As the stigma and legal boundaries once associated with cannabis use continue to decrease, it is important to recognize the implications that arise for both patient safety and outcomes as well as healthcare resource allocation. We found that IBD cannabis users required higher doses of propofol and longer endoscopic procedures compared to IBD cannabis non-users. The higher propofol doses and longer duration of endoscopy in cannabis users in our cohort raise the question of whether the higher sedation requirements were simply due to longer procedures. This was evaluated by comparing the propofol doses between groups when adjusted for the duration of endoscopy. There was not a significant difference in mg of propofol per minute (P = .082), suggesting the higher propofol requirements in cannabis users were likely a product of longer time under sedation rather than metabolically mediated.

Despite the paucity of standardized research regarding cannabis, there have been several studies that comment on the administration of propofol in the setting of cannabis use. Brand and colleagues¹⁷ found the sedative effect of a low propofol concentration in mice was eliminated by analgesic doses of THC; however, increasing propofol concentrations ultimately established sedation despite THC. Imasogie and colleagues¹⁸ conducted a case–control study in Canada that found higher doses of propofol were required for sedation in patients who reported cannabis use compared to those who denied cannabis use, with the highest doses required by daily cannabis users. Similarly, Kosirog and colleagues¹⁹ performed a prospective case–control study at the Oklahoma City Veterans Affairs Medical Center that found cannabis use to be an independent predictor for higher propofol requirement during routine endoscopy. Unlike our study, there was not a statistically significant difference in procedure duration between groups in either study by Imasogie and colleagues or Kosirog and colleagues. This reintroduces the question of potential metabolic consequences of cannabis use on propofol dosing requirements. Although the exact interaction between propofol and cannabinoids is presently unknown, there are hypotheses that suggest a pharmacokinetic explanation.²⁰ For example, both propofol and THC are highly lipophilic and have long half-lives.^21–23 In patients who use THC, these factors may produce an interaction when such patients receive propofol resulting in tachyphylaxis and the requirement of higher doses to achieve sedation. Another theory notes the common receptor that both propofol and cannabis act upon: GABA_A. The possibility of a CB1 and GABA_A receptor complex may lower GABA neurotransmission and result in higher propofol requirements.²² Twardowksi and colleagues¹⁴ performed a retrospective study that found the 25 cannabis users required 220.5% more propofol to achieve continuous sedation during endoscopy than the 225 cannabis non-users. This study did not report the duration of endoscopy to allow investigation into whether higher propofol doses were due to longer duration. King and colleagues²¹ investigated propofol requirements for patients undergoing EGD in 47 cannabis users and 23 cannabis non-users. No statistically significant differences were found in propofol requirements between groups. It should be noted that none of these studies examine patients with IBD specifically. Ours is the only study to our knowledge to date that investigates propofol requirements in patients with IBD who report cannabis use.

We found that IBD cannabis users were more likely to exhibit inflammation on endoscopy. The role of cannabis in the management of inflammation and associated symptomology of IBD has been studied with mixed conclusions. Anti-inflammatory properties of cannabis have been described through cannabinoid-mediated actions on adenylyl cyclase, beta-gamma pathways, and intracellular free calcium levels.²⁴ These properties were evaluated clinically by a prospective case–control study by Naftali and colleagues²⁵ when 21 patients with CD and CDAI scores greater than 200 with disease refractory to steroids, immunomodulators, and antitumor necrosis factor agents were randomized into either the treatment THC cigarette group or placebo cigarette groups. Clinical response, defined by a decrease in CDAI score of greater than 100, was observed in 90% of the subjects in the treatment group compared to 40% of subjects in the placebo group (P = .03). Importantly, this study addresses subjective, symptomatic improvement only. Changes in C-reactive protein (CRP) were not appreciated, suggesting the improvement in CDAI may be secondary to the central effects of cannabis rather than mitigation of inflammation. Similarly, Irving and colleagues²⁶ performed a randomized, double-blind, placebo-controlled, parallel-group pilot study assessing the efficacy of an oral CBD-rich extract on remission rates in patients with UC. Although there was no difference in remission rates between the treatment and placebo groups, the treatment group had significant improvement in the patient-reported global impression of change (P = .003). Both of these studies support our finding that cannabis use likely does not improve objective mucosal inflammation in patients with IBD.

Interestingly, there was not a significant difference in the use of medical IBD therapies between cannabis users and cannabis non-users. We initially hypothesized that this variable would differ between groups, as noncompliance with medical treatment would likely predispose the patient to worse disease and potentially prompt alternative means of symptom control such as cannabis. One possible explanation for the discrepancy in endoscopic inflammation between groups despite similar rates of medical IBD treatment may paradoxically be due to the symptomatic improvement cannabis can provide. A retrospective study by Glickman and colleagues found IBD cannabis users demonstrated an increased risk for requiring corticosteroids (risk ratios [RR], 1.095; 95% CI, 1.021-1.174; P = .011), emergency department visits (RR, 2.143; 95% CI, 2.034-2.257; P < .001), and hospitalizations (RR, 1.925; 95% CI, 1.783-2.079; P < .001).²⁷ This raises concern for a potential masking effect cannabis may have on IBD-related inflammation that may result in delayed care and worse outcomes.²⁸ Consideration was given to evaluating for differences in the exhibition of inflammation between patients on various classes of IBD therapy (biologic, aminosalicylates, immunomodulators, steroids); however, given the lack of statistically significant differences appreciated in the use of IBD therapies between cannabis users and cannabis non-users, the conclusion was drawn that further stratification by type of IBD therapy would produce a similarly insignificant result.

