Abstract
Background:
There has been growing interest in the use of cannabis and cannabinoids to treat chronic non-cancer pain (CNCP). Cannabis and cannabinoids have attracted attention because of their greater safety compared with opioids, and the possibility that their use can reduce opioid dose requirements via an opioid-sparing effect. Both factors have been proposed to contribute to fewer opioid-related deaths.
Methods:
We used The Pain and Opioids IN Treatment (POINT) study, a national cohort of 1,514 people living with CNCP prescribed opioids, to examine relationships between cannabis use, opioid use and pain outcomes over four years.
Outcomes:
Cannabis use was common, and by four-year follow-up, 24.3% had used cannabis for pain. Interest in using cannabis for pain doubled from 33% (baseline) to 60% (four years). We found that patients who had used cannabis had greater pain severity and interference, lower pain self-efficacy, and greater GAD severity than patients who had not used cannabis. We found no evidence of a temporal relationship between cannabis use and pain severity or pain interference, and no evidence that cannabis use reduced prescribed opioid use or increased rates of opioid discontinuation.
Interpretation:
Cannabis use was common in people living with CNCP prescribed opioids, but we found no evidence that cannabis use improved patient outcomes. Those who used cannabis had greater pain and lower self-efficacy in managing pain and there was no evidence that cannabis use reduced pain severity or interference or exerted an opioid-sparing effect.
Introduction
The use of prescribed opioids in the treatment of chronic non-cancer pain (CNCP) is controversial, due to a lack of evidence of their long-term effectiveness1,2 and increased opioid harms in countries as opioid prescribing for CNCP has increased3,4.
Alternatives to opioids are increasingly being debated and considered. Recent reviews of cannabinoids suggest they may have some efficacy in some CNCP conditions5–7. In Australia8, the United States (U.S.)9, Canada10 and the Netherlands11, CNCP is the most commonly cited reason for using cannabis for medicinal purposes. There has also been increasing discussion about the potential opioid-sparing effects of cannabinoids12. Changes in regulations mean that it is likely that there will be an increase in use of cannabinoid products for CNCP.
Longitudinal studies of cannabis use among people with CNCP are limited. Randomised controlled studies typically exclude those with complex physical, substance use and mental health comorbidities, which comprises a substantial number of people living with CNCP16. There is limited evidence on efficacy in the most common CNCP conditions, namely back or neck problems, arthritis and migraine7,13There is a lack of long-term follow-up in prospective studies, with the majority being 12-months or less17–19. Discussion about the opioid-sparing effects of cannabinoids has often been confined to ecological studies or cross-sectional surveys, which are poorly suited for testing causal hypotheses.
We used The Pain and Opioids IN Treatment (POINT) study, a national cohort of people living with CNCP prescribed opioids, to examine cannabis use and pain outcomes over four years. We aimed to examine:
Cannabis use over four years in people living with CNCP and who had been prescribed opioids, including their reasons for use and perceived effectiveness of cannabis;
Associations between level of cannabis use in the past month and pain, mental health and opioid-use;
The impact of cannabis use on pain severity and interference over time, while controlling for potential confounding of demographic and clinical variables;
Potential opioid-sparing effects of cannabis, controlling for potential confounding variables.
Method
The study was approved by the Human Research Ethics Committee of the University of New South Wales (HREC reference: #HC12149 and #HC16916). Full details of the study design and measure included have been published elsewhere16,20.
Participants
POINT participants were recruited through community pharmacies across Australia (see Appendix Figure B1 for more details). They were: 18 years or older; living with CNCP (defined in this study as pain lasting longer than three months); taking prescribed Schedule 8 opioids (including morphine, oxycodone, buprenorphine, methadone and hydromorphone) for CNCP for greater than six weeks; competent in English; mentally and physically able to participate in telephone and self-complete interviews; and did not have any serious cognitive impairments, as determined by the interviewer at the time of screening. A history of injecting drug use was not an exclusion criterion, but people currently prescribed pharmaceutical opioids for opioid substitution therapy for heroin dependence or for cancer were not eligible. Of 2,091 people assessed for eligibility, 90% (n=1,873) were eligible and 1,514 completed the baseline interview (n=359 refused after being deemed eligible and 74 could not be contacted, see Appendix FigureB2). At each follow-up wave, at least 80% of the original participants completed the interview (Figure 1). Details of the interview procedure are located in Appendix B1. Figure 1 here
Figure 1:
POINT study flow chart
. Baseline interviews were conducted in 2012–2014, wave 1 interviews in 2013–2014, wave 2 in 2015, wave 3 in 2016 and Wave 4 in 2017.
