The Quebec Cannabis Registry: Investigating the Safety and Effectiveness of Medical Cannabis

Abstract

Objective:

To investigate the safety and effectiveness of medical cannabis (MC) in the real-world clinical practice setting.

Design:

A 4-year prospective noncomparative registry of adult patients who initiated MC for a variety of indications. This paper reports on patients followed for up to 12 months, with interim visits at 3, 6, and 9 months after enrollment.

Setting:

Public or private outpatient clinics certified to authorize MC in the province of Quebec, Canada.

Participants:

Overall, 2991 adult (age ≥18 years) patients (mean age 51 years; 50.2% women) were enrolled between May 2015 and October 2018, with the last follow-up ending in May 2019.

Interventions/Exposures:

Cannabis products (dried, oil, or other) purchased from a Canadian licensed cannabis producer as authorized by physicians.

Main Outcome Measures:

The primary outcomes were self-reported pain severity, interference and relief (Brief Pain Inventory [BPI]), symptoms using the Revised Edmonton Symptom Assessment System (ESAS-r) and health-related quality of life dimensions (EQ-5D-5L) at baseline and each follow-up visit. The secondary outcomes were self-reported adverse events (AEs) and characteristics of cannabis treatment.

Results:

All patient-reported outcomes (BPI, ESAS-r, and EQ-5D-5L) showed a statistically significant improvement at 3 months (all p<0.01), which was maintained or further improved (for pain interference, tiredness, and well-being) over the remainder of the 12-month follow-up. Results also revealed clinically significant improvements in pain interference and tiredness, anxiety, and well-being from baseline. There were 79 AE reports (77 patients), 16 met the regulatory definition of seriousness, in which only 8 AEs were certainly or probably related to MC.

Conclusions:

MC directed by physicians appears to be safe and effective within 3 months of initiation for a variety of medical indications.

Introduction

Since 1999, the Canadian federal government has allowed access to cannabis for medical purposes, initially through a strict ministerial exemptions scheme, and then through more formal regulations such as the Medical Marihuana Access Regulations. This was replaced by the Marihuana for Medical Purposes Regulations in 2014 and subsequently by the Access to Cannabis for Medical Purposes Regulations in 2016. The latter allowed the production, sales, and distribution of dried cannabis flowers, oils, and capsules.^1,2 Finally, with the Cannabis Act, the recreational use of cannabis was legalized in October 2018.

At this time, access to medical cannabis (MC) has been legalized in over 50 countries.³ Since this access often precedes clinical guidelines and randomized clinical trials of MC are few and far between, health care practitioners feel unconfident in their ability to prescribe MC safely and appropriately.^4–7 Observational studies have been conducted to help inform policy and clinical practice.⁸ Many of these are surveys,^9–13 collecting opinions on MC from physicians or patients; others are retrospective analyses of government records,^14,15 clinic data,^16–19 or existing registries.^20–22 While the information they present is valuable, these studies are limited in scope and generalizability.

In 2014, the Quebec Medical College (Collège des médecins du Québec, CMQ) proposed a precautionary framework to allow access to MC while also ensuring the protection of patient health and well-being. These guidelines stipulated that MC could only be used within the context of research.²³ Consequently, between April 1, 2014, and October 17, 2018, physicians who agreed to direct the use of MC in Quebec were asked to do so only within an ethically approved research framework.¹² The Quebec Cannabis Registry (QCR) was established to follow the CMQ framework and was the first large MC registry launched in the world.

The QCR aimed to (1) systematically collect real-world data on indications, modes, and forms of administration, benefits, and risks of MC products; (2) allow for the conduct of observational studies on MC, and (3) support the ongoing development of a drug monitoring program. Countries around the world are now launching similar national MC registries—including Denmark, France, Germany, Ireland, Italy, and the United Kingdom—but many are still recruiting, and published data are lacking.

The main objective of the present study was to investigate the safety and effectiveness of MC in the real-world clinical practice setting and within the context of an active surveillance program.

