Opioid Utilization Following Lumbar Arthrodesis

Trends and Factors Associated With Long-Term Use

Piyush Kalakoti, MD; Nathan R. Hendrickson, MD; Nicholas A. Bedard, MD; Andrew J. Pugely, MD


Spine. 2018;43(17):1208-1216. 

In This Article


The rise in opioid abuse has become a national epidemic with daily news of illicit and prescription misuse. With governing bodies, both at federal and state-level, racing to implement new policies, an evidence-based approach is needed to guide change.[1,2,23] In this study, a commercial claims dataset was used to evaluate perioperative opioid utilization in patients undergoing lumbar arthrodesis. We found that nearly one-third patients to be on prescription opioids before surgery, with immediate postoperative prescription peaking to 60% and 83% (within 30 days following surgery) in ON and OUs. Our analysis showed preoperative use, as the strongest predictor of prolonged post-operative use, with nearly 50% using compared with 9% of ON patients using at 1 year after surgery. This epidemiological data will significantly aid providers to develop new protocols to limit long-term use. Several of these findings merit further discussion.

According to our results, a high number of patients (58%) with lumbar spine conditions were considered chronic OUs (> 3 months). In a study similar to our current investigation, Schoenfeld et al.[24] analyzed the TRICARE claims database and noted a higher proportion (84%) of opioid prescription filling rate at discharge following lumbar arthrodesis and discectomies, comparable to 1-month postoperative rate of 83% in OUs noted in our analysis. On the contrary, a lower proportion of narcotic filling rate was noted at 6 months compared with our study (0.1% vs. 45% in OUs, 9.6% in ONs).[24] These findings provide baseline estimates for future studies to assess atypical postoperative opioid requirements in patients undergoing lumbar arthrodesis.

In addition, we found preoperative opioid use as a major driver of prolonged postoperative consumption. We specifically found preoperative opioid use to be associated with over four-fold increase in the absolute risk ratio of using narcotics at 1 year following lumbar arthrodesis. Several studies have noted the adverse impact of preoperative opioid use on functional outcomes and morbidities following orthopedic[15–19,21,25] and spine surgery.[21,26–29] In workers' compensation patients that underwent lumbar arthrodesis, chronic opioid use negatively impacted their return to work and was associated with poor outcomes (higher revision rates and costs).[27] Lee et al.[21] elucidated the negative effect of preoperative opioid use following spine surgery on self-reported outcomes as assessed using EuroQol-5D (EQ-5D), 12-item Short-Form Health Survey (SF)12, and Oswestry Disability Index. In a recent investigation involving patients undergoing TLIFs, Villavicencio et al.20 noted preoperative OUs had significantly greater disability, higher VAS scores for low back pain, and fared poorly on physical component and mental component summary domains than ON patients at 1-year follow up. Interestingly, opioid dosing did not impact these outcomes. On the contrary, Armaghani et al.[26] quantified the association of preoperative morphine usage with length of hospital stay (LOS), translating to a 1.1 day increase in LOS for every 100 preoperative morphine equivalents. Similar impact of preoperative opioid use with higher costs and reoperation rates is noted in patients undergoing cervical arthrodesis.[29]

Accepting the negative effects to both the patient and health care system of chronic opioid use, this study has defined some modifiable risk factors to limit prolonged postoperative use. Most notably, we found over 40% of chronic preoperative users still were filling opioid prescriptions 12 months after surgery. In our experience, this prolonged opioid use after spine fusion surgery may be largely inappropriate. A very recent study previously discussed using the TRICARE database corroborates these findings.[24]

Although our study noted opioid use within 3 months before lumbar arthrodesis to be associated with 1-year narcotic usage, Connolly et al.[30] extended their analysis of preoperative opioid use to over 250 days before lumbar arthrodesis and identified it as a risk factor for prolonged postoperative use. Correlating the findings from the literature on OUs incurring unfavorable outcomes with our confirmatory assessment of preoperative opioid use as a major driver for postoperative narcotic consumption, our study underscores the relevance of preoperative weaning strategies in spine patients to optimize outcomes and minimize the illicit exchange of these medications.

In addition to perioperative opioid use, comorbidities also have an impact on outcomes and prolonged postoperative opioid consumption. Psychiatric comorbidities including depression and anxiety are linked with poor outcomes and lower propensity for postoperative opioid discontinuation.[24,30–32] Confirming these findings, our analysis noted depression/anxiety disorders as well as substance abuse (alcohol, drug abuse) to be strongly associated with prolonged postoperative narcotic use following lumbar arthrodesis, reflecting the relevance of comprehensive psychiatric consultations including counseling for substance abuse during preoperative visits.

Previous studies have demonstrated that surgery magnitude such as discectomies and decompression compared with lumbar interbody fusions[24] and revision surgeries[30] have been associated with opioid use. In this study, only lumbar fusion procedures were examined. Although there were some small differences in the ORs of the risk factors predicting prolonged narcotic use between the anterior and posterior lumbar fusion cohorts, we do not believe there is much clinical relevance to this difference. Despite this small variation, pre-op use remained a consistently top risk factor.

Despite the study findings, we do not advocate for elimination of opioids within spine surgery. Opioid overuse should not be confused with appropriate use. In spinal deformity patients, for example, opioids use has shown remarkable improvements in controlling postoperative pain.[33] While our study findings do not negate the plausible positive effect of narcotics in spine patients for pain management/control, we recommend appropriate patient selection, counseling, and regulation tailored upon individual needs to prevent potential abuse.

Limitations governing the use of administrative databases apply to the present investigation.[34] Coding inaccuracies including underreporting of events cannot be ruled out; however, the estimated bias is known to be minimal and is unlikely to impact our derived conclusions.[35] Although the database will incorporate more granular ICD-10 definitions in the coming years, our current analysis was based upon the broad-based ICD-9 coding definitions for diseases/procedures. Further, our analysis was limited by quantification of opioid usage, indications, behavioral patterns, and frequency of opioid consumption; however, unlike other administrative databases and single-institutional studies, it provides valuable overall insights into trends in pre and postoperative narcotic consumption. Although risk factors associated with opioid use were identified, more granular clinical information was not available: radiological parameters, indications for surgery, exact levels of fused, approach, patient-reported quality of life measures, and objective measures for opioid screening.[36] Lastly, our data account for associations of long-term narcotic use and like other observational studies, cannot predict causality. Despite several limitations, the Humana database provides a large sample size across various settings and enables longitudinal tracking of patients undergoing lumbar arthrodesis, including evaluation of monthly opioid prescription refill rates at multiple timelines.