) 6844 (93.9%) 5388 (93.1%) 11,517 (92.7%) 0.003
Elixhauser comorbidities, no. (%) <0.001
 0 3552 (48.7%) 3011 (52.0%) 6771 (54.5%)
 1 1950 (26.8%) 1485 (25.7%) 3171 (25.5%)
 ≥2 1785 (24.5%) 1293 (22.3%) 2488 (20.0%)
Common Elixhauser comorbidities, no. (%)
 Hypertension 2313 (31.7%) 1655 (28.6%) 3124 (25.1%) <0.001
 Diabetes 918 (12.6%) 638 (11.0%) 1337 (10.8%) <0.001
 Hypothyroidism 545 (7.5%) 435 (7.5%) 734 (5.9%) <0.001
 COPD 467 (6.4%) 376 (6.5%) 848 (6.8%) 0.475
 Cardiac arrythmia 340 (4.7%) 214 (3.7%) 386 (3.1%) <0.001
 Obesity 211 (2.9%) 169 (2.9%) 367 (3.0%) 0.973
Postoperative opioid use <0.001
 High frequency 442 (6.1%) 552 (9.5%) 1958 (15.8%)
 Low frequency 6845 (93.9%) 5237 (90.5%) 10,472 (84.2%)
Type of surgery, no. (%) <0.001
 Fusion with or without decompression, single level 964 (13.2%) 538 (9.3%) 763 (6.1%)
 Fusion with or without decompression, multilevel 1692 (23.2%) 925 (16.0%) 1151 (9.3%)
 Single-level decompression alone 3446 (47.3%) 3477 (60.1%) 8967 (72.1%)
 Multilevel decompression alone 1185 (16.3%) 849 (14.7%) 1549 (12.5%)

TABLE 2 - Baseline Characteristics of Patients Newly Diagnosed With Low Back or Lower Extremity Pain (LBP or LEP) Who Undergo Thoracic or Lumbosacral Spine Surgery Within 1 Year Following Diagnosis. Grouped by Opioids Use After Surgery
Characteristics High-frequency Postoperative Opioid Use (N = 2952) Low-frequency Postoperative Opioid Use (N = 22,554) P
Age group in years, no. (%) <0.001
 18–34 267 (9.0%) 2754 (12.2%)
 35–44 561 (19.0%) 4794 (21.3%)
 45–54 831 (28.2%) 5897 (26.1%)
 55–64 954 (32.3%) 6156 (27.3%)
 ≥65 339 (11.5%) 2953 (13.1%)
Median age, y (interquartile range) 53 (43–60) 51 (41–60)
Female sex, no. (%) 1329 (45.0%) 9041 (40.1%) <0.001
Depression, no. (%) 311 (10.5%) 1446 (6.4%) <0.001
Elixhauser comorbidities, no. (%) <0.001
 0 1312 (44.4%) 12,022 (53.3%)
 1 820 (27.8%) 5786 (25.7%)
 ≥2 820 (27.8%) 4746 (21.0%)
Common Elixhauser comorbidities, no. (%)
 Hypertension 941 (31.9%) 6151 (27.3%) <0.001
 Diabetes 382 (12.9%) 2511 (11.1%) 0.004
 Hypothyroidism 221 (7.5%) 1493 (6.6%) 0.084
 COPD 279 (9.5%) 1412 (6.3%) <0.001
 Cardiac arrythmia 105 (3.6%) 835 (3.7%) 0.732
 Obesity 645 (2.9%) 102 (3.5%) 0.081
Post diagnosis, preoperative opioid use, no. (%) <0.001
 High frequency 1958 (66.3%) 10,472 (46.4%)
 Low frequency 552 (18.7%) 5237 (23.2%)
 No opioid use 442 (15.0%) 6845 (30.3%)
Type of surgery, no. (%) <0.001
 Fusion with or without decompression, single level 452 (15.3%) 1813 (8.0%)
 Fusion with or without decompression, multilevel 839 (28.4%) 2929 (13.0%)
 Single-level decompression alone 1284 (43.5%) 14,606 (64.8%)
 Multilevel decompression alone 377 (12.8%) 3206 (14.2%)

Demographic Data

Among the postoperative high-frequency group, median age was 53 years (interquartile range [IQR] 43–60), compared to 51 years (IQR 41–60) among the postoperative low-frequency group (Table 1). The postoperative high-frequency cohort had a greater percentage of female patients (45.0% vs. 40.1%, P < 0.001).

