Secondary Logo

Journal Logo

Clinical Investigations

Opioid Utilization in Geriatric Patients After Operation for Degenerative Spine Disease

Nguyen, Anthony V. BA*; Ross, Evan MD; Westra, Jordan MPH; Huang, Nicole BS*; Nguyen, Christine Y. BS*; Raji, Mukaila MD§; Lall, Rishi MD; Kuo, Yong-Fang PhD

Author Information
Journal of Neurosurgical Anesthesiology: October 2021 - Volume 33 - Issue 4 - p 315-322
doi: 10.1097/ANA.0000000000000682

Abstract

The opioid epidemic continues to worsen; the Centers for Disease Control and Prevention reports that approximately 47,600 Americans died from opioid-related toxicity in 2017, up from 30,000 deaths in 2015.1 In parallel, the number of geriatric opioid-related overdose deaths has increased by 65% from 2014 to 2017.2 Major contributors to the epidemic include overprescribing, tolerance and dependence, and diversion.3,4 Because the surgical specialties account for ∼10% of all opioid prescriptions, there is a critical need to understand opioid use patterns after the operation.5

Pain, whether acute or chronic (lasting >3 to 6 mo), is the primary indication for prescription of opioids.6 Chronic pain is highly prevalent in community-dwelling individuals over the age of 65, affecting 28% to 39% of older adults.7,8 Degenerative spine disease (DSD) occurs in >90% of adults over the age of 50 years and can be a major source of chronic pain and focal neurological deficits.9,10 In older adults aged 65 years and above, low back pain is the most common cause of pain and disability.9

Spinal surgery is routinely performed for severe DSD, especially when there are neurosurgical indications such as neurological deficits or severe radiculopathy.11 Nearly half of lumbar spine surgeries are performed on patients aged over 65 years, and can even be considered for some patients over the age of 85.12,13 The severity of the disease is the most important predictor of cervical spinal surgical success, and older patients have more severe stenosis as well as higher rates of multilevel disease.14 Surgical management can potentially alleviate back pain or worsen it. Because back pain is a risk factor for opioid abuse, it is necessary to investigate the association of spinal surgery and opioid utilization.15

In addition, following any surgery and regardless of age, there is a risk for chronic opioid use, with between 0.1% and 9.9% of patients using opioids 1-year postoperatively.16–19 This further complicates the association between spine surgery and opioid utilization. Some studies have also identified older age as a risk factor for chronic opioid use following surgery.16,20–23 Thus, given the associations of DSD with chronic back pain, the rate of therapeutic failure, and their age, the effectiveness of spinal surgery in geriatric patients with DSD may be suboptimal.

Geriatric patients may have such advanced DSD that spinal surgery does not result in therapeutic success, leading to chronic opioid utilization.14,15 Chronic opioid utilization in this population can be potentially more harmful than chronic utilization in a younger cohort given the age-related decrease in drug metabolism and excretion, so understanding risk factors for continued utilization is critical.24 In addition, because DSD is highly prevalent in this elderly population, and spine surgery is readily performed for geriatric patients, there is a need to investigate postspine surgery opioid utilization to understand risk factors for persistent use. Therefore, the primary goal of the study was to calculate the proportion of patients who continuously filled prescriptions for opioid painkillers up to 1-year after spinal surgery for DSD. In addition, we sought to determine what patient characteristics, such as demographic data or preoperative opioid utilization, were associated with prolonged opioid utilization.