Methods

The STROBE statement for reporting observational studies was followed.²⁴

“Authorized,” “recommended,” “prescribed,” and “directed” MC are used interchangeably in this article. When authorizing MC in any form, it is mandatory for Canadian physicians to indicate a limit on the authorized amount, according to the equivalent weight of dried cannabis. This is known as the “authorized daily dose,” written in grams per day. Physicians may also limit the tetrahydrocannabinol (THC) content in the MC that they authorize.

Design

A prospective noncomparative registry of adult patients (age ≥18 years) who initiated MC, as directed by a physician, was created. Eligible patients were enrolled through participating physicians at initiation between May 2015 and October 2018 and were followed for up to 12 months. Patient follow-up ended at the first of the following events: discontinuation of MC due to lack of treatment efficacy or occurrence adverse event (AE), loss to follow-up, 1-year follow-up, or end of data collection (May 2019). At the end of follow-up, physicians completed an exit survey (Supplementary Appendix SA1.3). Neither patients nor physicians received financial incentives for participation, and patients were not reimbursed for MC. The full data collection forms are provided in Supplementary Appendix SA1. Data were collected prospectively at baseline and at 3, 6, 9, and 12 months.

Ethics

The project was initially approved by the central Research Ethics Committee of the Ministry of Health and Social Services (CCER) of Quebec. The CCER assumed the role of evaluating and monitoring the project for both private clinics and participating institutions, which were part of the health and social services network. On May 3, 2019, the Ethics Committee of the McGill University Health Centre took over the role of the CCER for the QCR. Written informed consent was obtained from each study participant, before collecting any baseline information.

Study procedures and measures

Baseline visit

Baseline visits took an average of 90 min to complete, including obtaining written informed consent. Pen-on-paper tools were used for data collection. MC directives were recorded by physicians in terms of (1) authorized daily dose (g/day); (2) primary mode of administration (oral, inhalation, or a combination of the two methods); (3) cannabinoid content ratio (THC-dominant, cannabidiol CBD-dominant, and/or balanced); and (4) number of products prescribed. The physician-reported indications for MC authorization were also recorded.

Patient sociodemographics (age, sex, and occupation), medical history, concomitant medications, smoking status, alcohol use, and recreational drug use were collected at baseline. Similar to the recommendations made by Health Canada²⁵ and the College of Family Physicians in Canada,²⁶ the presence of the following contraindications was assessed, as well as circumstances calling for cautious use of THC (known as “precautions”): unstable cardiopathy, personal or family history of psychosis, pregnancy/breastfeeding status, history of cannabis or substance use disorder, and concurrent psychiatric disorder (schizophrenia, attention deficit, obsessive-compulsive, and bipolar disorders.)

Patient-reported outcomes (PROs) were assessed using validated tools. The Brief Pain Inventory (BPI)²⁷ recorded patients' levels of pain severity, pain interference, and pain relief with current treatment. Scores range from 0 to 10. Pain relief is assessed on a scale from 0% to 100%. Several studies have supported the reliability and validity of the BPI to assess pain intensity and pain interference among diverse clinical populations.⁴

The revised Edmonton Symptom Assessment System (ESAS-r)²⁸ assessed the intensity of nine symptoms: pain, fatigue, somnolence, nausea, lack of appetite, shortness of breath, depression, anxiety, and well-being. Symptoms are rated from 0 to 10. This tool has been validated in cancer and noncancer patient populations.⁵

The EQ-5D-5L^29,30 instrument was also used, a health-related quality of life questionnaire that includes five dimensions (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression) and a global assessment (0–100% rating scale; 0—Worst health imaginable, 100—Best health imaginable).³⁰

Follow-up visits

The following were collected at each follow-up: AEs, PROs using BPI, ESAS-r, and EQ-5D-5L, physician-reported information regarding their MC directive (dose, mode of administration, cannabinoid content ratio, and number of products), MC discontinuation and reason for discontinuation, and change in medications. All moderate to severe AEs were also reported to Health Canada.

Follow-up visits typically took 30 min to complete. If a patient did not attend a follow-up visit, the time window allowed was 1.5 months (for example, for a follow-up at month 3, the patient could have the visit at month 1.5 to 4.5).