Patients in the postoperative high-frequency cohort were more likely to have two or more comorbidities (27.8% vs. 21.0%, P < 0.001). Common comorbidities included hypertension, diabetes, depression, hypothyroidism, chronic pulmonary disease, cardiac arrhythmia, and obesity. Prevalence of depression (10.5% vs. 6.4%, P < 0.001), hypertension (31.9% vs. 27.3%, P < 0.001), diabetes (12.9% vs. 11.1%, P = 0.004), and chronic pulmonary disease (9.5% vs. 6.3%, P < 0.001) were higher among postoperative high-frequency opioid users.

The probability of postoperative high-frequency opioid use varied geographically from 4.3% to 20.0% (Figure 2). States with highest rates of high-frequency postoperative opioid use included Idaho (20.0%), Oklahoma (18.3%), Alaska (16.7%), Rhode Island (15.9%), and West Virginia (15.4%) (Supplementary Table 2, These patterns appeared to be state-specific, and we did not find significant differences regionally.

Figure 2:
Geographic distribution of postoperative high-frequency opioid prescription in previously patients undergoing spine surgery.

Outcome Data

On average, daily MME declined preoperatively to postoperatively, both in aggregate and within every surgical cohort (Figure 3A). This was most pronounced in the single-level decompression group where MME decreased from 3.42 to 0.51 (P < 0.001). In multilevel decompression, MME decreased from 1.76 to 0.44 (P < 0.001). Minimal but statistically significant decrease was seen in the fusion cohort: single level from 1.01 to 0.70 (P < 0.001), and multilevel from 1.31 to 1.11 (P < 0.001). Of note, total MME was averaged across the number of days in the preoperative and postoperative periods; thus, it underestimates the MME of each individual script. We also present the average daily MME of the last preoperative script compared with the active postoperative script 1 year after surgery in Figure 3B. It shows a similar decrease in MME.

Figure 3:
(A) Change in total MME, averaged over number of days before and after surgery, in relation to frequency of opioid prescriptions preoperatively and postoperatively. Each dot represents the change for an individual patient. Green dots represent a decrease in MME after surgery while a red dot represents an increase in MME after surgery. (B) Average daily MME of final prescription leading up to surgery compared with daily MME of opioid prescription at 1 year after surgery, by surgical type. MME indicates morphine milligram equivalent.

Even though surgery led to a decrease in MME on average, many patients remained on opioid prescriptions 12 months after surgery (Figure 4). At 3 months after surgery, only 3.9% of high-frequency and 65.4% of low-frequency patients had discontinued opioids. By 6 months, 14.8% of high-frequency and 73.6% of low-frequency patients had discontinued opioids. Finally, 1 year after surgery, 44.3% of high-frequency and 87.0% of low-frequency patients had discontinued opioids. Patients prescribed opioids with low-frequency postoperatively were 3.78 times as likely those prescribed opioids with high-frequency to discontinue opioids by 12 months postoperatively, after adjusting for age, sex, and surgery type (95% CI 3.59–3.99, P < 0.0001).

Figure 4:
(A) Kaplan–Meier curve of ongoing filled opioid prescriptions by patients after thoracic or lumbosacral spine surgery. (B) Kaplan–Meier curve of active opioid prescription following spine surgery in patients, stratified by preoperative opioid use frequency. high-high indicates high preoperative opioid use, high postoperative opioid use; high-low, high preoperative opioid use, low postoperative opioid use; low-high, low preoperative opioid use, high postoperative opioid use; low-low, low preoperative opioid use, low postoperative opioid use; none-high, no preoperative opioid use, high postoperative opioid use; none-low, no preoperative opioid use, low postoperative opioid use.