MATERIALS AND METHODS

In this retrospective cohort study, we analyzed opioid use after surgery using a national 5% Medicare sample database randomly selected by the Centers for Medicare and Medicaid Services (CMS) based on the eighth and ninth digits (05, 20, 45, 70, 95) of their health insurance claim number. This standard database available for research purposes has been shown to be representative of the whole cohort.25 We included all patients aged 66 years and older who underwent an anterior cervical discectomy and fusion, posterior cervical fusion, 360-degree cervical fusion, lumbar microdiscectomy, lumbar laminectomy, anterior lumbar fusion, posterior lumbar fusion, or 360-degree lumbar fusion for a DSD-related diagnosis in the years 2008 to 2014 (Table S1, Supplementary Digital Content 1, http://links.lww.com/JNA/A263: Procedures, procedural codes, associated diagnoses). If claims were filed for multiple lumbar procedures on the same day, they were classified by the more invasive procedure (fusion>laminectomy>microdiscectomy).26 If a patient underwent multiple different procedures during the study period, the earliest procedure performed chronologically was considered to be the index operation. We excluded patients who were missing data, were enrolled in a Health Maintenance Organization, did not have continuous Medicare Parts A, B, and D coverage in the 12 months before surgery, or who underwent both lumbar and cervical surgery on the same date. All opioid prescriptions filled by a patient in the 12 months before and following surgery were recorded. Opioid utilization was defined as filling a prescription for an opioid painkiller. The National Drug Code (NDC) codes utilized can be found in the Supplementary Material (Table S2, Supplementary Digital Content 2, http://links.lww.com/JNA/A264: National drug code list for opioids). The University of Texas Medical Branch Institutional Review Board approved the research.

Variables of interest included patient sex, race, age (below 70, 70 to 74, 75 to 79, above 80 y), Medicaid dual eligibility, Charlson Comorbidity Index (0, 1, 2, 3+), geographic region and county characteristic (urban or rural), surgeon specialty, and procedure year. We also considered preoperative opioid utilization, benzodiazepine prescriptions, alcohol or other drug abuse, anxiety, depression, and diagnosis of cancer in the 12 months before the surgery. Active cancer treatment during the 12 months before or after surgery was also included. Comorbidities, such as drug abuse or psychiatric conditions, were determined by analyzing International Classification of Diseases, Ninth Revision (ICD-9) codes in other claims filed for the patient in the 12 months presurgery. Information about codes used for covariates are available in the Supplementary Material (Table S3, Supplementary Digital Content 3, http://links.lww.com/JNA/A265: Codes or source of codes for covariates in the main analysis of the study).

Preoperative opioid utilization was stratified by quarters with use, with opioid-naivety being defined as no prescriptions filled in days 5 to 365 before surgery. Patients who filled at least one prescription in each quarter leading up to surgery were noted as having 4 quarters with opioid use.

The Kaplan-Meier method was used to estimate the rate of continuing opioid prescription filling after surgery. Continuous opioid use was deemed related to the surgery only if a prescription was filled within 30 days of the end of the previous prescription’s intended supply duration or within 30 days of the surgery. For example, if a 90-day supply of medication was prescribed and filled on the day of surgery, the filling of another prescription before the 121st postoperative day was classified as continued use related to the operation. If the previously mentioned patient did not fill another prescription, their opioid utilization was labeled as 90 days of use. Patients who underwent outpatient surgery and did not fill a prescription within 30 days of surgery were considered to have no use. Because hospital-administered medications were not included in the prescription claims data, patients were considered to be utilizing opioids each day they were a hospital inpatient. Prescriptions filled in the 5 days before surgery but continued beyond the surgery date were counted for the number of days of supply intended past the date of surgery.

Patients were censored from the time-to-event analysis if they experienced trauma; a traumatic event was considered to be independent of surgery and likely to constitute a new prescription for opioids. However, additional surgery was considered to be potentially a result of surgical failure, so any more opioid utilization arising from additional surgery was considered to be related to the original operation. Patients were also censored if they lost Medicare coverage, enrolled in a Health Maintenance Organization, or died in the 12 months following surgery, whichever occurred first. Traumatic injuries were identified using the American College of Surgeons National Trauma Database definitions: ICD-9 codes 800 to 959.9, excluding 905 to 909.9, 910 to 924.9, and 930 to 939.9; ICD-10 codes S00 to S99 with a seventh character modifier of A, B, or C; T07, T14, and T20-T28 with a seventh character modifier of A; codes T30 to T32 and T79.A1 to T79.A9 with a seventh character modifier of A; excluding S00, S10, S20, S30, S40, S50, S60, S70, S80, and S90. The incidence of new-onset chronic opioid use, defined as a previously opioid-naive patient continuing to fill prescriptions 1-year postoperatively, was also calculated.

Cox proportional hazards regression was used to calculate hazard ratios of the effect of covariates on opioid use. The model included all covariates previously mentioned. We included State as a random effect to account for differences in opioid prescribing laws between States. The proportional hazards assumption for selected covariates (surgery type, prior opioid use, and procedure year) was examined by the interaction between the log of follow-up time and the covariate. When the assumption was violated, the time-dependent effect of a covariate up to a specific month and after that month was included in the model. The specific month was decided by visual inspection of hazard plots.