Safety data

Since this was an observational study, moderate to severe AEs were not reported to the study's REB but to Health Canada. The Health Canada Side Effect Reporting Form (Supplementary Appendix SA1.4) is used to report AEs of drugs and marketed health products like prescription and over-the-counter medications, natural health products, and biological products. It is used as a way to collect real-world data on a drug's safety and potentially inform health policy and regulation; it forms part of a postmarketing surveillance program by Health Canada that is called Canada Vigilance.

Several retrospective studies have analyzed the data in the Canada Vigilance Online Database.^31–33 To our knowledge, however, this is the first study to use it prospectively in this way.

Data analysis

Descriptive analyses were conducted on patient baseline characteristics, medical history, primary diagnosis, and symptoms. MC directives at baseline were described according to dose, mode of administration, THC/CBD content ratio, and number of products prescribed. Analyses were stratified by year of enrolment. Percentages were used for categorical variables and means, with standard errors of the mean (SEM), for continuous variables.

To assess the presence of selection bias due to cohort attrition, the baseline characteristics of patients who completed 12 months of follow-up (i.e., by attending all four follow-up visits) were compared to the baseline characteristics of all patients enrolled in the registry at baseline. In this way, we could assess the extent to which the characteristics of the patients who completed follow-up matched—and were therefore representative of—those of the whole baseline sample.

A series of one-way repeated-measures analyses of variance were conducted to compare the mean scores on the BPI, ESAS-r, and EQ-5D-5L between baseline and each follow-up visit. All analyses were conducted as paired analyses.

Patients were not excluded on the basis of missing visits or data, and there was no imputation of missing values. However, if a patient had two visits fall in the same time window, only the visit closest to the date it should have occurred was selected for analysis. For all analyses, the statistical significance was set to p<0.05 and normality distribution of all variables was assessed. All statistical tests were conducted using SAS version 9.4 (SAS Institute, Cary, NC).

Results

Patient characteristics

Overall, 2991 subjects were enrolled. The mean age of patients was 51 years and 50.2% were female (Table 1). Of the total, 22.4% presented with contraindications or precautions regarding the use of THC. Patients who presented with contraindications (e.g., cardiovascular or psychiatric disease) were not authorized products with >2% THC (in dried cannabis products) or 2 mg/mL THC (in oral formulations). However, if their primary indication for using MC necessitated a THC-rich or THC-CBD balanced product (e.g., appetite stimulation or insomnia), the prescribing physician consulted with the relevant specialist (e.g., cardiologist, psychiatrist) to attempt a judicious use of THC while maintaining the patient on the lowest effective dose.

Table 1.

Quebec Cannabis Registry Patient Baseline Characteristics

Characteristics	N (%)		Missing values
No. of participants	2991
Mean age (years)	50.9		0
Age range (years)	18 to 96		0
Female	50.2		9
Occupation
Full-time	740 (25.1)		38
Retired	685 (22.9)
Long-term disability	652 (21.8)
Short-term disability	293 (9.8)
Unemployed	234 (7.8)
Part-time	199 (6.7)
Student	79 (2.6)
Other	71 (2.4)
Precautions and contraindications to THC	670 (22.4)
Attention deficit disorder, obsessive-compulsive disorder, or bipolar disorder		276 (9.2)
Abuse (physical, psychological, or sexual)		271 (9.1)
Personal or family history of psychosis		71 (2.4)
Unstable cardiopathy		32 (1.1)
Schizophrenia		18 (0.6)
Pregnant/breastfeeding		2 (0.1)
Alcohol use
Occasional	1322 (45.7)		99
Never	1132 (39.1)
Regular	434 (15.0)
Alcohol consumption (unit/week) (SD)		9.01 (8.28)
Tobacco use
Never	1492 (51.0)		69
Current	732 (25.1)
Cigarette consumption (cigarettes/day) (SD)		15 (9.56)
Former	694 (23.7)
Illicita drug use
Cannabis		136 (4.5)
Psychostimulants		32 (1.1)
Hallucinogens		13 (0.4)
Opioids		8 (3.9)
Synthetic cannabinoids		2 (0.1)
Ketamine		1 (0.0)
Other		19 (0.6)
Unknown		55 (1.8)

Values are numbers (percentages) unless stated otherwise.