In multivariate analysis, the strongest predictors of postoperative high-frequency opioid prescription were preoperative high-frequency opioid prescription (OR 4.53, 95% CI 4.00–5.15) and fusion surgery (OR 4.41, 95% CI 4.01–4.84). Similarly, preoperative opioid use is strongly associated with discontinuation of opioids 1 year after surgery. Within the postoperative high-frequency cohort, patients with preoperative low-frequency opioid prescriptions were 1.23 times as likely to discontinue opioids as those who were prescribed opioids with high-frequency preoperatively (95% CI of risk ratio, 1.08–1.39). Patients without any preoperative opioid prescriptions were 1.40 times as likely to discontinue opioids as those who were prescribed opioids with high-frequency preoperatively (95% CI 1.23–1.61; Figure 4B).

On average, patients prescribed opioids with high-frequency filled their first postoperative opioid script earlier and had more days supplied by it than the low-frequency cohort (Supplementary Table 3,

The postoperative high-frequency cohort had more NSAID prescriptions and epidural steroid injections in both the preoperative and postoperative periods compared with the low-frequency cohort; however, physical therapy usage in the high-frequency group was lower in the preoperative setting and higher in the postoperative setting (Table 3).

TABLE 3 - Postoperative Opioid-prescribing Patterns in Relation to Other Pain Treatment Modalities, Including Physical Therapy, Epidural Steroid Injections, and NSAID Prescriptions Before and After Index Surgery
Postoperative High-frequency Opioid Use Patients (N = 2952) Postoperative Low-frequency Opioid Use Patients (N = 22,554) P
Time between LBP/LEP diagnosis and surgery
 Mean no. of PT visits ± SD 1.9 ± 4.8 2.3 ± 5.1 0.001
 Underwent ESI 1294 (43.8%) 7956 (35.3%) <0.001
 Mean no. of ESI ± SD 0.9 ± 1.2 0.7 ± 1.0 <0.001
 Mean no. of NSAID prescriptions filled ± SD 1.0 ± 1.6 0.7 ± 1.3 <0.001
One year after surgery
 Mean number of PT visits ± SD 4.3 ± 8.7 2.9 ± 6.6 <0.001
 Underwent ESI 511 (17.3%) 952 (4.2%) <0.001
 Mean number of ESI ± SD 0.3 ± 0.8 0.1 ± 0.4 <0.001
 Mean number of NSAID prescriptions filled ± SD 1.5 ± 2.4 0.6 ± 1.6 <0.001
ESI indicates epidural steroid injection; LBP, lower back pain; LEP, lower extremity pain; NSAID, non-steroidal anti-inflammatory drug; PT, physical therapy; SD, standard deviation.

Utilization Data

Within 12 months after surgery, patients prescribed opioids with high-frequency had more combined health care encounters than low-frequency patients (12 vs. 6, P < 0.0001; Supplementary Table 4,, and incurred higher health care costs ($36,883 vs. $19,756, P < 0.0001; Table 4) even after propensity score matching for age, sex, surgical type, depression, and common comorbidities. The high-frequency cohort (11.6% of total study population) accrued 19.6% of the total 12-month postoperative costs of the combined cohorts (Table 4). Services that increased costs included imaging, PT, clinic visits, emergency department / urgent care visits, and repeat spine surgery (Supplementary Table 5,

