Abstract
Study Design:
Retrospective review.
Objective:
Our purpose was to evaluate factors associated with increased risk of prolonged post-operative opioid pain medication usage following spine surgery, as well as identify the risk of various post-operative complications that may be associated with pre-operative opioid usage.
Methods:
The MarketScan commercial claims and encounters database includes approximately 39 million patients per year. Patients undergoing cervical and lumbar spine surgery between the years 2005-2014 were identified using CPT codes. Pre-operative comorbidities including DSM-V mental health disorders, chronic pain, chronic regional pain syndrome (CRPS), obesity, tobacco use, medications, and diabetes were queried and documented. Patients who utilized opioids from 1-3 months prior to surgery were identified. This timeframe was chosen to exclude patients who had been prescribed pre- and post-operative narcotic medications up to 1 month prior to surgery. We utilized odds ratios (OR), 95% Confidence Intervals (CI), and regression analysis to determine factors that are associated with prolonged post-operative opioid use at 3 time intervals.
Results:
553,509 patients who underwent spine surgery during the 10-year period were identified. 34.9% of patients utilized opioids 1-3 months pre-operatively. 25% patients were still utilizing opioids at 6 weeks, 17.3% at 3 months, 12.7% at 6 months, and 9.0% at 1 year after surgery. Pre-operative opioid exposure was associated with increased likelihood of post-operative use at 6-12 weeks (OR 5.45, 95% CI 5.37-5.53), 3-6 months (OR 6.48, 95% CI 6.37-6.59), 6-12 months (OR 6.97, 95% CI 6.84-7.11), and >12 months (OR 7.12, 95% CI 6.96-7.29). Mental health diagnosis, tobacco usage, diagnosis of chronic pain or CRPS, and non-narcotic neuromodulatory medications yielded increased likelihood of prolonged post-op opioid usage.
Conclusions:
Pre-operative narcotic use and several patient comorbidities diagnoses are associated with prolonged post-operative opioid usage following spine surgery. Chronic opioid use, diagnosis of chronic pain, or use of non-narcotic neuromodulatory medications have the highest risk of prolonged post-operative opioid consumption. Patients using opiates pre-operatively did have an increased 30 and 90-day readmission risk, in addition to a number of serious post-operative complications. This data provides spine surgeons a number of variables to consider when determining post-operative analgesia strategies, and provides health systems, providers, and payers with information on complications associated with pre-operative opioid utilization.
Introduction
The Centers for Disease Control and Prevention reports the United States is experiencing an opioid overdose epidemic. 1 The age-adjusted rate of drug overdose deaths attributable to opioids has increased 3% per year from 2009 to 2016; additionally, the rate of drug overdose deaths attributable to synthetic opioids has increased 88% per year from 2013 to 2016. 2 In 2016, Opioids killed 42,000 people, 40% of which used opioids prescribed to them by a physician. 1 The US Department of Health and Human services has now declared a public health emergency to focus attention and resources on the rising opioid crisis.
Notwithstanding the addictive nature of opiate medications, and subsequent risk of overdose death, multiple studies investigated the detrimental effects prolonged opioid use can have on a patient’s course of treatment. Prolonged opioid use following orthopaedic surgery has been associated with prolonged length of stay and subsequent higher numbers of adverse events. 3
Spine surgeons are placed in a challenging situation when trying to treat post-operative pain. They are the third highest prescribers of opioids among physicians in the United States.4-6 This is concerning given that nearly 20-29% of patients prescribed opioid pain medications misuse them and roughly 12% go on to develop an opioid addiction. 7
As physicians are increasingly tasked with reducing opioid prescriptions, it is imperative to identify patients who may be at risk for prolonged narcotic use following surgery and the potential increased risk of complications post-operatively in patients who are on chronic opioids. The purpose of our study was to evaluate factors that are associated with prolonged post-operative opioid pain medication usage following common spine surgeries, and to identify associations between pre-operative opioid usage and post-operative complications.
Methods
Database
MarketScan commercial claims and encounters database, consisting of nearly 40 million patients per year, was utilized. The database is constructed with data from employers, health plans and Medicaid groups. The data is de-identified by MarketScan in order to comply with requirements set forth by Health Insurance Portability and Accountability Act (HIPAA). Data is collected annually when nearly 100 percent of the claim has been paid which helps improve the reliability of capturing patients for the duration of a 1-year period. The de-identified records include laboratory results, health risk assessments, hospital discharges, prescription filling records, and electronic medical records. The Institutional Review Board committee at Penn State University Milton S. Hershey Medical Center reviewed this project’s proposal and approved an exemption for this study as it did not directly involve any human subjects. It was therefore also not necessary to obtain informed consent as this study did not involve any human subjects.