^a Data collection occurred before cannabis legalization in Canada.

THC, tetrahydrocannabinol.

The most common primary diagnoses were pain disorders (61.3%), mental health diagnoses (13.8%), cancer (12.1%), and neurological disorders (9.0%) (Supplementary Appendix SA2). The top three primary indications for MC prescription were pain (76.9%), anxiety (9.6%), and insomnia (4.8%). The majority of patients (63.5%) presented with more than one indication at baseline. Thus, when considering both primary and secondary indications for MC use, pain was again observed most commonly (53.2%), followed by insomnia (21.9%) and anxiety (19.0%).

The median duration of registry participation was 25 weeks: 1813 patients completed 3-month visits, 1281 completed 6-month visits, 886 completed 9-month visits, and 687 completed 12-month visits. The most common reason for study exit was the decision of the Registry's scientific committee to end the follow-up of patients 6 months after legalization of recreational cannabis in 2018 (60.3%) (Supplementary Appendix SA3). This was done to preserve data integrity, as patients were no longer limited to MC and this could bias results. The next most common reason was loss to follow-up (18.1%). Notably, lack of treatment effectiveness (5.4%), inability to afford MC (2.2%), and AEs (1.1%) were among the most uncommon reasons.

When comparing baseline characteristics of patients who completed all four follow-up visits (N=687) with the baseline characteristics of all patients enrolled (N=2991), significant differences (i.e., equal or greater than 10%) were found only in the mode of MC administration: 35.5% of patients with a 12-month follow-up received oral MC at baseline compared to 52.0% in the overall population, and 38.1% of patients with a 12-month follow-up used inhaled MC compared to 25.3% overall (Supplementary Appendix SA4).

Patterns of MC directives by physicians

The majority of patients were authorized ≤1 g/day at baseline (64%). There was an increase in the proportion of patients authorized between 1.1 and 3.0 g/day, from 42% at baseline to 64% at 12 months. Less than 7% at baseline and <18% at 12 months were authorized ≥3.1 g/day (Supplementary Appendix SA5).

Over the course of study enrollment years, there was a significant increase in oral administration, use of CBD-dominant products and number of MC products recommended at treatment initiation (Fig. 1A, C, E, respectively).

FIG. 1.

Mode of administration, cannabinoid content ratios, and number of medical cannabis products recommended at baseline and per follow-up. (A) Mode of administration at baseline per year. (B) Mode of administration per follow-up. (C) Cannabinoid content ratio at baseline per year. (D) Cannabinoid content ratio per follow-up. (E) No. of products at baseline per year. (F) No. of products per follow-up.

Over the 12 months, patients were gradually recommended more combinations of oral and inhalational routes, more THC-dominant products, and were recommended two or more MC products concurrently (Fig. 1B, D, F, respectively). These changes were suggested by the treating physicians based on clinical outcomes, for example, MC effect on pain relief and/or occurrence of AEs.

Effectiveness of MC

Half of patients (50.3%) met the criteria for pain based on the tools administered at baseline. Mean pain severity scores according to the BPI were 5.3 (SEM: 0.043) out of 10 at baseline, 4.5 at 3 months (SEM: 0.052), 4.3 at 6 months (SEM: 0.057), 4.5 at 9 months (SEM: 0.065), and 4.4 at 12 months (SEM: 0.071) (Fig. 2A). The p-value between the baseline visit and all follow-up visits was <0.0001. There was also further improvement in this score at 6 and 12 months, compared with 3 months (p<0.01 and p<0.05, respectively). The estimated effect size of improvements in this variable was 0.048.

FIG. 2.

Mean scores of patient-reported outcomes, over time. (A) Mean score over time of patient-reported pain severity as measured by four BPI subscales (worst pain in 24 h, least pain in 24 h, average pain, and current pain). *p < 0.05, **p < 0.01, ***p < 0.0001. (B) Mean score over time of patient-reported pain interference as measured by seven BPI subscales (pain interference with general activity, mood, walking ability, normal work, relationships with other people, sleep, and enjoyment of life). **p < 0.01, ***p < 0.0001. (C) Mean percentage of pain relief reported by patients on the BPI questionnaire. ***p < 0.0001. (D) Mean patient-reported health state score as measured by the EQ-5D-5L questionnaire. **p < 0.005, ***p < 0.001. BPI, Brief Pain Inventory.