In order to identify a possible explanation for the significantly longer duration of endoscopy in cannabis users compared to cannabis non-users, the duration of endoscopy in cannabis users who were found to have endoscopic inflammation was compared to cannabis users. Our hypothesis was that cannabis users with endoscopic inflammation would have longer endoscopic durations, potentially pointing towards the presence of inflammation as the reason for the difference in endoscopic duration between cannabis users and non-users. However, there was not a significant difference in endoscopic duration between cannabis users with and without inflammation, suggesting that the presence of inflammation alone is unlikely to explain differences in the duration of endoscopy. This analysis leads to further questions rather than an answer, further underscoring the need for randomized control studies on the outcomes associated with cannabis use. With reclassification to Schedule III status under the Controlled Substance Act, these studies might be more feasible.

Consistent with the existing literature, our findings emphasize the importance of identifying all therapeutic agents, both prescribed and not prescribed, used by patients. The initial office visit with patients with IBD should include a thorough medication reconciliation and substance abuse history to optimally identify the appropriate level of sedation for prospective endoscopies. Additionally, this will allow for a more accurate estimation of the time needed to complete an endoscopy. The higher incidence of endoscopic inflammation appreciated in IBD cannabis users may reflect a propensity to pain that may have contributed towards the significantly longer endoscopic duration in these patients. This information can help practices avoid the extra cost associated with procedures that extend past the assigned hours.

Although the average difference in the median duration of endoscopy between IBD cannabis users and cannabis non-users was only 7 and a half minutes, the cost accrued during this time is worth noting. This is especially true among these patients who more frequently have commercial insurance as opposed to Medicare. Across the 3 academic medical centers in which the endoscopies in this study were performed, the IBD procedure room is open 480 minutes per day. Considering the professional and facility fees, it is estimated that each procedure room generates 20 000 dollars daily; therefore, 312.50 dollars is generated in 7 and a half minutes. Compounded over even a handful of patients, this presents a significant financial burden if not accounted for preemptively. Additionally, these delays in patient turnover can result in changes to endoscopy throughput and the ability to attend to patients in an organized system of care.

One limitation of this study is the retrospective design, which may introduce biases in the data collection. This design restricted the ability to collect certain data, including labs to corroborate adherence to prescribed IBD treatments, urine drug tests to confirm recent cannabis use and frequency, dose, and route of administration of cannabis use. However, it should be noted that even if this were a prospective study, current quantitative cannabis testing is not sensitive or specific enough to confirm or negate recent cannabis use due to the nonlinear relationship between plasma or urine cannabinoid levels and degree of intoxication.²⁹ Additionally, the retrospective design prevented the implementation of a standardized endoscopic scoring system to more reliably characterize the degree of inflammation. The 124 endoscopies included in our study were done across 3 campuses by 82 endoscopists resulting in appreciable variety in the type and degree of detail documented regarding endoscopic inflammation. This led to our binary method of evaluating the presence of inflammation in order to prevent or at least limit inaccurate, biased, or subjective interpretations. It should also be noted that documentation of NSAID use or inclusion of these agents in the medication list of patients in our cohort was minimal. This is likely due to the perspective of some patients that over-the-counter therapies like NSAIDs are conceptualized differently than those that are prescribed by a healthcare provider, resulting in possible under-reporting of their use. This introduces the possibility that endoscopic inflammation may be secondary to or contributed to by unidentified NSAID use.

The modest sample size may affect the power to detect associations and thus present the possibility of a type II error. A substantial contributor to propofol administration during endoscopy is the comfort of the endoscopist with various levels of sedation. For example, some endoscopists may be comfortable with intermittent movement from the patient while others may only proceed if the patient is completely still. As mentioned, there were 82 endoscopists responsible for the 124 endoscopies included in our study; therefore, the variation in comfort with patient sedation may have contributed to the differences in the amount of propofol administered between groups.

The possibility exists that patients included in the cannabis non-users group were using cannabis but did not report use to their healthcare provider. Although the previously pejorative connotations associated with cannabis use are diminishing, some patients may fear the social or legal repercussions of disclosing cannabis use. This further emphasizes the need for providers to be educated on complementary and alternative therapies to accurately counsel patients and identify areas for management optimization. Due to the legal barriers to cannabis prescription, standardization of dose, route of administration, and strain were not possible, limiting the generalizability of our definition of cannabis use.

In summary, our study identified an association between cannabis use in patients with IBD with longer procedure time and a higher likelihood of exhibiting endoscopic inflammation. These correlations may be important in identifying the appropriate level of sedation and time allotment for endoscopic procedures in this population to optimize patient care and healthcare resource allocation. Future studies are needed to establish the role cannabis may have in IBD treatment as well as potential adverse effects and implications on patient outcomes.