Measures
The measures, tools, and data domains were based on recommendations made by the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT)21,22.
Demographics
Data on age, gender, relationship status and current work status were collected.
Pain and pain-related measures
Participants were asked about lifetime and past year chronic pain conditions, and duration of CNCP. Pain is only one of a range of core outcomes to consider when evaluating interventions for CNCP21; we used the pain severity and interference (how pain impacts on sleep, daily living, working ability and social interaction) subscales of the Brief Pain Inventory (BPI)23, with higher scores indicating greater pain severity/interference (score range 0–10).
Pain self-efficacy relates to a person’s beliefs about the extent to which they can carry out daily activities despite their pain; this was measured using the Pain Self-Efficacy Questionnaire (PSEQ)24 (score out of 60, higher scores indicating greater self-efficacy).
Participants were asked at baseline “Is your pain neuropathic? That is, pain that burns or tingles (either diagnosed by self or doctor)”.
Opioid treatment
Daily oral morphine equivalent (OME) doses of opioids, in mg per day, were estimated using conversion units established through synthesis of clinical references25, using the medication diary. At each follow-up, there was confirmation of whether participants were still taking a Schedule 8 opioid.
Cannabis use
Participants were asked about lifetime and past 12-month use, and number of days used in the past month, in general and for pain specifically. Frequency of cannabis use in the past month was categorised as ‘no use’ (0 days), ‘less frequent cannabis use’ (1–19 days) and ‘near daily/daily cannabis use’ (20+ days of cannabis use, approximately five times a week or more frequently).
Participants who reported lifetime use of cannabis for pain but had discontinued use were asked their reasons for doing so. Those who reported past 12-month cannabis use were asked further questions about reasons for use (See Appendix B). All participants were asked ‘if you had access to cannabis, would you want to use it?’ at each wave (excluding the one-year follow-up). Based on a similar question in the BPI, we asked participants to rate the effectiveness of cannabis on their pain on a scale of 0 (‘no relief’) to 10 (‘complete relief’).
Mental health
Current depression and generalised anxiety disorder were measured by the PHQ-9 and GAD-7 modules of the Patient Health Questionnaire26,27. Moderate-severe depression was defined as PHQ-9 score ≥ 1026; moderate-severe anxiety was defined as GAD-7 score ≥ 1027. The Composite International Diagnostic Interview 3.0 (CIDI) substance use module assessed lifetime ICD-10 harmful use and dependence28.
Statistical analysis
Analyses were conducted using STATA, version 15.0 (Stata Corporation, College Station, TX, USA). For descriptive statistics, means and standard deviations were computed where data were normally distributed; medians and inter-quartile ranges where data were skewed.
29Cross-sectional associations with cannabis use frequency
Multinomial logistic regression models were also used for univariate comparisons of people at each wave who reported ‘less frequent cannabis use’ and ‘near daily/daily cannabis use’ (cf. people who had not used cannabis). Variables included in the multinomial regressions were selected in the same manner as above. For interpretability, RRR’s for OME are reported per 100 units.
Additional analysis on the demographic and clinical associations between prevalent and incident cannabis use are presented and discussed in Appendix C pages 14–15.