Mean pain interference scores were 5.4 (SEM: 0.048) out of 10 at baseline, 4.3 (SEM: 0.062) at 3 months, 4.0 at 6 months (SEM: 0.067), 4.1 at 9 months (SEM: 0.074) and 4.0 at 12 months (SEM: 0.080) (Fig. 2B). The p-value between the baseline visit and all follow-up visits was <0.0001. There was further improvement in this score at 6, 9, and 12 months, compared to 3 months (p<0.05). The estimated effect size of improvements in this variable was 0.078.

In the 24 hours before their baseline visit, patients reported 39% (SEM: 0.578) pain relief from pain treatments/medications. Pain relief scores became 49% at 3 months (SEM: 0.776), 51% at 6 months (SEM: 0.870), 52% at 9 months (SEM: 1.036), and 53% at 12 months (SEM: 1.091) (Fig. 2C). The p-value between the baseline visit and all follow-up visits was <0.0001, the estimated effect size being 0.027.

Normality assumption conditions were met for all BPI outcome results. The improvements in mean pain severity, interference, and relief scores were statistically significant at 3 months, with clinically significant (i.e., ≥30% change in mean scores^9,10) reductions in pain interference at the same visit.

All symptoms recorded via the ESAS-r significantly improved at 3, 6, 9, and 12 months, compared to baseline (all p<0.0001), with tiredness and well-being further improving at 9 and 12 months, compared to the 3-month visit (p<0.05). Well-being continued to improve at 12 months, compared to 6 months (p<0.05) (Table 2).

Table 2.

Patient-Reported Outcomes

	Baseline		3 Monthsa		6 Monthsa		9 Monthsa		12 Monthsa		Effect size
	N	Mean±SEM	N	Mean±SEM	N	Mean±SEM	N	Mean±SEM	N	Mean±SEM
Mean score of patient-reported symptoms measured by ESAS-r questionnaire ESAS-r domains: differences between visits
Pain	2830	5.03±0.05	1545	4.26±0.06	1120	4.13±0.07	799	4.22±0.08	616	4.17±0.09	0.036
Tiredness	2837	5.85±0.05	1546	4.84±0.07	1121	4.69±0.08	800	4.56±0.09b	619	4.54±0.10b	0.050
Nausea	2822	1.49±0.05	1542	1.13±0.05	1117	1.14±0.06	801	1.16±0.07	622	1.10±0.07	0.007
Depression	2819	3.36±0.06	1539	2.71±0.07	1117	2.60±0.07	796	2.64±0.09	622	2.61±0.10	0.020
Anxiety	2819	4.23±0.06	1534	3.32±0.07	1115	3.27±0.08	797	3.20±0.09	621	3.10±0.09b	0.034
Drowsiness	2821	3.58±0.06	1540	2.86±0.07	1117	2.78±0.08	801	2.69±0.09	616	2.77±0.10	0.020
Appetite	2821	2.95±0.06	1543	2.24±0.07	1118	2.25±0.08	797	2.27±0.09	620	2.29±0.10	0.016
Wellbeing	2766	5.18±0.05	1507	4.46±0.07	1097	4.30±0.07	792	4.19±0.09b	617	4.08±0.10b,c	0.027
Shortness of breath	2816	2.16±0.05	1541		1.84±0.06	1120	1.77±0.06	796	1.87±0.07	621	1.77±0.08	0.006

