Opioid consumption patterns after lumbar microdiscectomy or decompression


Francis Lovecchio (Hospital for Special Surgery [HSS], New York, USA) recently gave a presentation on opioid consumption patterns after lumbar microdiscectomy or decompression at the 19th Annual Conference of the International Society for the Advancement of Spinal Surgery (ISASS; 3–5 April, Anaheim, USA). Here, for Spinal News International, he gives a summary of the talk.

In an effort to ensure adequate pain control and a functional recovery, providers have tended to overprescribe narcotics in the post-discharge period.1,2 In turn, patients are often left with an excess of pills, which may become available for diversion or increase the risk of misuse. Excess pain medication in the home is a key driver of the current opioid crisis.3 In patients who become dependent on prescription opioids, spinal conditions are the most common diagnoses for the first prescription, putting spine surgeons on the front lines of the opioid epidemic.4 While opioid prescribing guidelines have been shown to reduce overprescribing, there is a paucity of data on the amount of opioid to provide after spine surgery.5 Determining the “minimum necessary quantity” to provide after specific spine procedures is essential to create evidence-based, national prescribing guidelines. The purpose of the current study is to prospectively record daily opioid use and pain levels after one-level lumbar microdiscectomy or decompression at two high-volume spine centres.

Following approval from the respective institutional review boards, a prospective observational study was conducted at HSS and at Midwest Orthopaedics at Rush (Chicago, USA) from September 2017–2018. During this time period, any adult patient undergoing a single-level decompression or microdiscectomy was prospectively enrolled with one of eight fellowship trained, participating surgeons, with anywhere from six to greater than 20 years of experience. These two surgeries were chosen as they are often performed on an outpatient basis, leaving providers without any inpatient narcotic consumption data on which to base an estimate for a post-discharge prescription. In order to generate conclusions for a realistic spine practice population, a selective exclusion criteria was used: non-English-speakers and those who had used opioids on a daily basis in the six months preceding surgery.

Data collection was performed using a Health Insurance Portability and Accountability Act (HIPAA) compliant automated SMS-based data collection platform, created and piloted by the first two authors specifically for use in this study. The platform sent daily queries to enrolled patients asking for 24 hour opioid consumption and average Numeric Rating Scale (NRS) pain score. At six weeks or after patients reported cessation of opioid use, the final survey questions were asked, including the number of pills remaining, method of disposal, and side effects. Prescription and refill data was taken from the electronic medical record and/or the state prescription registry. Surgeons were not asked to change their preoperative, intraoperative, or postoperative pain management protocols in any way.  Patients who completed at least 50% of the daily queries were included in the final statistical analysis.

Out of 104 patients approached, 89 patients enrolled, and 85 patients completed at least 50% of the daily queries. The average age of the patient cohort was 50.5±14.6 years, and average body mass index was 28±5.5 kg/m2. The majority of patients were male (68.2%). Fifty-seven patients (67.1%) underwent microdiscectomy and 28 (31.8%) underwent decompression. Twenty patients (23.5%) had an inpatient or overnight stay in the hospital. Thirteen percent of patients had a history of psychiatric disorder, and 31.8% of patients had a history of intermittent opioid use, defined as non-daily opioid use in the last 6 months before surgery. The most commonly prescribed opioid was oxycodone (with or without acetaminophen), and the second most commonly prescribed was tramadol (40% and 24.1%, respectively).

Daily opioid use and pain scores declined steadily through the initial postoperative period. On POD1, the average amount of opioid consumed was 16.2 oral morphine equivalents (OME), which dropped to 7.6 OME by POD7. On POD1, the median NRS pain score was 5, which dropped to 3 by POD7. The average time of opioid cessation was between POD8 and 9. To further facilitate comparisons between patients who were prescribed different types of pain medication, opioid quantities were converted into oxycodone 5 mg equivalents. A conversion ratio of 7.5 OME to one oxycodone 5 mg “pill” was used. Total opioid consumption in the postoperative period was widely variable, ranging from 0-118 pills. Median opioid consumption was 32 pills, while 75th percentile was 57 pills. Initial prescription size was equally variable. The most common prescription ranges were between 70-80 pills (18.8%), 30-40 pills (17.6%), and 10-20 pills (17.6%). While only 22.4% of patients completed their initial prescription, merely 9.4% of patients obtained a refill. Out of the 77.6% of patients who had leftover medications, only approximately 1/5th reported appropriate methods of disposal (destroy/throw away or return to pharmacy/authorities), while the others reported they were either “unsure” of what they would do or would keep the medications.