Prospective associations between cannabis use and outcomes
Lagged mixed-effects models examined temporal associations between cannabis use and pain severity and interference and OME, incorporating a random-intercept for individuals to account for the repeated measures design. We examined unadjusted and adjusted associations between cannabis use (the exposure) and three outcomes: pain severity, pain interference and OME. We analysed data from four annual waves, with outcomes for the following year compared with people who used cannabis (‘less frequent cannabis use’ and ‘near daily/daily cannabis use’) and people who had ‘never used cannabis’. Variables identified in previous research as related to the outcomes were included in adjusted models29. We conducted four models. In the first model, we adjusted for previous wave outcome; in the second, we adjusted for age, gender, duration of pain, GAD severity and history of substance use. Additionally, for the analysis on pain severity we also controlled for OME; for pain interference we adjusted for pain severity and OME; for OME we adjusted for pain severity. In the third model, we further adjusted for PSEQ (we had some missing data, since the PSEQ was not collected at the one-year interview). We utilised the Stata command margins (or mimrgns for multiple imputation) to obtain adjusted means. Details of sensitivity analysis are located in Appendix D
Role of the funding source
The funder had no role in the design, conduct, analysis, interpretation of findings, or decision to publish this work. The corresponding author had full access to all the data in the study and final responsibility for the decision to submit for publication.
Results
Characteristics of the cohort at baseline
At baseline, the cohort (n=1,514) was 44% male (IQR 42–47) with a median age of 58 years (IQR 48–67) (Table 1). Just under half were unemployed (48.8%) and 31% had retired from work. Participants had been living with CNCP for a median of 10 years (IQR 4.5–20) and had been prescribed a strong opioid for a median of four years (IQR 1.5–10). The median OME taken was 75mg per day (IQR 36–150). The most common types of pain reported at baseline was back/neck pain (76.6%) followed by arthritis (61.6%), and comorbid pain was common, with participants reporting a median of 2 (IQR 2–3) chronic pain conditions at baseline in the preceding 12 months. Approximately two-thirds (61.9%) reported neuropathic pain at baseline.
Table 1:
Socio-demographic characteristics, pain, prescribed opioid use and cannabis use amongst the POINT sample, by wave
| Baseline (N=1514) % (n) |
1-year (N=1216) % (n) |
2-year (N=1277) % (n) |
3-year (N=1211) % (n) |
4 year (N=1214) % (n) |
|
|---|---|---|---|---|---|
| Demographics | |||||
| Age MDN (IQR) | 58 (48–67) | 58 (49–68) | 59 (50–69) | 60 (50–69) | 60 (50–70) |
| % Male | 44 (672) | 44 (542) | 44 (555) | 43 (524) | 43 (524) |
| Pain | |||||
| BPI pain severity (M, SD) | 5.1 (1.79) | 5.3 (1.9) | 5.0 (1.9) | 4.9 (1.9) | 4.8 (1.9) |
| BPI pain interference (M, SD) | 5.7(2.3) | 5.7 (2.4) | 5.4 (2.4) | 5.5 (2.4) | 5.4 (2.4) |
| Prescribed opioid use | |||||
| OME (MDN, IQR) | 75 (36–150) | 61 (24–135) | 63 (25–135) | 60 (22–126) | 56.8 (15–125) |
| % Discontinued opioids | - | 10.7 (131) | 13.7 (174) | 16.7 (162) | 20.2 (246) |
| Cannabis use | |||||
| % Lifetime use | 43.2 (649) | - | - | - | - |
| % Past 12 months | 12.9 (195) | 11.1 (135) | 13.3 (170) | 14.3 (173) | 15.8 (190) |
| % Past month use | 8.7 (126) | 9.4 (112) | 9.7 (123) | 10.9 (132) | 12.9 (155) |
| Proportion past month use: None 1–19 days 20–31 days |
91.3 (1319) 5.4 (78) 3.3 (48) |
90.6 (1085) 5.4 (65) 3.9 (47) |
90.4 (1151) 5.5 (70) 4.2 (53) |
89.1 (1078) 5.8 (70) 5.1 (62) |
87.1 (1047) 6.4 (78) 6.5 (79) |
| % Ever used for pain relief | 15.6 (237) | 18.0 (220) | 20.4 (260) | 22.0 (267) | 24.3 (295) |
| % Used for pain relief past 12 months | - | 9.8 (123) | 11.8 (151) | 12.0 (145) | 14.0 (168) |
| % Used for pain relief past month | 5.6 (85) | - | 8.7 (111) | 10.0 (121) | 11.1 (134) |
| Effectiveness of cannabis for pain (out of 10) (M, SD) | 6.5 (2.9) | 5.0 (3.5) | 7.3 (2.2) | 7.0 (2.2) | 7.2 (2.3) |
| % Would use it if had access | 32.6 (364) | - | 44.1 (562) | 53.6 (649) | 60.1 (723) |
Note. BPI: Brief Pain Inventory; OME: oral morphine equivalent; MDN: median; IQR: interquartile range; CI: confidence interval; M: mean; SD: standard deviation, -data not collected
Cannabis use
At baseline, two-fifths of the cohort (43.2%) reported ever using cannabis, 12.9% reported use in the past 12-months, and 8.7% reported past month use. Both past 12-month and past-month use increased steadily from baseline to the four-year timepoint (15.8% and 12.9%, respectively; see Table 1).