	Baseline		3 Monthsa		6 Monthsa		9 Monthsa		12 Monthsa		Effect size
	N	Mean±SEM	N	Mean±SEM	N	Mean±SEM	N	Mean±SEM	N	Mean±SEM
Mean score of health dimensions measured by the EQ-5D-5L questionnaire EQ-5D-5L domains: difference in mean scores between visits
Mobility		2146	2.41±0.02	1279	2.22±0.03	1024	2.15±0.03b	770	2.19±0.03	614	2.18±0.03	0.014
Self-care		2142	1.73±0.02	1274	1.55±0.02	1023	1.54±0.02	769	1.56±0.03	611	1.56±0.03	0.012
Usual activities		2141	3.00±0.02	1273	2.61±0.03	1021	2.54±0.03b	766	2.53±0.03b	609	2.44±0.04b,c,d	0.043
Pain/discomfort		2147	3.31±0.02	1276	2.94±0.03	1025	2.84±0.03b	770	2.82±0.03b	612	2.77±0.03b,c	0.054
Anxiety/depression		2137	2.61±0.02	1273	2.29±0.03	1017	2.23±0.03b	767	2.21±0.03b,c	609	2.13±0.03b,c,d	0.034

^a All comparisons between baseline and months 3, 6, 9, and 12 were significantly different, p<0.0001.

^b Significantly different from 3 months, p<0.05.

^c Significantly different from 6 months, p<0.05.

^d Significantly different from 9 months, p<0.05.

ESAS-r, Revised Edmonton Symptom Assessment System; SEM, standard error of the mean.

Clinically significant improvements (i.e., ≥1-point reduction in mean scores¹¹) were recorded for fatigue at 3 months compared to baseline, as well as for anxiety and well-being at 9 months compared to baseline. These clinically significant improvements persisted at subsequent follow-ups. Normality assumption conditions were met for all variables.

The average global health state, as measured by the EQ-5D-5L, was 54.5% (SEM: 0.4%) at baseline (Fig. 2D). This significantly improved to 62.4% at 3 months, 63.7% at 6 months, 65.1% at 9 months, and 65.7% at 12 months (all p<0.0001 compared to baseline); further significant improvements were demonstrated at the 9- and 12-month visits when compared to the 3-month visit (p<0.005). The estimated effect size of improvements in this variable is 0.040. Normality assumption conditions were validated. Individual health dimensions all showed significant improvements at all follow-up visits, compared to baseline (Table 2).

Safety of MC

Overall, 158 AEs were observed among 77 patients (2.6%) (Supplementary Appendix SA6). Of the 79 AE reports, 63 (79.7%) were non-serious, and 16 (20.3%) met the regulatory criteria for seriousness, including 1 death. The death was secondary to myocardial infarction and deemed unlikely caused by MC.

Among nonserious reports, the three most common AEs were nausea (10/63, 15.9%) dizziness (9/63, 14.3%), and somnolence (8/63, 12.7%). Among serious reports, myocardial infarction was the most common AE (3/16, 18.8%). Of the serious reports, 8/16 (50.0%) led to hospitalization, 4/16 (25.0%) required intervention, and 3/16 (18.8%) caused disability.

Causality of the 16 serious reports (23 AEs) was assessed as Unrelated for one AE (urinary tract infection), Unlikely for ten AEs (myocardial infarction [n=3], abnormal respiration, alopecia, parasitic gastroenteritis, blood test abnormal, abdominal distension, dyspnea, pain), Possible for two AEs (pneumonia, intraocular pressure increased), Probable for seven AEs (tachycardia, dry mouth, loss of confidence, hot flushes, asthma, rash, pruritus), Certain for one AE (suicidal ideation) and, Unable to assess for two AEs (vomiting, angina pectoris).

Discussion

This is the first large-scale study launched to assess the safety and effectiveness of MC among adults in a province-wide longitudinal prospective registry. The QCR allowed for the collection of uniform and reliable data regarding MC directives, PROs, and safety in a large number of patients with a wide variety of medical conditions.

Data on MC directives showed that physicians initially opted for 1 g/day of dry equivalent for MC, and mostly remained within 3 g/day during follow-up, as recommended by Health Canada.²⁵ This is consistent with results of previously published research.^34–38

This study showed a fivefold increase in the rates of initiating patients on oral MC over the course of enrollment years. The increase over time in variety and availability of oral products influenced the likelihood of initiating patients on the oral route. Consequently, physicians may have become more aware of the long-acting symptom control offered by oral MC. Inhaled MC, on the contrary, has a faster onset and shorter duration of action, making it useful for breakthrough symptoms. This may explain why, over the course of follow-up, there were higher rates of concurrent routes of cannabinoid administration.