TABLE 4 - Twelve-month Postoperative Costs in Patients Newly Diagnosed With Back or Lower Extremity Pain and Opioid-naïve for at Least 6 Months Before Diagnosis, Who Undergo Thoracic or Lumbosacral Spine Surgery Within 1 Year Following Diagnosis
Cost in US Dollars Postoperative High-frequency Opioid Use Patients (N = 2952) Postoperative Low-frequency Opioid Use Patients (N = 22,554) P
Unadjusted per patient, mean (SEM) $36,883 (555) $19,756 (189) <0.0001
Adjusted per patient, mean (SEM) $23,311 (490) $13,771 (511) <0.0001
Adjusted, matched, per patient, mean (SEM) $23,553 (864) $14,099 (1527) <0.0001
Total cost, no. (% of total cost across cohorts) $108,878,321 (19.6%) $445,585,846 (80.4%)
Multivariate linear regression model was used.
Adjustment variables were age (<45, ≥45), sex, and surgery type (fusion, decompression only).
A total of 2952 postoperative high-frequency opioid use patients were matched to 2952 postoperative low-frequency opioid use patients. Patients were matched for age (<45, ≥45), sex, and surgery type (fusion, decompression only).


Key Results, Interpretation, and Generalizability

Opioid overprescribing after surgery is associated with subsequent opioid use disorder and diversion,19–22 yet guidelines on postoperative outpatient pain management do not exist. The issue is especially challenging in patients with LBP/LEP who may initiate opioids before undergoing surgery. We present the first national descriptive study of opioid use in adult patients with LBP/LEP who undergo spine surgery. The strengths of this study lie in the large cohort size and the use of pharmaceutical claims which captured opioid prescriptions patients received from all health care providers.

The literature contains a wide range of postoperative opioid prescribing rates and use. In a study of commercially insured patients undergoing lumbar fusion, 29.3% remained on prescription opioids 12 months after surgery,8 which is higher than our finding of 15.6%. In a study of patients undergoing anterior cervical discectomy and fusion at 2 academic centers, 9.9% were on opioids at 3 to 6 months postoperatively.10 A study of patients undergoing lumbar fusion in a workers’ compensation setting found that 57.4% remained on opioids 1 year after surgery.13 In a study of opioid-naïve veterans who underwent lumbar decompression or fusion surgery, a minimal number used opioids 6 months after surgery (0.1%).12 The variations in opioid prescription duration and usage rates following surgery underscore the many factors influencing opioid use in the postoperative setting. Patient demographics, surgical type/intensity, and preoperative opioid use all contribute to postoperative opioid use.

Here, the strongest predictor of postoperative high-frequency opioid status is frequency of preoperative opioid prescription. Additionally, preoperative frequency predicts long-term opioid use following surgery in a stepwise fashion. Within the postoperative high-frequency group, the likelihood of ongoing active opioid prescription 12 months after surgery increased progressively depending on frequency of use preoperatively (none, low, high). Previous studies have also identified preoperative opioid use as a predictor of postoperative opioid use,10,13,21 and as a predictor of increased surgical complications, 90-day readmissions, and increased costs.11 The increased use of other pain treatment modalities among the high-frequency group suggests that this group has higher overall pain management needs.

This study also shows that MME on average decreased postoperatively compared with preoperatively, and the effect is modified by surgery type. Patients who underwent decompression alone had greater decrease in daily MME compared to those undergoing arthrodesis. It is conceivable that escalating pain may lead to increase in opioid use preoperatively and may also lead physicians to intervene surgically. Additionally, the modification of effect by surgical type may possibly reflect the different pain characteristic of the patients undergoing decompression alone versus those requiring arthrodesis.

This study and others show that initial opioid prescribing patterns impact chronic use in nonsurgical and surgical patients.16,23,24 Patients who had their first outpatient postoperative opioid prescription filled sooner and for a longer period have a greater incidence of becoming high-frequency and long-term users. Although we cannot discern the clinical reasoning for these prescribing practices with the present study, prescribers should consider exhausting other modalities to address pain postoperatively.25,26

Our analysis shows geographic variability in rates of postoperative opioid prescription. The underlying causes for these differences were not investigated here. Evidence suggests that policy changes may lead to more prudent opioid prescribing practices. Through supplementary analysis we found that following creation of guidelines aimed at reducing misuse of prescription opioids in 2013 in Oklahoma, the yearly proportion of high-frequency postoperative opioid cohort decreased in that state, but this change was not statistically significant (Supplementary Figure 2, In Rhode Island, statutory limits on opioid prescribers successfully reduced opioid prescription following lumbar surgery and did not increase ED visits or hospital readmissions.27