Approximately one-in-six (15.6%) reported that they had used cannabis for their pain in their lifetime. Past 12-month and past month reporting of cannabis use for pain also increased steadily over time. The percentage reporting use between 1 and 19 days (‘less frequent cannabis use’) in the month preceding interview remained relatively stable. The percentage reporting use 20+ days in the past month-31 (approximately five days a week or more, i.e. ‘near-daily/daily cannabis use’) increased from 3.3% at baseline to 6.5% at the 4-year follow-up.
At baseline, participants who had used cannabis for pain rated its mean effectiveness for their pain as around seven out of 10 (with 10 being “extremely effective”; Table 1). The percentage of participants reporting that they would use cannabis if they had access to it doubled from 33% at baseline to 60% at the four-year follow-up.
At the three and four-year follow-ups, participants who reported past month cannabis use were asked whether it influenced their use of opioid medication. The majority reported that cannabis had no effect on their use of opioid medication (3-year 77%; 4-year71%); one-quarter reported that they ‘sometimes’ or ‘regularly’ reduced their opioid medication when using cannabis (3 year 21%; 4 year 29%) see Figure Appendix C, page 11). There were no differences in age, gender, pain severity/interference or OME between cannabis users who reported cannabis ‘sometimes’ or ‘regularly’ reduced their opioid use, compared with those who said it had no such effect.
Of those currently using cannabis, the most common reasons for use at both the three-year and four-year follow-up were to relieve pain (82% at each year) and pain-related distress (3 year 64; 4 year 73%), to improve sleep (3 year 66%; 4 year 63%) and for general relaxation (3 year 72% and 4 year 64%) (see Figure Appendix C, page 12). Participants who had previously used cannabis for pain, but were no longer doing so, were asked about their reasons for stopping. The most common reasons were side effects (3 year 28%; 4 year 23%), legal concerns 3 year 26%; 4 year 18%), difficulties accessing cannabis (3 year 18%; 4 year 20%) and its ineffectiveness in relieving pain (3 year 22%; 4 year 12%) (Figure Appendix C, page 13).
Cross-sectional associations of pain-related factors with cannabis use
Table 2 presents univariate analyses of associations between level of involvement in cannabis use and a range of clinical variables. With few exceptions, at each wave, people who were using cannabis (less frequent or daily/near daily use) reported greater pain severity and pain interference, lower pain self-efficacy and higher levels of GAD than those not using cannabis (Table 2). The associations were consistent for less frequent and near daily users (Table 2). For example, at the 4 year interview, compared to people with no cannabis use, less frequent and daily/near daily had greater pain severity score (RRR 1.14, 95%CI 1.01–1.29; RRR 1.17 95%CI 1.03–1.32), greater pain interference score (RRR1.21 95%CI 1.09–1.35; RRR 1.14, 95%CI 1.03–1.26), lower pain self-efficacy scores (RRR 0.97, 95%CI 0.96–1.00; RRR 0.98, 95%CI 0.96–1.00) and greater GAD severity scores (RRR 1.07, 95%CI 1.03–1.12; RRR 1.10, 95%CI 1.06–1.15)
Table 2.