A sensitivity analysis was performed in the same manner as above, using Cox proportional hazards regression and including all covariates in the adjusted models. The sensitivity analysis used total preoperative morphine milligram equivalents starting 5 days before surgery going back 1 year, instead of stratifying preoperative opioid use by quarters. The analysis was also performed using the criterion set by Soneji et al18 (no greater than a 90-day gap between prescription fill dates) to examine the difference in the incidence of new-onset chronic opioid utilization between the definitions. All statistical analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC). All P-values were 2-sided with <0.05 considered statistically significant.

RESULTS

A total of 14,583 patients in the national 5% Medicare sample database met study criteria (Fig. 1). The number of surgeries performed, and patient characteristics are shown in the Supplementary Material (Table S4, Supplementary Digital Content 4, http://links.lww.com/JNA/A266: Demographic data).

FIGURE 1
FIGURE 1:
Flow diagram outlining the formation of the study cohort. Patients were required to be above 66 years of age because of the requirements (1) old age being the reason for Medicare entitlement, and: (2) 12 months continuous enrollment before the surgery date.

The majority of patients utilized opioids preoperatively (Table S4, Supplementary Digital Content 4, http://links.lww.com/JNA/A266: Demographic data). Median opioid use was 11 days for patients who had an anterior cervical discectomy and fusion, posterior cervical fusion, lumbar microdiscectomy, or lumbar laminectomy (Fig. 2). The median utilization for all other operations was at least 19 days. At 1-year postoperation, 6.0% (95% confidence interval [CI]: 5.6-6.5) of all patients in the study continued to fill prescriptions for opioids. Disregarding surgery type, 0.3% (95% CI: 0.2-0.6) of all opioid-naive patients continued to utilize opioids 1-year after surgery. Opioid utilization through 90 days after surgery is shown in the Supplementary Material (Fig. S1, Supplementary Digital Content 5, http://links.lww.com/JNA/A267: Proportion of patients with continued opioid utilization stratified by procedure; Fig. S2, Supplementary Digital Content 6, http://links.lww.com/JNA/A268: Proportion of patients with continued opioid utilization following cervical spine surgery; Fig. S3, Supplementary Digital Content 7, http://links.lww.com/JNA/A269: Proportion of patients with continued opioid utilization following lumbar spine surgery). The following proportion of patients with preoperative opioid use continued their use at 1-year after surgery: 0.7% (95% CI: 0.5-1.1) of patients with any previous use in 1 quarter, 2.0% (95% CI: 1.4-2.8) of patients with use in 2 quarters, 4.9% (95% CI: 3.7-6.2) of patients with use in 3 quarters, and 23.6% (95% CI: 21.9-25.4) of patients with 4 quarter use (Fig. 3). The definition used by Soneji et al18 resulted in an overall chronic, continuous utilization rate of 7.6% (95% CI: 7.2-8.0) at 1-year and a rate of 0.9% (95% CI: 0.1-1.2) for opioid-naive patients.

FIGURE 2
FIGURE 2:
Proportion of patients with continued opioid utilization following spine surgery, stratified by the procedure. There was no significant difference at 1-year after surgery. Patients were censored from the Kaplan-Meier analysis if they experienced trauma, lost Medicare coverage or died. ACDF indicates anterior cervical discectomy and fusion; PCF, posterior cervical fusion.
FIGURE 3
FIGURE 3:
Proportion of patients with continued opioid utilization following spine surgery, stratified by duration of preoperative opioid utilization. The duration of preoperative opioid utilization was significantly associated with continued utilization. Patients were censored from the Kaplan-Meier analysis if they experienced trauma, lost Medicare coverage or died.