Our study has several constraints related to using a health care claims dataset. We lack granular clinical information including pain levels and rationale for opioid prescription. Claims-based databases may underestimate overall patient comorbidities. Accuracy of procedural coding is reliant upon accuracy of physician billing. Additionally, this study population was comprised of employed individuals with steady insurance coverage; thus, there is likely inadequate representation of socioeconomically vulnerable demographics at high risk for sustained opioid use. We used filled drug prescriptions as a surrogate for opioid use, and the correlation between usage and prescription is not known. This approach fails to directly measure patient consumption and does not capture possible diversion.

Our study shows a stepwise association between opioid use after LEP/LBP diagnosis and frequency and duration of opioid prescriptions after surgery. Simultaneously, the strength of prescriptions as measured by daily MME decreased following surgery. Most patients were prescribed opioids before surgery and many continued these medications long after surgery; thus, different providers are likely involved in prescribing opioids at different times. Additionally, we note great variability across states in frequency of opioid prescription following surgery. Judicious opioid prescribing in this patient population, therefore, requires a consensus-based approach that includes public health officials, primary care, surgeons, anesthesiologists, and pain management specialists.

Key Points

  • As the frequency of opioid prescriptions to patients with LBP or LEP increases, so does their risk of frequent and chronic opioid use after spine surgery.
  • As measured by average daily MME, patients’ prescribed opioid dosage was higher before surgery than following surgery over 1 year; this difference was further modified by surgical type.
  • Patients with postoperative high-frequency opioid prescriptions utilize other pain management modalities (e.g., physical therapy, NSAID use, and epidural steroid injections) in higher proportion than patients with low-frequency opioid prescriptions.
  • Patients who are prescribed opioids with higher frequency postoperatively incur, on average, higher 1-year postoperative health care cost.
  • There exists wide variability in opioid prescription following spine surgery across states in the United States.