Bivariate cross-sectional associations between level of cannabis use in the past month (days of use) and pain, anxiety and medication use in the POINT cohort, by wave
| No cannabis use (A) |
Less frequent cannabis use (<20 days) (B) |
Daily/near daily cannabis use (20+ days) (C) |
Unadjusted1 | ||||
|---|---|---|---|---|---|---|---|
| B vs A | C vs A | ||||||
| RRR | p | RRR | p | ||||
| Duration of pain +(years; MDN, IQR) | 10 (4–20) | 12.5 (6–21) | 13 (5–21.5) | 1.00 (0.99–1.02) | 0.484 | 1.00 (0.98–1.02) | 0.901 |
| BPI Pain Severity score (M, SD) | |||||||
| Baseline | 5.1 (1.8) | 5.3 (1.9) | 5.1 (1.4) | 1.09 (0.96–1.24) | 0.19 | 1.00 (0.86–1.19) | 0.906 |
| 1-year | 5.3 (2.0) | 5.4 (1.8) | 5.6 (1.6) | 1.03 (0.90–1.17) | 0.703 | 1.09 (0.93–1.27) | 0.27 |
| 2-year | 5.0 (1.9) | 5.4 (1.9) | 5.6 (1.9) | 1.12 (0.98–1.27) | 0.090 | 1.20 (1.03–1.39) | 0.020 |
| 3-year | 4.8 (1.9) | 5.4 (1.8) | 5.5 (1.6) | 1.19 (1.04–1.36) | 0.011 | 1.21 (1.05–1.40) | 0.0081 |
| 4-year | 4.7 (1.9) | 5.2 (1.9) | 5.3 (1.8) | 1.14 (1.01–1.29) | 0.031 | 1.17 (1.03–1.32) | 0.013 |
| BPI Pain Interference score (M, SD) | |||||||
| Baseline | 5.6 (2.3) | 6.0 (2.2) | 6.2 (1.5) | 1.08 (0.98–1.21) | 0.13 | 1.13 (0.99–1.30) | 0.078 |
| 1-year | 5.6 (2.4) | 6.2 (2.2) | 6.4 (2.0) | 1.11 (0.99–1.24) | 0.076 | 1.15 (1.01–1.31) | 0.039 |
| 2-year | 5.3 (2.4) | 6.2 (2.3) | 6.2 (1.8) | 1.18 (1.05–1.31) | 0.0035 | 1.18 (1.04–1.33) | 0.010 |
| 3-year | 5.4 (2.4) | 6.5 (2.0) | 6.4 (2.0) | 1.23 (1.10–1.38) | <0.001 | 1.22 (1.08–1.38) | 0.0011 |
| 4-year | 5.3 (2.4) | 6.3 (2.3) | 6.0 (2.3) | 1.21 (1.09–1.35) | <0.001 | 1.14 (1.03–1.26) | 0.0091 |
| PSEQ Pain self-efficacy score (M, SD) | |||||||
| Baseline | 29.7 (13.6) | 26.4 (12.7) | 25.6 (9.8) | 0.98 (0.97–1.00) | 0.039 | 0.98 (0.96–1.00) | 0.048 |
| 1-yearb | - | - | - | - | - | - | - |
| 2-year | 33.7 (13.4) | 27.8 (11.2) | 29.6 (11.4) | 0.97 (0.95–0.99) | <0.001 | 0.98 (0.96–1.00) | 0.029 |
| 3-year | 34.4 (13.2) | 28.1 (12.3) | 28.6 (13.1) | 0.96 (0.95–0.98) | <0.001 | 0.97 (0.95–0.99) | 0.0008 |
| 4-year | 34.2 (13.9) | 30.2 (12.7) | 30.6 (13.3) | 0.97 (0.96–1.00) | 0.015 | 0.98 (0.96–1.00) | 0.026 |
| GAD-7 severity score (M, SD) | |||||||
| Baseline | 5.3 (5.3) | 7.2 (5.5) | 8.0 (5.5) | 1.06 (1.02–1.10) | 0.0023 | 1.09 (1.04–1.14) | 0.0007 |
| 1-yeara | 5.1 (5.3) | 7.4 (5.3) | 8.9 (7.2) | 1.07 (1.03–1.12) | 0.0012 | 1.11 (1.06–1.16) | <0.001 |
| 2-year | 4.5 (4.8) | 7.5 (5.5) | 6.9 (5.8) | 1.11 (1.06–1.15) | <0.001 | 1.09 (1.03–1.14) | <0.001 |
| 3-year | 4.5 (4.8) | 6.7 (5.5) | 8.1 (5.9) | 1.08 (1.03–1.13) | <0.001 | 1.12 (1.07–1.17) | <0.001 |
| 4-year | 4.3 (4.9) | 6.4 (5.1) | 7.3 (6.1) | 1.07 (1.03–1.12) | <0.001 | 1.10 (1.06–1.15) | <0.001 |
| OME (MDN, IQR) | |||||||