This study also found that Quebec physicians preferred initiating MC with balanced cannabinoid content ratios. This is possibly due to the wider spectrum of therapeutic effects associated with combining CBD and THC, and thus better chances of positive clinical response. As physicians gained a better understanding of how to titrate and fine-tune the authorization of THC-dominant products, it follows that there was also an increase in the authorization of THC-dominant products over time. Furthermore, although patient symptoms may respond to CBD alone, the use of high doses (≥50 mg/day) can be impractical due to cost. CBD may work effectively at lower doses when combined with THC, even at subpsychoactive levels (≤2.5 mg THC/day).³⁹ The increase in the use of balanced MC is consistent with a potentially synergistic effect. Furthermore, this approach minimizes the risk of AEs associated with high-dose THC.

The increase in the number of MC products prescribed at baseline over study enrollment years could be related to physicians becoming more aware of different MC products. Similarly, over the course of follow-up, physicians may have recommended an increased number of MC products as they became more aware of patient needs and familiar with the responsiveness of their symptoms to MC.

This study shows that safety and effectiveness of MC may be continuously monitored by prescribing physicians. Doses, strains, and modes of administration can be adjusted according to safety and response. Once safety and tolerability are established, the physician can then focus on effectiveness, titrating the dose of THC and/or CBD to achieve a response at the minimal effective dose. This is consistent with the “start low, go slow, and stay low” method well-described in the literature^40,41 and national guidelines.²⁶

All PROs measured via BPI, ESAS-r, and EQ-5D-5L showed a statistically significant improvement at 3 months, which was maintained over the course of follow-up. The largest estimated effect size was on the improvement in pain interference. For pain interference, fatigue, anxiety, and well-being, a clinically significant reduction in scores was also observed. Altogether, this suggests that the clinical potential of MC is not limited to pain but may extend to domains such as function and quality of life, supporting the role of cannabinoids in complementary patient care. Our data suggest that, in our population, the benefits of MC may be achieved relatively quickly and maintained over time.

Given the different routes of administration and personalized titration schedules, MC therapy can be highly heterogenous. Our data confirm that, under the guidance of health care professionals, MC may achieve clinical success through different regimens. This is particularly important when considering that, in the context of many jurisdictions in the United States, cannabis products are suboptimally regulated and physicians may not be required to have specific training or qualifications in prescribing MC.⁴²

This study's results are consistent with the available literature regarding the main indications for which MC is prescribed.^11,43–45 QCR's data on MC effectiveness are comparable to those of other real-world studies that examined pain relief and/or quality of life.^18,46–50 The safety data from the current study are also consistent with the literature.¹⁹ MC was generally well-tolerated, possibly because of physician supervision, per QCR protocol. Slow and controlled titration allowed patients to find the lowest dose effective at controlling their symptoms. The low frequency and severity of AEs were achieved through regular follow-ups, as recommended by the CMQ.

A number of limitations must be considered: (1) this is a noncomparative study, therefore true association between the occurrence of AEs and PROs with MC usage is difficult to confirm, (2) half of the initial sample exited by 6 months (retention bias), (3) MC use was reported in terms of what was directed by physicians, rather than what was actually used by patients, (4) the MC effect on any particular PRO may be diluted by considering patients who had other indications for starting MC, and (5) some patients may have had access to illegal cannabis markets and illicit drug use was only assessed at baseline, despite its potential to confound safety and effectiveness of MC over the course of the study.

Despite its limitations, our study was prospective, included a large number of patients, included multiple assessment time points, and its findings can be generalized due to the large number of physicians and clinic settings involved. Medical diagnoses and the wide range of indications for MC were reported through standardized forms. Finally, the majority of published studies have <6-months' follow-up data^18,45–48 and higher attrition rates,^25–49 compared to our study.

In conclusion, physician-directed MC appears to be safe and effective within 3 months of initiation for a variety of indications. This large prospective and longitudinal surveillance study should be followed by disease- or indication-specific registries and trials, which should confirm MC effectiveness with a comparator. Furthermore, real-world data on the patient-use of MC, in addition to the directives by physicians as recorded in this study, may help with developing specific protocols for starting and monitoring MC.