1. CDC/NCHS. CDC/NCHS, National Vital Statistics System, Mortality. In Department of Health and Human Services CfDCaP ed. Atlanta, GA, 2018.
2. (CBHSQ) CfBHSaQ. 2017 National Survey on Drug Use and Health: Detailed Tables. In Administration SAaMHS ed. Rockville, MD, 2018.
3. Dowell D, Haegerich TM, Chou R. CDC guideline for prescribing opioids for chronic pain—United States, 2016. MMWR Recomm Rep 2016; 65:1–49.
4. Chou R, Gordon DB, de Leon-Casasola OA, et al. Management of postoperative pain: a clinical practice guideline from the American Pain Society, the American Society of Regional Anesthesia and Pain Medicine, and the American Society of Anesthesiologists’ Committee on Regional Anesthesia, Executive Committee, and Administrative Council. J Pain 2016; 17:131–157.
5. Gordon DB, de Leon-Casasola OA, Wu CL, et al. Research gaps in practice guidelines for acute postoperative pain management in adults: findings from a review of the evidence for an American Pain Society Clinical Practice Guideline. J Pain 2016; 17:158–166.
6. Care DAoH. Spine Surgery: A Report by the Dartmouth Atlas of Health Care, CMS - FDA Collaborative: Center for the Evaluative Clinical Sciences, 2006.
7. Jiang X, Orton M, Feng R, et al. Chronic opioid usage in surgical patients in a large academic center. Ann Surg 2017; 265:722–727.
8. Connolly J 3rd, Javed Z, Raji MA, et al. Predictors of long-term opioid use following lumbar fusion surgery. Spine (Phila Pa 1986) 2017; 42:1405–1411.
9. Boscarino JA, Rukstalis M, Hoffman SN, et al. Risk factors for drug dependence among out-patients on opioid therapy in a large US health-care system. Addiction (Abingdon, England) 2010; 105:1776–1782.
10. Karhade AV, Ogink PT, Thio Q, et al. Machine learning for prediction of sustained opioid prescription after anterior cervical discectomy and fusion. Spine J 2019; 19:976–983.
11. Vakharia RM, Donnally CJ 3rd, Rush AJ 3rd, et al. Comparison of implant survivability in primary 1- to 2-level lumbar fusion amongst opioid abusers and non-opioid abusers. J Spine Surg (Hong Kong) 2018; 4:568–574.
12. Schoenfeld AJ, Nwosu K, Jiang W, et al. Risk factors for prolonged opioid use following spine surgery, and the association with surgical intensity, among opioid-naive patients. J Bone Joint Surg Am 2017; 99:1247–1252.
13. Anderson JT, Haas AR, Percy R, et al. Chronic opioid therapy after lumbar fusion surgery for degenerative disc disease in a workers’ compensation setting. Spine (Phila Pa 1976) 2015; 40:1775–1784.
14. Kim LH, Vail D, Azad TD, et al. Expenditures and health care utilization among adults with newly diagnosed low back and lower extremity pain. JAMA Netw Open 2019; 2:e193676.
15. Morden NE, Munson JC, Colla CH, et al. Prescription opioid use among disabled Medicare beneficiaries: intensity, trends, and regional variation. Med Care 2014; 52:852–859.
16. Deyo RA, Hallvik SE, Hildebran C, et al. Association between initial opioid prescribing patterns and subsequent long-term use among opioid-naive patients: a statewide retrospective cohort study. J Gen Intern Med 2017; 32:21–27.
17. Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care 1998; 36:8–27.
18. National Center for Injury Prevention and Control. CDC compilation of benzodiazepines, muscle relaxants, stimulants, zolpidem, and opioid analgesics with oral morphine milligram equivalent conversion factors, 2018 version], 2018. Available at:
19. Neuman MD, Bateman BT, Wunsch H. Inappropriate opioid prescription after surgery. Lancet (London, England) 2019; 393:1547–1557.
20. Sharma M, Ugiliweneza B, Aljuboori Z, et al. Factors predicting opioid dependence in patients undergoing surgery for degenerative spondylolisthesis: analysis from the MarketScan databases. J Neurosurg 2018; 29:271–278.
21. Kalakoti P, Hendrickson NR, Bedard NA, et al. Opioid utilization following lumbar arthrodesis: trends and factors associated with long-term use. Spine (Phila Pa 1976) 2018; 43:1208–1216.
22. Wright AK, Sikora M, Leveque JC. Characterizing the risk of long-term opioid utilization in patients undergoing lumbar spine surgery. Spine (Phila Pa 1976) 2020; 45:E54–E60.
23. Ge DH, Hockley A, Vasquez-Montes D, et al. Total inpatient morphine milligram equivalents can predict long term opioid use after transforaminal lumbar interbody fusion. Spine (Phila Pa 1976) 2019; 44:1465–1470.
24. Shah A, Hayes CJ, Martin BC. Characteristics of initial prescription episodes and likelihood of long-term opioid use—United States, 2006-2015. MMWR Morb Mortal Wkly Rep 2017; 66:265–269.
25. Devin CJ, McGirt MJ. Best evidence in multimodal pain management in spine surgery and means of assessing postoperative pain and functional outcomes. J Clin Neurosci 2015; 22:930–938.
26. Scully RE, Schoenfeld AJ, Jiang W, et al. Defining optimal length of opioid pain medication prescription after common surgical procedures. JAMA Surg 2018; 153:37–43.
27. Reid DBC, Shah KN, Ruddell JH, et al. Effect of narcotic prescription limiting legislation on opioid utilization following lumbar spine surgery. Spine J 2019; 19:717–725.

chronic; low back; lower extremity; opioid; outpatient; pain; postoperative; spine

Supplemental Digital Content

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.