| Baseline | 70 (35–140) | 84 (38–188) | 90 (33–171) | 1.21 (1.01–1.44)** | 0.040 | 1.05 (0.80–1.37) | 0.72 |
| 1-year | 60 (23–135) | 88 (44–152) | 90 (31–240) | 1.05 (0.85–1.30) | 0.64 | 1.39 (1.18–1.63) | <0.001 |
| 2-year | 60 (24–135) | 87 (52–191) | 80 (30–165) | 1.12 (0.98–1.27) | 0.082 | 1.14 (0.99–1.30) | 0.063 |
| 3-year | 60 (22–120) | 71 (39–180) | 60 (23–138) | 1.15 (0.89–1.29) | 0.072 | 1.07 (0.89–1.29) | 0.47 |
| 4 year | 55 (15–124) | 63 (23–135) | 49 (8–135) | 1.04 (0.88–1.22) | 0.65 | 1.01 (0.85–1.21) | 0.89 |
| % discontinued opioids | |||||||
| 1-year | 10.8 (9.1–12.8) | 9.2 (4.1–19.4) | 10.6 (4.3–23.8) | 0.59 (0.21–1.65) | 0.31 | 0.88 (0.31–2.52) | 0.81 |
| 2-year | 13.8 (11.9–15.9) | 7.1 (2.9–16.3) | 18.9 (10.2–32.1) | 0.48 (0.19–1.21) | 0.12 | 1.44 (0.71–2.94) | 0.304 |
| 3-year | 16.8 (14.7–19.1) | 15.7 (8.8–26.5) | 16.1 (8.7–27.8) | 0.92 (0.48–1.79) | 0.81 | 0.95 (0.48–1.91) | 0.89 |
| 4-year | 20.9 (18.6–23.5) | 9.0 (5.1–19.4) | 21.5 (13.7–32.2) | 0.38 (0.17–0.83) | 0.016 | 1.05 (0.60–1.84) | 0.85 |
Note. Table 1 provides n for each wave.
only asked at baseline
not asked at this wave.
RRR based on per 100 unit
Values highlighted in bold where p<.05.
BPI: Brief Pain Inventory; PSEQ: Pain Self-Efficacy Questionnaire; OME: oral morphine equivalent; GAD-7: Generalized Anxiety Disorder 7-item scale.
Data on PSEQ not collected at 1-year timepoint
Results of multivariate are provided in the appendix.
There were very few differences in OME consumption or rates of opioid discontinuation between those using cannabis at different frequencies. The exception was that people who reported ‘less frequent’ cannabis use had lower opioid discontinuance rates at 4 years (9%) than those reporting no use (21%), despite no difference in OME at the 4-year.
Temporal associations between prior cannabis use and current pain severity, pain interference and oral morphine equivalent (OME) consumption
Lagged mixed effect models examined the effect of past cannabis use on current pain severity (Table 3), current pain interference (Table 4) and current OME consumption (Table 5) in people using cannabis compared with those not using cannabis (complete case analysis; for multiple imputation analysis see Appendix C). In the unadjusted model, near daily or daily cannabis users had significantly greater pain severity (β 0.53, 95%CI 0.27–0.80) than people who had not used cannabis (difference of 0.5 on a 10-point scale). This difference, though still significant, was mediated by the inclusion of previous pain severity score (β 0.21, 95%CI 0.01–0.40). In adjusted models, including clinical covariates and pain self-efficacy there was no association between past cannabis use and current pain severity.
