Abstract
Context:
Opioids may be prescribed for the short-term management of acute-onset back pain in the setting of trauma or for long-term management of chronic back pain. More than 50% of regular opioid users report taking them for back pain.
Objective:
To investigate whether surgical intervention reduces opioid requirements by patients taking opioids for back pain and whether there is a difference between county and managed care hospitals in this postoperative reduction of opioid requirement.
Methods:
A retrospective medical record review of 118 patients who underwent elective lumbar fusion at 4 hospitals (2 county hospitals and 2 managed care hospitals) was conducted. Opioid requirements before and after surgical intervention and at the 30-day outpatient follow-up were evaluated.
Results:
Forty medical records were included in the study. An overall decrease in opioid use was found in the postoperative follow-up phase after lumbar fusion in both the county and managed care hospitals. This reduction was statistically significant at 3 of 4 hospitals (P<0.01). When the data were pooled by facility type, the significance remained for county facilities (P<.01) but not managed care facilities (P=.18). Moreover, there was a significant decrease in opioid use during the postoperative inpatient phase for county compared with managed care facilities (P=.0427). The pain rating reported by patients during the hospital stay was significantly higher at county compared with managed care hospitals (P=.0088); however, the difference at discharge was not significant (P=.14).
Conclusion:
Our study indicates that lumbar fusion is associated with a significant decrease in opioid use (P<.05) compared with nonsurgical management. Overall, the difference in decreased opioid use between county and managed care hospitals after lumbar fusion was not significant.
Back pain is a major contributor to the opioid crisis; more than 50% of patients with chronic pain cite back pain as one of their reasons for starting opioid therapy.
1-3 Opioids may be prescribed for the short-term management of acute-onset back pain in the setting of trauma or for long-term management of chronic back pain.
Opioid use and abuse has been exponentially increasing in the United States since 2000.
4 A Centers for Disease Control and Prevention (CDC) review of drug-related deaths and overdoses from 2000-2014 found that the death rate from all drug overdoses increased by 137% and the rate of opioid-related deaths increased by over 200%.
4 By 2016, the number of opioid-related deaths was 5 times higher than it was in 2000.
5 In October 2017, the United States declared the opioid epidemic a national public health emergency.
Back pain can be stratified into acute, subacute, and chronic.
6 Patients with acute and subacute back pain may benefit from short-term opioid therapy; however, patients with chronic back pain and radicular symptoms for more than 16 weeks
7 may need longer-term opioid therapy. The purpose of the current study was to evaluate whether surgical intervention for chronic back pain and radicular pain can reduce opioid use and whether surgical intervention has an effect on opioid use at a county vs managed care hospital.
Four hospitals (2 county and 2 managed care) were enrolled in this retrospective review of a prospectively collected database. Patient medical records from 2016 to 2017 were reviewed from each hospital during 2017. Authorization was obtained from the institutional review board of each hospital. Inclusion criteria were age 18 years or older and elective lumbar fusion for degenerative back and radicular pain lasting more than 16 weeks. Levels fused were defined as the number of disc spaces traversed by the surgery, eg, fusion of L4 and L5 was defined as 1 level fused. Patients who required fusions because of traumatic injuries to the spine and patients taking opioids for other causes were excluded. Patients with chronic pain conditions that affected the nonaxial skeleton, such as fibromyalgia, rheumatoid arthritis, complex regional pain syndrome, and gout, and patients with previous back operations were also excluded.
The average daily preoperative opioid use was calculated on the basis of patient-reported opioid medication use or the prescriptions written in the medical record by the primary care or pain management physicians and further verified using the Controlled Substance Utilization Review and Evaluation System (CURES) database to ensure that no other sources of opioid medications were found. Pain medication use in the postoperative inpatient phase was retrieved using the medication administration record. An average of the daily reported pain was also assessed; the perioperative pain rating was evaluated on the 1 to 10 Numeric Pain Rating Scale from the electronic medical record. The patients’ opioid use at their initial 15- to 30-day postoperative outpatient follow-up was assessed on the basis of patient-reported pain rating, prescriptions written on hospital discharge, and patient report of medication use.
All opioid use was converted to oral morphine milligram equivalents (MMEs)
8 for data analysis and direct comparison of the various opioid medications. All patient data were stratified by hospital for initial
t test analysis. Next, the 2 county hospitals and the 2 managed care hospitals were pooled and compared for differences in opioid use to assess for any socioeconomic differences. Additional factors such as nonopioid therapies, number of levels being fused, and hospital length of stay were also reviewed.
Significant values were defined as P<.05. Our primary end point was to determine whether lumbar fusion had a positive or negative effect on opioid use. As a secondary end point, we sought to compare the similarities and differences between the 2 patient populations of county (uninsured or underinsured) and managed care (insured) patients.
Back and radicular pain are multifaceted, and an appropriate workup needs to be completed to determine whether a patient is an appropriate surgical candidate. When patients initially present with back pain, a detailed history and physical examination should be completed to assess for possible causes. If anatomical injury to the spine is suspected, computed tomographic imaging can be used to assess for bony pathologic changes or magnetic resonance imaging for compression of neural elements. In many cases, conservative care in the form of anti-inflammatory analgesics, physical therapy, and bracing are sufficient.
6,9,10 However, if a patient does not respond to conservative treatments or begins to demonstrate signs of neurologic deficits, surgical consultation should be sought.