Selected Patients and Procedures
Common spine surgical procedures queried are shown in Table 1. The codes selected for this study were done so based on procedures performed on a routine basis by most spine surgeons. All patient data matching these CPT codes was included in our study, thus creating an applicable group of procedures for surgeons who routinely perform spine surgery. The database was queried for claims between 2005-2014.
Spine Procedures Queried.a
a The above CPT codes were identified as common spine procedures and utilized to identify patients in the database.
Pre-operative Opioid Usage
Specific narcotic medications were identified (acetaminophen/oxycodone, acetaminophen/hydrocodone, fentanyl, hydrocodone, oxycodone, oxymorphone, hydromorphone, tramadol) to determine which patients were utilizing narcotics before and after surgery.
Narcotic prescriptions filled between one and 3 months prior to surgery were identified to determine the number of patients taking opioids pre-operatively. We elected to stop our pre-operative opioid search at one month prior to surgery because some providers might routinely provide an opioid prescription intended for pre-operative and/or post-operative pain control at a pre-operative appointment. This would have included those patients prescribed an opioid at pre-operative appointment in our cohort, which does not accurately reflect the group who is taking routine opioids prior to surgery.
Pre-operative Variables
Pre-operative variables were identified and the cohort was queried. Variables selected included pre-operative co-morbidity diagnoses: Diagnostic and Statistical Manual of Mental Disorders (DSM-V) diagnoses, chronic pain, complex regional pain syndrome (CPRS), obesity, tobacco usage, and diabetes. The use of prescription non-narcotic medications pre-operatively was also queried. The non-narcotic neuromodulatory medications specifically queried included neuropathic medications (gabapentin and pregabalin), benzodiazepines, and anti-depressants (citalopram, escitalopram, paroxetine, fluoxetine, sertaline). These were queried to further capture those patients with chronic pain. Non-steroidal anti-inflammatory medications and acetaminophen were not queried.
Post-operative Opioid Usage and Complications
Patients utilizing opioids post-operatively were then identified by determining those patients who filled prescriptions, and/or refills, at specific post-operative time intervals (6 weeks - 3 months, 3-6 months, 6-12 months, >1 year). Though the use of opioid pain medications can be variable based on the procedure, published literature would support that opioid pain medications are generally required by the majority of patients for only the first few days after surgery, regardless of soft tissue or bone procedures.8,9 We thus set a threshold of 6 weeks as defining prolonged post-operative opioid medication use. While we believe most patients should be off opioid pain medications well before this, the 6-week threshold allows for differences in procedures and practices.
Post-operative complications (based on ICD-9 codes) were then queried. Complications included: deep vein thrombosis (DVT), pulmonary embolism (PE), irrigation and debridement of surgical site infection, myocardial infarction (MI), urinary tract infection (UTI), ileus, and wound dehiscence. We also queried for inpatient readmission rates (30 and 90 days).
Statistical Analysis
A multivariate regression analysis was utilized (SAS, Cary, NC) to determine the effect of pre-operative opioid use and its association with prolonged post-operative usage, along with specific medical comorbidities (mental health diagnosis, tobacco usage, diabetes, chronic pain, complex regional pain syndrome, use of non-narcotic neuromodulatory medications) and post-operative complications (readmission, DVT, PE, irrigation and debridement for infection, MI, UTI, ileus, and wound dehiscence). For all statistical analysis, P < .05 was considered statistically significant. We calculated odds ratios (OR) and associated 95% confidence interval. Pre-operative variables were statistically regressed across each procedure and post-operative outcome with an odds ratio with 95% confidence intervals.
Results
During the ten-year period, 553,509 patients who underwent spine surgery were identified (Table 2) by querying common spine surgery CPT codes (Table 1). Over one-third of patients (34.9%) utilized opioids 1-3 months pre-operatively. One-fourth of patients (25%) were still utilizing opioids at 6 weeks, 17.3% at 3 months, 12.7% at 6 months, 9.0% at 1 year after surgery (Table 3). For each time point, those patients still using narcotics had an over 5 times higher likelihood of having used pre-operative opiate medications compared to patients who did not.
Age and Sex Demographics of those Patients Included in the Study.
Number of Patients Utilizing Opioids Post-Operatively as well as Associated Odds Ratio of Pre-Operative Opioid Use on Post-Operative Opioid Use.a
a Pre-operative opioid use was associated with an increased post-operative use at 6-12 weeks, 3-6 months, 6-12 months, and >12 months.