Table 3:
Lagged mixed-effects linear regression examining the impact of cannabis use in the previous wave on pain severity in the following wave (complete case analysis)
| Current level of pain severity | ||||
|---|---|---|---|---|
| Adjusted mean (SE) |
β | 95% CI | P value | |
| Cannabis use in previous wave | ||||
| No cannabis use (ref) | 5.0 (0.05) | - | - | - |
| Less frequent use | 5.1 (0.12) | 0.16 | −0.07–0.39 | 0.18 |
| Near daily/daily use | 5.5 (0.13) | 0.53 | 0.27–0.80 | <0.001 |
| …adjusted for pain severity in previous wave | ||||
| No cannabis use (ref) | 5.0 (0.02) | - | - | - |
| Less frequent use | 5.0 (0.10) | 0.06 | −0.12–0.26 | 0.51 |
| Near daily/daily use | 5.2 (0.10) | 0.21 | 0.01–0.40 | 0.037 |
| …adjusted for previous pain severity and clinical covariates1 | ||||
| No cannabis use (ref) | 4.9 (0.03) | - | - | - |
| Less frequent use | 5.0 (0.13) | 0.35 | −0.01–0.71 | 0.061 |
| Near daily/daily use | 5.1 (0.14) | 0.45 | −0.21–1.11 | 0.18 |
| …adjusted for previous pain severity and clinical covariates1 and PSEQ | ||||
| No cannabis use (ref) | 4.9 (0.03) | - | - | - |
| Less frequent use | 5.1 (0.13) | 0.37 | −0.01–0.75 | 0.056 |
| Near daily/daily use | 5.2 (0.14) | 0.43 | −0.23–1.10 | 0.201 |
Covariates include: previous wave BPI severity, age, gender duration of pain, OME, GAD severity, baseline lifetime ICD10 substance use disorder, time. PSEQ: Pain Self-Efficacy Questionnaire
Table 4:
Lagged mixed-effects linear regression examining the impact of cannabis use in the previous wave on pain interference in the following wave (complete case analysis)
| Current level of pain interference | ||||
|---|---|---|---|---|
| Adjusted mean (SE) |
β | 95% CI | P value | |
| Cannabis use in previous wave | ||||
| No cannabis use (ref) | 5.4 (0.06) | - | - | - |
| Less frequent use | 5.8 (0.14) | 0.38 | 0.11–0.66 | 0.0065 |
| Near daily/daily use | 5.9 (0.15) | 0.46 | 0.15–0.77 | 0.0034 |
| …adjusted for pain interference in previous wave | ||||
| No cannabis use (ref) | 5.4 (0.03) | - | - | - |
| Less frequent use | 5.8 (0.12) | 0.32 | 0.08–0.55 | 0.0087 |
| Near daily/daily use | 5.6 (0.11) | 0.15 | −0.08–0.37 | 0.20 |
| …adjusted for previous OME and clinical covariates1 | ||||
| No cannabis use (ref) | 5.3 (0.03) | - | - | - |
| Less frequent use | 5.6 (0.14) | 0.33 | −0.23–0.89 | 0.25 |
| Near daily/daily use | 5.2 (0.15) | −0.56 | −1.41–0.28 | 0.19 |
| …adjusted for previous OME and clinical covariates1 and PSEQ | ||||
| No cannabis use (ref) | 5.4 (0.04) | - | - | - |
| Less frequent use | 5.7 (0.16) | 0.35 | −0.22–0.92 | 0.23 |
| Near daily/daily use | 5.2 (0.19) | −0.63 | −1.46–0.19 | 0.13 |
Covariates include: previous wave BPI interference, age, gender, duration of pain, BPI severity score, OME, GAD severity, baseline lifetime ICD10 substance use disorder, time. PSEQ: Pain Self-Efficacy Questionnaire
Table 5:
Lagged mixed-effects linear regression examining the impact of cannabis use in the previous wave on level of opioid use in the following wave (complete case analysis)
| Current opioid morphine equivalent (OME) mg use per day | ||||
|---|---|---|---|---|
| Adjusted mean (SE) |
β | 95% CI | P value | |
| Cannabis use in previous wave | ||||
| No cannabis use (ref) | 97.5 (2.77) | - | - | - |
| Less frequent use | 100.7 (7.46) | 3.31 | −11.74–18.36 | 0.67 |
| Near daily/daily use | 105.3 (13.44) | 7.84 | −18.75–34.44 | 0.56 |
| …adjusted for OME in previous wave | ||||
| No cannabis use (ref) | 96.3 (1.32) | - | - | - |
| Less frequent use | 91.7 (5.15) | −4.56 | −15.13–6.01 | 0.40 |
| Near daily/daily use | 100.3 (7.43) | 4.08 | −10.79–18.95 | 0.59 |
| …adjusted for previous OME and clinical covariates1 | ||||
| No cannabis use (ref) | 91.2 (1.45) | - | - | - |
| Less frequent use | 88.2 (6.78) | 1.05 | −31.25–33.35 | 0.95 |
| Near daily/daily use | 91.5 (8.88) | 27.64 | −28.87–84.15 | 0.34 |
| …adjusted for previous OME and clinical covariates1 and PSEQ | ||||
| No cannabis use (ref) | 85.5 (1.74) | - | - | - |
| Less frequent use | 95.1 (8.85) | 7.00 | 26.97–40.96 | 0.69 |
