Weeding Out the Problem: The Impact of Preoperative Cannabinoid Use on Pain in the Perioperative Period

Christopher W. Liu, BSc, MBBS, MMED; Anuj Bhatia, MBBS, MD, FRCA, FRCPC, FIPP, FFPMRCA, EDRA; Arlene Buzon-Tan, RN, MN, NP; Susan Walker, RN, MN, NP; Dharini Ilangomaran, BSc; Jamal Kara, BSc, PG, Diploma; Lakshmikumar Venkatraghavan, MBBS, MD, FRCA, FRCPC; Atul J. Prabhu, MBBS, FRCA, FRCPC


Anesth Analg. 2019;129(3):874-881. 

In This Article


This is the first study examining the impact of preoperative cannabinoid use on postoperative pain- and analgesia-related outcomes in patients undergoing major orthopedic surgery. Propensity score matching was used to generate 2 matched cohorts of cannabinoid and noncannabinoid users. Preoperative cannabinoid use was found to be associated with a small but statistically significant increase in pain scores at rest and with movement as well as higher incidence of moderate-to-severe pain in the early postoperative period. There was also a higher incidence of sleep impairment in the early postoperative period in cannabinoid users.

The actual and reported incidence of cannabinoid use is increasing in the developed world due to greater societal acceptance and more relaxed legal attitudes toward use of cannabinoids. Many federal and state or provincial governments have either allowed medical prescription or legalized recreational use (or both) of cannabinoids. In Canada, the prevalence of self-reported cannabis consumption was 12.3% in 2015, compared to 5.6% in 1985.[21] This represents a more than 2-fold increase in cannabis consumers in 2 decades. The reported incidence of cannabinoid use in our study cohort was relatively low (4.1%). This may be due to several reasons including reluctance of patients to admit to using this substance and missing data secondary to lack of data entry in our database. The noncannabinoid cohort was also older than the cannabinoid cohort in our study. This is expected because the incidence of cannabinoid use declines with age.[21,22]

This was an observational study that had the goal of understanding impact of cannabinoid use on analgesia- and pain-related outcomes in the early postoperative period. We were able to perform this study because we had robust postoperative pain-related data available in our locally developed Acute Pain Service database (Networked Online Processing of Acute Pain Information). Though a randomized controlled trial is less likely to be affected by confounding or bias, it is challenging to perform a prospective study to assess impact of recreational use of cannabinoids on pain because of logistical and ethical challenges. Logistical difficulties include a lack of methodology to quantify cannabinoid use (unlike other recreational substances such as cigarettes and alcohol or medications such as opioids). Ethical dilemmas also arise because cannabinoids are associated with several adverse effects including psychosis and schizophrenia.[23]

We performed propensity score matching on our study cohort to reduce the probability of bias due to confounding arising from the nonrandom assignment of exposure.[24] The use of propensity score methods (ie, using the probability of being exposed given covariates) can reduce bias due to measured confounders.[25] Use of propensity score matching in our study resulted in the cannabinoid and noncannabinoid cohorts being well matched on a number of measured baseline covariates (age, type of surgery, sex, preoperative opioid use, depression, anxiety, and administration of a regional analgesia) that could have affected pain- and analgesia-related outcomes. Though we had a relatively large number of noncannabinoid users in our study, we chose to match the cohorts in a ratio of 1:1 because matching in higher ratios can increase bias.[24] We believe our study was adequately powered to study the primary outcome (difference in intensity of pain with movement between the 2 cohorts compared using the paired t test) because a post hoc power calculation indicated 86 patients were required in each group to detect a mean difference of 1 (on a 0–10 pain numerical rating score) with an SD of 2 for a study with a type I error of 5% and a type II error of 10%. A 10 mm change in Verbal Analogue Scale scores for pain (or a 1-point difference in numerical rating scores) has previously been found to signify a clinically important difference for acute postoperative pain.[26]

Cannabinoid users in our study had a higher intensity of pain at rest and with movement in the early postoperative period as compared to noncannabinoid users. However, this difference was small, and the requirement for opioids in both cohorts was similar. The lack of difference in opioid requirements between the 2 propensity-matched cohorts may be due to an inadequate sample size for detecting this difference. Our findings also indicate that acute postoperative pain in patients undergoing major orthopedic surgery is still poorly treated despite our current best efforts. However, our results suggest that patients who use cannabinoids may be at risk of a higher intensity of pain after surgery. We also found that cannabinoid users who had spine surgery had higher intensity of pain with movement, whereas those who had hip or knee surgery had more pain at rest (Table 3). Though multilevel spine surgery is known to be associated with significant pain in the early postoperative period and mobilization after spine surgery can take longer than following major lower limb joint surgery,[27] the reasons for our findings are unclear. It should also be noted that the actual incidence of physical impairment was high in both cohorts in our study.