Surgery type did not meet the proportional hazards assumption, so time-dependent analysis of hazard for surgery type in 2 separate time intervals postoperatively (<180 d, ≥180 d) was reported. A few variables were significantly associated with prolonged opioid utilization at 1-year after surgery: prior benzodiazepine use (hazard ratio [HR]: 0.93; 95% CI: 0.87-0.98), anxiety (HR: 0.89; 95% CI: 0.85-0.94), prior opioid use, and Medicaid dual eligibility (HR: 0.90; 95% CI: 0.83-0.97) (Table 1). The association of surgery (anterior lumbar fusion, posterior lumbar fusion, and 360-degree lumbar fusion compared with anterior cervical discectomy and fusion) with prolonged opioid use was observed only within the first 180 days after surgery; there was no significant difference at 1-year. The number of preoperative quarters with prior opioid use was the strongest predictor of prolonged utilization: patients who used opioids during all 4 quarters before surgery were 68% more likely than opioid-naive patients to have continued opioid utilization at any point during the study period. Variables associated with shortened duration of opioid utilization included Asian race (compared with white, HR: 1.23; 95% CI: 1.05-1.45), and increased age (greatest difference seen in 80+ years age group compared with below 70 y, HR: 1.32; 95% CI: 1.25-1.40). The results of the sensitivity analysis can be found in the Supplementary Material (Table S5, Supplementary Digital Content 8, http://links.lww.com/JNA/A270: Adjusted relative hazard of ceasing opioid utilization; sensitivity analysis using preoperative total milligram morphine equivalents) and are largely similar to the main analysis except that female sex is associated with prolonged utilization and benzodiazepine use is not significantly correlated with duration of opioid utilization. Time-to-event analysis for the definition used by Soneji et al18 is also shown in the Supplementary Material (Table S6, Supplemental Digital Content 9, http://links.lww.com/JNA/A271: Adjusted relative hazard of ceasing opioid utilization; sensitivity analysis using 90-day gap between prescriptions for continuous use) and is largely similar to both analyses, but depression (HR: 0.94; 95% CI: 0.89-0.99), a Charlson comorbidity score of 2 (HR: 0.93; 95% CI: 0.88-0.99), and active cancer treatment after surgery (HR: 0.86; 95% CI: 0.75-0.98) were significant.

TABLE 1 - Adjusted Relative Hazard of Ceasing Opioid Utilization Among Geriatric Patients Undergoing Spinal Surgery for Degenerative Spine Disease-related Diagnoses
Variable Hazard Ratio 95% Confidence Interval
Surgery (days 1-179 postoperative)*
 ACDF Reference Reference
 PCF 0.91 0.81-1.01
 360-degree cervical fusion 0.84 0.67-1.05
 Lumbar microdiscectomy 1.03 0.97-1.10
 Lumbar laminectomy 1.00 0.93-1.07
 Lumbar fusion, posterior approach 0.80 0.75-0.84
 Lumbar fusion, anterior approach 0.74 0.60-0.92
 360-degree lumbar fusion 0.68 0.62-0.74
Surgery (days 180-365 postoperative)*
 ACDF Reference Reference
 PCF 1.47 0.66-3.25
 360-degree cervical fusion 1.38 0.33-5.80
 Lumbar microdiscectomy 1.02 0.62-1.69
 Lumbar laminectomy 1.29 0.76-2.17
 Lumbar fusion, posterior approach 1.48 0.97-2.25
 Lumbar fusion, anterior approach 0.92 0.22-3.88
 360-degree lumbar fusion 1.38 0.76-2.49
Sex
 Male Reference Reference
 Female 0.97 0.93-1.00
Race
 White Reference Reference
 Black 1.08 0.98-1.19
 Asian 1.23 1.05-1.45
 Hispanic 1.15 0.95-1.39
 American Indian/Alaska Native 1.00 0.73-1.36
 Other 1.19 1.00-1.42
 Unknown 1.26 0.91-1.74
Age group (y)
 <70 Reference Reference
 70-74 1.04 0.99-1.09
 75-79 1.16 1.10-1.22
 80+ 1.32 1.25-1.40
Region
 WE Reference Reference
 MW 1.04 0.98-1.09
 NE 1.06 0.99-1.12
 SO 0.99 0.95-1.04
Surgeon specialty
 Neurosurgery Reference Reference
 Other 1.02 0.98-1.06
Charlson Comorbidity Index
 0 Reference Reference
 1 1.00 0.96-1.04
 2 0.99 0.94-1.05
 3+ 1.03 0.97-1.09
Benzodiazepine use† 0.93 0.87-0.98
Alcohol abuse† 0.93 0.77-1.13
Drug abuse† 0.99 0.85-1.15
Anxiety† 0.89 0.85-0.94
Depression† 0.99 0.95-1.04
History of cancer† 0.96 0.88-1.05
Active cancer treatment before surgery† 1.02 0.91-1.13
Active cancer treatment after surgery‡ 0.93 0.82-1.06
Medicaid eligible at surgery 0.90 0.83-0.97
Quarters with preoperative opioid use§
 No use Reference Reference
 1 quarter 0.85 0.81-0.89
 2 quarters 0.74 0.70-0.78
 3 quarters 0.58 0.54-0.62
 4 quarters 0.32 0.31-0.34
Procedure year
 2008 Reference Reference
 2009 1.01 0.94-1.09
 2010 1.00 0.93-1.08
 2011 0.96 0.89-1.03
 2012 0.94 0.88-1.01
 2013 1.02 0.95-1.10
 2014 0.97 0.91-1.04
Urban/rural
 Rural Reference Reference
 Metro 1.03 0.94-1.14
 Urban, nonmetro 0.99 0.90-1.10
*All hazards ratios were calculated over the at-risk period, which was defined as 1-year postoperatively for this study except in the case of surgery type because it did not meet the proportional hazards assumption when the at-risk period was considered the year following surgery.
†Measured in the 12 months before surgery
‡Measured in the 12 months following surgery
§Measured in days 5 to 365 before surgery. Quarter 1 was defined as days 6 to 95 days before surgery, quarter 2 was defined as days 96 to 185 before surgery, etc.
ACDF indicates anterior cervical discectomy and fusion; PCF, posterior cervical fusion.