The study had several strengths. While recall bias was inherently present, the prospective design combined with automated daily data collection likely minimised such bias. Second, given the narrow exclusion criteria, the patient population was generalisable to a real-world spine surgery practice. Third, refill and prescription data was verified from the state registry for the majority of patients. However, there were also limitations. While a multivariate logistic regression was performed in an attempt to identify independent predictors of increased postoperative opioid use, the analysis was underpowered. Second, the variations in prescription quantity possibly affected opioid use, i.e., those who were given larger prescriptions may have subsequently utilised more opioids simply by virtue of having excess pills. Lastly, the variability within the surgical procedures could not be quantified. However, any prescribing guideline must make assumptions of surgical consistency in order to be clinically applicable.

There are several crucial findings that can be taken from this study to influence the content of national, state-wide, or institutional opioid prescribing guidelines. Fewer than half of the patients who finished their prescriptions obtained a refill, a pattern that has been seen before in similar research in other subspecialties.2 This suggests that the expected increase in refill requests by lowering initial prescription quantities is likely overestimated. Lowering initial prescription quantities will likely minimise the number of leftover medications without leading to an unreasonable increase in refill requests (a clinically important concern given the increased potential workload surgeons and their staff). Thirty-two oxycodone 5 mg pills (the median use) is likely a reasonable initial prescription quantity. While 57 pills, the 75th percentile of use, may lead to near-zero refill requests, such a prescribing strategy will likely lead to very large number of leftover pills. This is a critical point considering that the majority of patients did not know what to do with their leftover pills, and by default are likely to keep them in the home, continuing the risk for diversion or misuse. Thus, patient education on proper disposal and access to pill disposal centres is a crucial area for improvement as we optimise opioid prescribing strategies in the future.

While the quantities provided by this study are helpful to reduce overprescribing, blind adherence to a quantitative prescribing quantity can be potentially harmful, especially considering that opioid consumption is difficult to predict and widely variable on an individual level. Thus, along with a quantitative recommendation, any opioid prescribing guideline must include large components of patient and provider education on responsible opioid prescribing, consumption, disposal, and follow up.


  1. K Kumar, LV Gulotta, JS Dines et al. Unused opioid pills after outpatient shoulder surgeries given current perioperative prescribing habits. Am J Sports Med 2017;45:636–41.
  2. R Wojahn, L Bogunovic, R Brophy, R Wright, M Matava, J Green, C Zalomek, A Haas, W Holloway, E Garofoli, M Smith. Opioid consumption after knee arthroscopy. J Bone Jt Surg Am 2018;100A:1629–36.
  3. American Society of Addiction Medicine. Opioid Addiction 2016 Facts & Figures. Available at https://www.asam.org/docs/default-source/advocacy/opioid-addiction-disease-facts-figures.pdf. 2016.
  4. AJ Schoenfeld, W Jiang, MA Chaudhary et al. Sustained prescription opioid use among previously opioid-naïve patients insured through TRICARE (2006-2014). JAMA Surg. Epub ahead of print 2017. DOI: 10.1001/jamasurg.2017.2628.
  5. F Lovecchio, JG Stepan, A Premkumar et al. An institutional intervention to modify opioid prescribing practices after lumbar spine surgery. J Neurosurg Spine 2018;1–8.

Francis Lovecchio is currently a third-year orthopaedic surgery resident at HSS. He received his medical degree from the Northwestern University Feinberg School of Medicine (Chicago, USA). His research interests include opioid prescription practices after orthopaedic procedures, bone turnover markers in spinal fusion, and proximal junctional kyphosis in adult spinal deformity.


Please enter your comment!
Please enter your name here