This is the first study that evaluates the impact of preoperative cannabinoid use on the postoperative pain experience. The findings of this study are significant because prevalence of cannabinoid use in the community for both recreational and medical purposes is increasing. This will lead to an increasing number of patients presenting for surgery, especially in the younger age groups, who will be on cannabinoids preoperatively. As such, it is important to know how preoperative cannabinoid use can impact postoperative outcomes and, therefore, allow us to study how we can mitigate adverse pain experiences in the future. There are a few possible reasons for increased pain and its alleviation by cannabinoids in the postoperative period in patients who use these substances on a regular basis. Both cannabinoid 1 and 2 receptors have been shown to be involved in pain modulation, and peripheral agonist activity at these receptors has shown inhibitory effects on pain responses.[28,29] Laboratory studies have also shown that long-term exposure of cannabinoid 1 receptors to cannabinoids may lead to receptor downregulation, internalization, as well as desensitization.[30,31] Cannabinoids have also been used to aid weaning of opioids in the postoperative period in a patients with persistent severe pain and high opioid requirement following liver transplantation.[32] Finally, studies on neuroanatomical distribution of cannabinoid and opioid receptors show colocalization in the sites in the central nervous system involved in processing painful stimuli, notably in the spinal cord, locus coeruleus, and striatum.[33]

There was also a higher incidence of sleep impairment in the cannabinoid cohort in our study during the early postoperative period. The poorer quality of sleep in the cannabinoid cohort may reflect higher intensity of pain and or acute withdrawal from cannabinoids.[34] This suggests that patients on cannabinoids in the preoperative period may need treatment for insomnia in the postoperative period. Administration of synthetic cannabinoids in the perioperative patients who have a history of cannabinoid use may be another therapeutic option to consider to reduce the incidence of insomnia. We did not find any difference in the incidence of pruritus in the 2 cohorts. It is of interest to note that cannabinoids have been used for the treatment of pruritus in patients with dermatological disorders.[35]

There are some limitations of this study. Data on variables such as a history of anxiety or depression, physical, and sleep impairment were based on patient self-reporting. Compared to the use of validated screening questionnaires, this form of reporting can lead to either over- or under-reporting.[36] Though we used propensity score matching to reduce the probability of bias and confounding, an inherent drawback in observational studies is of unknown confounders affecting the outcome variables. During the matching process, we were unable to stratify patients based on the preoperative pain scores or the absolute opioid dosage because we did not have access to these data. Establishing incidence of use of cannabinoids and other substances in patients with chronic pain would also be informative. A study on patients with low back pain reported higher odds of using marijuana and other recreational substances, and patients with history of substance abuse were also found to be more likely to be on prescription opioids.[36]

We were able to assess pain and related outcomes only in the first 24 early postoperative period, but a study that follows patients over a period of weeks to a few months may yield a more accurate profile of postoperative pain trajectories in this population. Our study also did not address the issue of using cannabinoids to alleviate pain in the postoperative period for patients who have a history of using these substances. However, a case report describes refractory pain after an emergency repair of an abdominal wound dehiscence in a patient. Addition of nabilone to the analgesic regimen resulted in alleviation of pain.[37] Finally, the reported rate of cannabinoid use in our cohort was relatively low at 4.1%. This may reflect reluctance to report use of cannabinoids.

To conclude, this study demonstrates an association between cannabinoid use and increased postoperative pain and worsening of sleep in patients undergoing major orthopedic surgery. We recommend prospective randomized controlled trials to determine the mechanism behind this phenomenon as well as possible strategies that can mitigate this. A possible study design would be to assess the effect of continuation of preoperative cannabinoids during the perioperative period in patients undergoing surgery while measuring pain and related outcomes using validated tools or questionnaires.