Our retrospective dataset consisted of all patients undergoing elective lumbar fusion for back and radicular pain with symptoms lasting greater than 16 weeks. Contrary to the belief that surgical procedures can increase pain medication requirements, our study demonstrates that surgical intervention can play a crucial role in reducing the level of opioid use for degenerative disease of the lumbar spine. While opioid requirements, measured in MME, increased in the acute postoperative phase at all hospitals as expected, they tapered down to below preoperative requirements at outpatient follow-up, between 14 and 30 days postoperatively.
Our primary end point was to determine whether surgery had a positive or negative effect on opioid use. The use of opioid medication decreased at all 4 hospitals (
Figure) and was statistically significant at 3 of the 4 hospitals. Our secondary end point was to compare the socioeconomic differences between county and managed care populations. In a pooled data analysis comparing opioid use by hospital type—county vs managed care—the county hospitals maintained a statistically significant reduction in opioid requirements in the outpatient postoperative phase compared with the managed care hospitals (
Table 2). Multiple factors may have influenced this finding; increased preoperative opioid use and dose in managed care patients may have been related to prolonged conservative management. Most patients in a managed care setting have easier access to health care and are given a protocol-driven treatment regimen. The initial pain improvement with conservative treatment, often due to opioid use, may delay surgical referrals—a study
11 of total knee replacements found that 54% of patients delayed surgical intervention by 2 years secondary to opioid use. Similar findings may be seen in patients with back pain who may opt to postpone spine surgery; long-term nonsurgical management can lead to opioid tolerance and dependence.
12
The statistically significant difference (
P=.0427) in opioid use during the postoperative inpatient phase (
Table 5) between county and managed care hospitals may have been related to patient satisfaction ratings.
13-15 These hospital ratings based on patient satisfaction scores, although subjective, affect reimbursement rates
15,16; private insurers are adopting such policies.
14,15 The increased use of opioids in the managed care hospitals can likely be explained by the hospitals’ aim to make patients happier through treating postoperative pain more aggressively with opioid medication rather than nonopioid alternatives. The added benefit of opioid-induced euphoria may serve as a secondary bonus for higher patient satisfactory scores. This area needs further study, but a health care system driven by patient satisfaction may be putting patients at increased risk for opioid addiction.
The average number of vertebrae fused at each hospital was similar, ranging from 1.2 to 2.4 (
Table 3). The difference was not substantial but was approximately 1 level greater at the county hospitals. The addition of a fusion level leads to a larger incision and may explain the increase in average number of inpatient days as well as increased pain reported by patients at the county compared with the managed care hospitals. Given a larger number of levels fused at the county hospitals, one would expect an increased use of pain medications in the acute inpatient postoperative phase. However, the contrary was found. Another confounding factor that could have contributed to length of stay, particularly in the county hospitals, is the difficulty with disposition in those who may require rehabilitation services.
Another factor to be cognizant of is the use of nonopioid medications as adjuvant therapy for pain control. There were a variety of muscle relaxers and anti-inflammatory medications used in the postoperative phase; these medications varied to a slight degree among our patients and was dependent on the hospitals’ formulary and surgeon preference. Future studies may be done to monitor and standardize these medications and find optimal combinations that may work better than others. Many surgeons are cautious to use certain nonsteroidal anti-inflammatory drugs (NSAIDs) because of their antiplatelet effects, which can cause postoperative hematomas; furthermore, their anti-inflammatory properties can inhibit bony fusion.
17 At county hospital A, all patients were receiving a continuous infusion of ketorolac tromethamine (120 mg at 10 mL/hour for 50 hours) in the immediate postoperative phase. This hospital also had the greatest reduction in opioid use of the 4 hospitals studied. Although the data for NSAID use in the immediate postoperative phase is controversial,
17-19 our data show that NSAIDs may limit opioid use and therefore can be considered in the immediate postoperative period when excessive bleeding is not an issue. Icing, manipulation, and early mobilization have also been shown to be effective for managing low back pain.
9
The differences in opioid use between the county and managed care hospital patients studied highlights an important difference in the access to appropriate management of back pain. From this perspective, patients in a county hospital system seemingly are at a disadvantage because of the limited resources and availability for specialty care at county facilities. However, our data appear to support a more conservative approach to opioid use in the management of back pain vs the current approach used by managed care hospitals. The nonopioid medications used before and after an operation seem to have contributed a significant positive effect in reducing opioid use and should be further studied to replicate this finding.
A limiting factor in this study is the patients’ self-reporting of preoperative and outpatient postoperative pain medication use. We used data from the prescriptions given to the patients and their reported medication use to gather these data, and the data were confirmed using the CURES database to ensure that no other opioid prescriptions had been written. Opioid use was strictly monitored during the inpatient phase using the medication administration record; however, this information was not available in the outpatient phases of this study given its retrospective nature. In future studies, a stricter regimen may be formulated for a prospective trial so that this information can be kept in a medication administration record. Moreover, the possibility of patients using unprescribed opioid medications or recreational drugs during the outpatient phases of this study was unable to be regulated.
Another limiting factor was the use of various nonopioid medications. Given that multiple hospitals were involved and that this was a retrospective study, we were unable to ensure that all patients received the same adjuvant therapies. In future studies, we hope to standardize the adjuvant medications so that any potential synergistic analgesic effects can be standardized.
In addition, the opioids used in this study were varied; we attempted to rectify this factor by converting all of the medications to oral MMEs. However, the quality of the opioids used and the possibility that some formulations were combined with acetaminophen may have affected their analgesic effects. A related confounding variable was the lack of information with regard to the duration of opioid therapy before the operation. A longer duration of therapy may make it more difficult to wean patients off opioids. Given that this was a retrospective medical record review, we were unable to ascertain this variable.