DISCUSSION

Previous studies have utilized cross-sectional analyses to determine rates of chronic opioid use after surgery. We instead performed a time-to-event analysis to track continued postoperative opioid utilization. Using this approach in Medicare beneficiaries undergoing spinal surgery for DSD, we found that 0.3% of opioid-naive patients continuously used opioids up to 1-year after surgery, while 6.0% of patients overall continuously used opioids at 1 year. Duration of preoperative opioid use strongly correlated with prolonged postoperative utilization: 23.6% of patients with preoperative opioid use in all 4 yearly quarters continuously filled prescriptions during the year following spine surgery.

The 0.3% incidence rate of new-onset chronic opioid use following spinal surgery among opioid-naive patients is similar to those of Sun et al16 (0.1% to 1.4% among nongeriatric adult patients who underwent 1 of 11 nonspinal surgeries), Soneji et al18 (0.4% in geriatric patients who underwent 1 of 6 major elective nonspinal surgeries), and Schoenfeld et al27 (0.02% among nongeriatric adult patients with TRICARE insurance who underwent spine surgery). Importantly, Sun et al16 excluded codeine from their analysis, a decision that may potentially account for the differences in observed rates of use. Meanwhile, Clarke et al17 (major nonspinal surgery) and Alam et al19 (low-risk nonspinal surgery) reported much higher incidence rates of 3.1% and 7.7%, respectively, in patients aged 66+ years undergoing nonspinal surgeries.

Sun et al16 defined chronic use as >10 prescriptions or >120 days of use, Clarke et al17 and Alam et al19 defined prolonged utilization as a prescription between 91-180 and 305-425 days after surgery, respectively. Notably, each of these approaches is limited by the possibility that the opioids observed in the record were being prescribed for reasons unrelated to the index operation. Although our definition was similar to the one used by Soneji et al,18 who defined cessation as the absence of an opioid prescription for 90 days, we further refined that definition to account for the putative duration of opioid therapy that was prescribed, similar to the method used by Schoenfeld et al.27 We analyzed our cohort utilizing the method employed by Soneji et al,18 which was similar to ours but did not account for opioid supply, and this resulted in a rate of 0.9% among opioid-naive patients. Although the results are similar, additional studies comparing the accuracy and utility of various methodologies are necessary.

Along with censoring patients who experienced a traumatic injury, the requirement of use without prolonged gaps (as employed in the time-to-event analysis but not cross-sectional analysis) provides greater confidence that the observed opioid utilization is directly related to the surgery. In addition, cross-sectional analysis requires patients to have the continuous enrollment to Medicare throughout the entire study period to observe whether opioids were utilized. Patients who lose coverage or die in the study period are thus excluded. In the time-to-event analysis, these patients can be included and are censored when such an event occurs; this allows for their opioid utilization during the study period to be included in the analysis, reducing the selection bias associated with required enrollment.