Mental health diagnosis, tobacco usage, diagnosis of chronic pain or CRPS, and non-narcotic neuromodulatory medications (pregabalin, gabapentin, benzodiazepines, anti-depressants) all yielded significantly increased likelihood of post-operative opioid usage at 6-12 weeks, 3-6 months, and 6-12 months (Table 4). While all were found to be statistically significant, diagnosis of chronic pain or CRPS, and usage of non-narcotic neuromodulatory medications trended toward higher odds ratios than mental health diagnosis or tobacco usage. Diabetes and obesity were not found to be associated with an increased likelihood of post-operative opioid usage (Table 4).
Odds Ratio With 95% Confidence Intervals Associated With Specific Pre-Operative Risk Factors and their Effect On Post-Operative Opioid Use.a
a There was an increased risk of post-operative opioid use for those patients who had a mental health diagnosis, complex regional pain syndrome, chronic pain, were smokers and those who took non-opioid analgesic medications (Gabapentin, Pregabalin, Benzodiazepines, and Anti-Depressants stated above) at 6-12 weeks, 3 months-6 months and 6 months-12 months.
For patients utilizing opioids pre-operatively, there was an increased risk of readmission at 30 and 90 days post-operatively (Table 5). Patients taking opioids pre-operatively were at a statistically significant increased risk for DVT, irrigation and debridement of surgical site infection, and wound dehiscence. There was no associated increased risk for pulmonary embolism, myocardial infarction, ileus or urinary tract infection.
Odds Ratio of Specific Post-Operative Complications and their Associations With Pre-Operative Opioid Use.a
a Pre-Operative Opioid Use was Associated with Deep Venous Thrombosis, Irrigation and Debridement of Surgical Site, Wound Dehiscence, as well as 30 and 90 Day Readmission. There was Not an Increased Risk of Pulmonary Embolism, Myocardial Infarction, Post-Operative Ileus, Or Urinary Tract Infection.
Discussion
Utilizing the Marketscan database, with its nearly 39 million patients per year, we have identified multiple pre-operative variables that place patients at an increased risk for prolonged opioid use following spinal surgery. Specifically, we determined that patients who take opioid medications pre-operatively are at an increased risk for post-operative opioid use. Additionally, we found that patients taking opioid pain medications pre-operatively are at significantly increased risk for a number of serious complications as well as hospital readmission following spine surgery.
Pre-operative opioid use and specific associations has been studied within certain areas of spine surgery. One study, looking at peri-operative opioid use in spine surgery, found that patients taking opioids pre-operatively were more likely to require increased opiates in the recovery room following surgery. 10 Additionally, this study found that patients given increased pre-operative opioids were found to have a decreased opioid independence at 12 months following surgery. 10 Our data is consistent with their outcome that pre-operative opioid consumption increases opioid usage post-operatively and places patients at risk of being on opiates long-term following surgery. While their study was prospective in nature, their volume of only 583 patients was significantly lower than our data. The number of patients in our study allows for increased statistical power, which can demonstrate important trends in opioid research such as those outlined in our data.
Previous literature showed that a diagnosis of chronic pain, mood disorder, pre-operative narcotic use and tobacco exposure placed patients at an increased risk for opioid dependence long-term for lumbar spine conditions. 11 Our data is concordant with their findings, strengthening the assertion that psychiatric diagnoses, pre-operative narcotic use, and habitual behaviors such as smoking are risk factors for prolonged opioid use following spine surgery. As previously stated, our large dataset provides high power analysis to this discussion, and adds to the growing body of literature that aims to identify patient risk factors for prolonged opioid dependence. Still other studies found similar associations between opioid misuse and psychiatric illness, citing a statistically significant increase in opioid dependence among psychiatric patients compared to the general population. 12
Schoenfeld et. al performed a retrospective review of opioid-naïve patients in an attempt to determine if certain spine surgery procedures would predispose them to post-operative opioid dependence. They stratified patients into “high-intensity”, such as a posterolateral and interbody fusion procedures, and “low-intensity”, such as a discectomy and decompression, to determine if procedure “intensity” played a role in the risk of post-operative opioid use. They found that 0.1% of opioid-naïve patients remained on opioid medication up to 6 months following surgery, and also noted that patients who had “low-intensity” procedures were quicker to wean from narcotics. The only association they found for prolonged opioid consumption in the post-operative period was a pre-operative diagnosis of depression, which is in keeping with our data. They concluded that the intensity of surgery was not an independent risk factor for prolonged opioid use in opioid-naïve patients. 13 This information is concordant with our large data set showing that mental health plays an important role in how patients cope with pain in the acute post-operative period, as well as long-term after surgery. Ideally, this information will allow surgeons to work with patients to wean down on these medications before surgery so that post-operative pain control is optimized.