Abbreviations Used

AEs

adverse events

BPI

Brief Pain Inventory

CBD

cannabidiol

CCER

Le Comite' central d'ethique de la recherche

CMQ

Collège des médecins du Québec

EQ-5D-5L

EuroQol-5 dimensions-5 levels

ESAS-r

Revised Edmonton Symptom Assessment System

MC

medical cannabis

PROs

patient-reported outcomes

QCR

Quebec Cannabis Registry

SEM

standard errors of the mean

THC

tetrahydrocannabinol

Authors' Contributions

M.W. conceived the study. M.W., A.V., Y.M., P.B., A.N., W.B., and J.D. contributed to the study design. M.W., A.V, W.B., and J.P. contributed to the data collection phase. A.V., Y.M., R.G., M.K., and M.C.M contributed to data management, data analysis, drafting the methods section, and interpreting the results of this article. A.V., Y.M., and M.C.M wrote the article and designed the figures and tables. M.C.M. managed the administration of the study (2018–2019). A.V., P.B., A.N, Y.M, M.M., J.P., and W.B. provided scientific oversight and leadership of the prospective study. All authors contributed to refinement of the study protocol and approved the final article.

Public and Patient Involvement Statement

Jacques Laliberté, President ex-officio of the Association québécoise de la douleur chronique, was involved in the conception, design, development, interim and final results of the study. Members of the Advisory Board (Collège des medecins, Association québécoise de la douleur chronique, Association des Anesthésiologistes du Québec, Quebec Pain Research Network) were shown interim results and provided feedback.

Ethics Approval

The project was initially approved by the Central Research Ethics Committee of the Ministry of Health and Social Services (CCER). The CCER assumed the role of evaluating and monitoring the project for both private clinics and participating institutions, which were part of the health and social services network. On May 3, 2019, the Ethics Committee of McGill University Health Centre took over the role of the CCER for the QCR.

Author Disclosure Statement

M.W. initiated the QCR in 2015 and was the Principal Investigator until July 1, 2018, when he joined Canopy Growth Corporation, initially as Medical Director and most recently as Consultant. He has since served as a nonvoting Advisory Board member to the QCR. As of July 1, 2018, the QCR was lead by Dr. William Barakett (in memoriam) until March 2019 and is presently lead by A.V. A.V. is currently an advisory board member for EmpowerPharm and has formerly served in this capacity for Spectrum Therapeutics/Canopy Health, Tilray, Tetrabiopharma, Syqe. He was the former Research Director of Santé Cannabis in Montreal, Canada. Y.M. was a consultant for Spectrum Therapeutics/Canopy Health. No other authors have disclosures or conflicts of interest to declare.

Funding Information

This research was supported by grants from the Canadian Consortium for the Investigation of Cannabinoids, the Collège des Médecins du Québec and unrestricted grants from several licensed cannabis producers (Bedrocan, Mettrum and Tweed; these three companies merged during the study conduct into one company called Canopy Growth Corporation). Funding for data analysis has been received from the Alan Edwards Pain Management Unit of the McGill University Health Centre. Cedars Cancer Foundation and Rossy Cancer Network supported the Medical Cannabis Program in Oncology of the McGill University Health Center, which provided most data on oncology patients to the QCR. No funding agency had input into the design and conduct of the QCR, although the CMQ was present at development meetings

Supplementary Material

Supplementary Appendix SA1

Supplementary Appendix SA2

Supplementary Appendix SA3

Supplementary Appendix SA4

Supplementary Appendix SA5

Supplementary Appendix SA6

Cite this article as: Vigano A, Moride Y, Hachem Y, Canac-Marquis M, Gamaoun R, Kalaba M, Martel, MO, Perez J, Néron A, Beaulieu P, Desroches J, and Ware M (2023) The Quebec Cannabis Registry: investigating the safety and effectiveness of medical cannabis, Cannabis and Cannabinoid Research 8:6, 1106–1116, DOI: 10.1089/can.2022.0041.