The low incidence rate of new-onset chronic opioid utilization is reassuring and suggests that surgery per se does not contribute significantly to the development of chronic opioid use. For patients with preoperative opioid utilization, there may even be a reduction in continuous utilization rate; the continuous utilization rate at 1-year for patients with the heaviest preoperative opioid use was 23.6%.

Our results are consistent with some previous findings but also reveal that some risk factors for continued opioid utilization in the geriatric population differ from risk factors observed in younger adult populations. Although more intense surgery has been reported to be associated with greater opioid utilization, this correlation was not observed in our population at 1-year after surgery.27 However, lumbar fusions were associated with prolonged opioid utilization within the first 6 months after surgery, which is expected as a result of acute surgical pain. In our study, preoperative opioid use was the strongest predictor of continued use.28–30 The relationship between anxiety, benzodiazepines, and opioids has also been reported previously in the literature.16,17,31 Contrary to studies reporting older age as a risk factor for opioid dependence,16,20–23 younger geriatric patients in our study were more likely to continuously utilize opioids. Various studies have reported that gender, Charlson Comorbidity Index, previous alcohol or drug abuse, history of depression, and living in a rural county are associated with prolonged opioid use, associations we did not find.16,17,27–30,32 Also interestingly, although opioid-related deaths and hospitalizations have increased over time, the year of operation was not significant in a time-to-event analysis.1,33

Furthermore, the presence of additional independent predictors suggests that a multiplicity of different reasons may explain why patients continue to use opioids. The association of Asian race with decreased utilization could reflect less physician willingness to prescribe opioids to Asian peoples or may reflect cultural attitudes.34,35 Our results indicated that older age was associated with decreased duration of opioid use, which may also be a result of physician unwillingness to prescribe, although the association between age and opioid utilization is controversial.16,17,29 The significance of Medicaid dual-eligibility is consistent with previous studies’ results and raises the question about potential socioeconomic factors such as the incentives that may drive opioid diversion.32,36

Additional prospective studies should be considered to elucidate whether the relationship between preoperative opioid use and chronic postoperative use is a simple correlation or causation. The cases of prolonged preoperative utilization may reflect more severe disease or may result in opioid tolerance, opioid dependence, or increased sensitivity to pain. The possibility of surgical failure in alleviating chronic pain due to DSD in older patients must be considered. If opioid utilization is a reflection of disease severity, then interventions aimed at earlier identification and referral of patients with DSD should be developed. However, if utilization is resulting in dependence, then interventions aimed at reducing dependence would be highly beneficial. One possible avenue for future research is testing the effectiveness of preoperative opioid-weaning protocols on postoperative opioid use. Mixed modality pain control regimens should be studied as an alternative to opioid utilization in this population. For example, perioperative ketorolac, regional anesthesia such as epidural catheters with ropivacaine, and consultation with an acute pain service have all been shown to decrease opioid consumption.37–39 Other research topics could include the prospective evaluation of preoperative frailty as a predictive factor, or employment of different databases to compare age-stratified cohorts. Nongeriatric patients enrolled in Medicare reflect a disabled population that is intrinsically different from those whose reason for entitlement is old age, which prevents comparison of younger adults against older adults enrolled in Medicare.

There are several important limitations of this study. These data are not indicative of opioid consumption and do not provide clarity on the indication for the opioid prescriptions, nor do the data reflect patient pain level. The inability of an administrative claims database to provide information on indications for opioid prescriptions, pain level, and quality of life necessitates additional prospective studies which may help to delineate which definition is most accurate when assessing chronic opioid utilization following surgery. Patients without Part D coverage were also excluded, so this analysis cannot be considered representative of all Medicare patients. Moreover, State regulations on opioids were not controlled for. Information about prescribers was not included, and the data do not reflect the use of opioids obtained through means outside of provider prescriptions. In addition, the determination of risk factors present in the 12 months before surgery is dependent on patients seeing providers and on claims data, so it is possible that not all conditions or diseases were entirely captured by the database. Another important limitation is that the criteria set for preoperative and postoperative opioid utilization were different. This was necessary given that DSD is a slowly developing process, and a definitive timepoint cannot be set for when the disease began. However, the index operation occurred on a set date, and continuous utilization can be evaluated after that timepoint. As a result, this methodology does not allow for a direct evaluation of changes in utilization.