Utilizing the Marketscan database, we were able to include roughly 550,000 patients. We limited the pre-operative opioid usage from 1 to 3 months prior to surgery in order to limit the amount of bias, which may have been introduced through post-operative pain prescriptions being provided in the clinic within 30 days of surgery. Our data shows that patients taking opioids within 1-3 months prior to surgery were at an increased risk of remaining on narcotic medications for up to a year. Our data showed a significantly increased odds ratio of 5.45 at 6-12 weeks, 6.48 at 3-6 months, 6.97 at 6-12 months, and 7.12 at >12 months. These high odds ratios show the risk patients are at for remaining on narcotics post-operatively when they consume them pre-operatively.
Readmission rates at both 30 and 90-days were increased in patients who take opioid pain medications pre-operatively. There are certainly multiple possible explanations for this, one of which is that these same patients may run out of pain medications sooner than other patients with similar procedures, prompting them to present to a local emergency department due to uncontrolled pain. This may be the more common reason for ED visits in this population, but given the elevated risks for serious post-operative complications (i.e. DVT, surgical site infection, wound dehiscence) found in our study, this readmission rate may also be attributable to these patient’s risk of suffering one of these complications, which would also prompt a return trip to the hospital.
This information can help surgeons counsel patients who consume opiates pre-operatively. Specifically, they can counsel them that it may be difficult to decrease their opiate use post-operatively, and that they may benefit from multi-modal pain management strategies after surgery. These can come in the form of regional anesthesia, anti-inflammatory medications, as well as local anesthesia used both pre and post procedure. Additionally, to further help the orthopedic spine surgeon identify those patients who may be at an increased risk of opioid dependence, we identified a number of pre-operative patient variables associated with prolonged opioid use following surgery. These included patients with a mental health diagnosis, diagnosis of chronic pain, tobacco users and those patients taking non-narcotic neuromodulatory medications. This in combination with previous literature provides the surgical team specific patient variables to consider when determining patients who may be at risk for remaining on opioids following a procedure.
The purpose of our study was to evaluate post-operative opioid utilization and assess risk factors which may place patients at an increased risk for opioid usage following specific spine procedures. Compared to these previous works, we feel that our study provides additional findings which will ultimately help orthopedic spine surgeons discuss post-operative pain control with patients. The high-volume data we pooled allows us to determine, with increasing certainty, that pre-operative opioid use can be detrimental to patients undergoing spine surgery in that it places them at significant risk for multiple concerning post-operative complications, and drastically reduces their chances of completely weaning off all narcotics.
Our study does have several limitations. First, we considered pre-operative opioid use as a dichotomous variable, which does not take into account the heterogenous amounts of morphine equivalents between certain narcotics (ie, tramadol vs. oxycodone) and dosing schedules. The nature of the queries and coding structure of the Marketscan database did not permit granular data analysis such as specific dosages and schedules of opioid prescriptions used by patients prior to surgery. Undoubtedly this information would provide a more clear picture of how different opiates affect patients in the pre- and post-operative period. This information would likely affect a patient’s risk of prolonged post-operative use. Furthermore, our study also included a limited number of CPT codes. It would be beneficial to expand the number of CPT codes considered and to break them down on an individual basis to determine if the findings identified in this study are more specific to one procedure compared to another. Additionally, it is possible that a patient had a re-injury or separate injury in the post-operative period that required them to have an opioid pain medication filled. This factor may eliminate a number of patients included in this study. With the use of the database, we are unable to elucidate the exact reason the opioid prescription was filled. Lastly, there is a potential loss of database patient due to changes in insurance. This loss of patient due to insurance logistics would potentially decrease the database numbers of patients on prolonged opioids as well as those with post-operative complications out to the 1-year post-operative period. As such, this analysis represents a conservative estimate of risks and complications.
With these limitations in mind, this study provides spine surgeons useful information regarding trends in prolonged opioid usage and risk factors for dependence and complications following surgery. As prescribers of opioid medications, it is our responsibility to take part in reducing the opioid epidemic. This study has identified specific patient variables that place patients at an increased risk of remaining on opioids long term, which can lead to addiction. It is our duty as surgeons to provide patients with an understanding of the risks associated with surgery, and the increased risks of consuming narcotics during the pre-operative period.
Pre-operative diagnoses and narcotic use are associated with prolonged post-operative opioid usage following spine surgery. Chronic opioid use, diagnosis of chronic pain, and use of non-narcotic neuromodulatory medications have the highest risk of post-operative opioid consumption. Patients using opiates pre-operatively did have an increased 30 and 90-day readmission risk, in addition to a number of serious post-operative complications. This data provides spine surgeons a number of variables to consider when determining post-operative analgesia strategies, and provides health systems, providers, and payers with information on complications associated with pre-operative opioid utilization.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