CONCLUSIONS

On the basis of our time-to-event analysis approach, spinal surgery for DSD is not a significant contributor to the development of chronic opioid use. However, the fact that 23.6% of patients with 4 quarters of preoperative opioid use continuously used opioids up to 1-year after surgery is alarming and may be indicative of inadequate trialing of nonopioid alternatives, opioid use disorders, advanced DSD, or inappropriate overprescription of opioids for noncancer pain. Geriatric patients, particularly at risk for chronic opioid utilization following spinal surgery, are those who use opioids preoperatively, use benzodiazepines, have a diagnosis of anxiety, are younger (particularly those aged 66 to 74 y), and are Medicaid dual-eligible. Although clinician awareness of these risk factors can help guide opioid-reduction strategies at the patient level, we also believe there is a need to trial perioperative opioid-sparing interventions and address opioid prescribing practices.

ACKNOWLEDGMENT

The authors would like to acknowledge and thank Dr. Sarah Toombs Smith for proofreading our manuscript and providing insightful editing suggestions.

REFERENCES

1. Scholl L, Seth P, Kariisa M, et al. Drug and opioid-involved overdose deaths—United States, 2013–2017. MMWR Morb Mortal Wkly Rep. 2019;67:1419–1427.
2. Rudd R, Seth P, David F, et al. Increases in drug and opioid-involved overdose deaths—United States, 2010–2015. MMWR Morb Mortal Wkly Rep. 2016;65:1445–1452.
3. McDonald DC, Carlson KE. Estimating the prevalence of opioid diversion by “doctor shoppers” in the United States. PLoS One. 2013;8:e69241.
4. Kharasch ED, Brunt LM. Perioperative opioids and public health. Anesthesiology. 2016;124:960–965.
5. Levy B, Paulozzi L, Mack KA, et al. Trends in opioid analgesic-prescribing rates by specialty, U.S., 2007-2012. Am J Prev Med. 2015;49:409–413.
6. Volkow ND, McLellan AT. Opioid abuse in chronic pain—misconceptions and mitigation strategies. N Engl J Med. 2016;374:1253–1263.
7. Johannes CB, Le TK, Zhou X, et al. The prevalence of chronic pain in United States adults: results of an internet-based survey. J Pain. 2010;11:1230–1239.
8. Kennedy J, Roll JM, Schraudner T, et al. Prevalence of persistent pain in the U.S. adult population: new data from the 2010 National Health Interview Survey. J Pain. 2014;15:979–984.
9. Teraguchi M, Yoshimura N, Hashizume H, et al. Prevalence and distribution of intervertebral disc degeneration over the entire spine in a population-based cohort: the Wakayama Spine Study. Osteoarthritis Cartilage. 2014;22:104–110.
10. Alshami AM. Prevalence of spinal disorders and their relationships with age and gender. Saudi Med J. 2015;36:725–730.
11. Eliyas JK, Karahalios D. Surgery for degenerative lumbar spine disease. Dis Mon. 2011;57:592–606.
12. Wang MY, Widi G, Levi AD. The safety profile of lumbar spinal surgery in elderly patients 85 years and older. Neurosurg Focus. 2015;39:E3.
13. Kalff R, Ewald C, Waschke A, et al. Degenerative lumbar spinal stenosis in older people: current treatment options. Dtsch Arztebl Int. 2013;110:613–624.
14. Nagata K, Ohashi T, Abe J, et al. Cervical myelopathy in elderly patients: clinical results and MRI findings before and after decompression surgery. Spinal Cord. 1996;34:220–226.
15. Burcher KM, Suprun A, Smith A. Risk factors for opioid use disorders in adult postsurgical patients. Cureus. 2018;10:e2611.
16. Sun EC, Darnall BD, Baker LC, et al. Incidence of and risk factors for chronic opioid use among opioid-naive patients in the postoperative period. JAMA Intern Med. 2016;176:1286–1293.
17. Clarke H, Soneji N, Ko DT, et al. Rates and risk factors for prolonged opioid use after major surgery: population based cohort study. BMJ. 2014;348:g1251.
18. Soneji N, Clarke HA, Ko DT, et al. Risks of developing persistent opioid use after major surgery. JAMA Surg. 2016;151:1083–1084.
19. Alam A, Gomes T, Zheng H, et al. Long-term analgesic use after low-risk surgery: a retrospective cohort study. Arch Intern Med. 2012;172:425–430.
20. Papaleontiou M. Outcomes associated with opioid use in the treatment of chronic non-cancer pain among older adults: a systematic review and meta-analysis. J Am Geriatr Soc. 2010;58:1353–1369.
21. Campbell CI, Weisner C, LeResche L, et al. Age and gender trends in long-term opioid analgesic use for noncancer pain. Am J Public Health. 2010;100:2541–2547.
22. Parsells Kelly J, Cook SF, Kaufman DW, et al. Prevalence and characteristics of opioid use in the US adult population. Pain. 2008;138:507–513.
23. Holman JE, Stoddard GJ, Higgins TF. Rates of prescription opiate use before and after injury in patients with orthopaedic trauma and the risk factors for prolonged opiate use. J Bone Joint Surg Am. 2013;95:1075–1080.
24. Smith HS. Opioid metabolism. Mayo Clin Proc. 2009;84:613–624.
25. Mues KE, Liede A, Liu J, et al. Use of the Medicare database in epidemiologic and health services research: a valuable source of real-world evidence on the older and disabled populations in the US. Clin Epidemiol. 2017;9:267–277.
26. Mirza SK, Deyo RA, Heagerty PJ, et al. Development of an index to characterize the “invasiveness” of spine surgery: validation by comparison to blood loss and operative time. Spine (Phila Pa 1976). 2008;33:2651–2661; discussion 2662.
27. 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.
28. Connolly J III, Javed Z, Raji MA, et al. Predictors of long-term opioid use following lumbar fusion surgery. Spine (Phila Pa 1976). 2017;42:1405–1411.
29. Pugely AJ, Bedard NA, Kalakoti P, et al. Opioid use following cervical spine surgery: trends and factors associated with long-term use. Spine J. 2018;18:1974–1981.
30. Armaghani SJ, Lee DS, Bible JE, et al. Preoperative opioid use and its association with perioperative opioid demand and postoperative opioid independence in patients undergoing spine surgery. Spine (Phila Pa 1976). 2014;39:E1524–E1530.
31. Skurtveit S, Furu K, Bramness J, et al. Benzodiazepines predict use of opioids—a follow-up study of 17,074 men and women. Pain Med. 2010;11:805–814.
32. Goodwin JS, Kuo Y, Brown D, et al. Association of chronic opioid use with presidential voting patterns in us counties in 2016. JAMA Netw Open. 2018;1:e180450.
33. Barnett ML, Olenski AR, Jena AB. Opioid-prescribing patterns of emergency physicians and risk of long-term use. N Engl J Med. 2017;376:663–673.
34. Mossey JM. Defining racial and ethnic disparities in pain management. Clin Orthop Relat Res. 2011;469:1859–1870.
35. Pletcher MJ, Kertesz SG, Kohn MA, et al. Trends in opioid prescribing by race/ethnicity for patients seeking care in us emergency departments. JAMA. 2008;299:70–78.
36. Kuo YF, Raji MA, Chen NW, et al. Trends in opioid prescriptions among part D Medicare recipients from 2007 to 2012. Am J Med. 2016;129:221.e21–221.e30.
37. De Oliveira GS Jr, Agarwal D, Benzon HT. Perioperative single dose ketorolac to prevent postoperative pain: a meta-analysis of randomized trials. Anesth Analg. 2012;114:424–433.
38. Gottschalk A, Freitag M, Tank S, et al. Quality of postoperative pain using an intraoperatively placed epidural catheter after major lumbar spinal surgery. Anesthesiology. 2004;101:175–180.
39. Said ET, Sztain JF, Abramson WB, et al. A dedicated acute pain service is associated with reduced postoperative opioid requirements in patients undergoing cytoreductive surgery with hyperthermic intraperitoneal chemotherapy. Anesth Analg. 2018;127:1044–1050.
Keywords:

analgesics; opioids; epidemiology; geriatrics; neurosurgery; osteoarthritis; spine

Supplemental Digital Content

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