| Near daily/daily use | 97.1 (12.66) | 32.76 | −25.04–90.57 | 0.27 |
Covariates include: previous wave OME, age, gender, duration of pain, BPI severity score, GAD severity, baseline lifetime ICD10 substance use disorder, time. PSEQ: Pain Self-Efficacy Questionnaire
People who had used cannabis in the previous wave had greater pain interference in the subsequent wave than those who had not used cannabis (for less frequent use β 0.38, 95%CI 0.11–0.66; for near daily/daily β 0.46, 95%CI 0.15–0.77). In adjusted models, after controlling for age, gender, previous pain interference, pain factors (e.g., duration of pain, pain severity and pain self-efficacy) and OME, prior cannabis use was not independently associated with current pain interference.
We also failed to detect an association between cannabis use in the previous wave and reduced OME in the subsequent wave. A complete case analysis of the effect of past cannabis use on current OME using is presented in Table 5 (for analysis based on multiple imputation, which found similar results, see Appendix D, pages 16–18)). There was no association in the univariate model and no independent association after controlling for other variables (Table 5).
Sensitivity analysis
To examine the robustness of the findings we conducted the following sensitivity tests. Sensitivity analyses that used log transformations of OME (Appendix D, pages 19) and used OME in categories (0, 1–20mg, 21–90mg, 91–199mg and 200+ greater mg, Appendix D, pages 20) found similar results to those presented here. Post-hoc, we ran the mixed-effects models among participants who self-reported neuropathic pain (Appendix D, pages 21–23) and also adjusting for neuropathic pain (see Appendix D, pages 24–25) and found no significant effect of past cannabis use on pain severity, interference or OME.
Discussion
To our knowledge, this is one of the longest in-depth prospective studies of a community cohort of people with CNCP that examines the impacts of cannabis use on pain and prescribed opioid use. Cannabis use was common in the cohort, patients reported that it reduced their pain and the proportion interested in using cannabis for pain doubled over the four-year follow up. We found that patients who had used cannabis had greater pain severity and interference, lower pain self-efficacy, and greater GAD severity than patients who had not used cannabis.
We found no evidence of a temporal relationship between cannabis use and pain severity or pain interference, and no evidence that cannabis use reduced prescribed opioid use or increased rates of opioid discontinuation. The most common reasons for discontinuing cannabis use included side effects, lack of efficacy, access difficulties and legal concerns. Nonetheless, our data and other population surveys32, highlight growing community interest in using cannabis for pain.
It is possible that perceptions of efficacy and safety will increase in Australia after a recent legislative change33 allowing medicinal use. Specifically, amendments to the federal Narcotic Drugs Act 1967 decriminalising medicinal use and supply of cannabis and cannabinoids came into effect on 30 October 201613. Only a minority of the data in the 4-year follow-up were collected after this change, and very few individuals nationally have accessed cannabinoids for medicinal purposes, so our cohort primarily used illicitly produced cannabis. It is not surprising, then, that two of the main reasons for discontinuing cannabis use for pain were access difficulties and legal concerns, similar to previous findings8. The increased availability of medicinal cannabinoids may increase use among people living with CNCP in Australia, though there is still limited accessibility and licensed cannabinoid medications are expensive. Additionally,it is unlikely cannabis was consumed under the guidance of a medical practitioner. Expectancies around cannabis reducing pain and opioid use may be different from participants using medicinal cannabis. High quality double-blind, randomised, placebo-controlled trials, which also examine expectancy effects, which are lacking for most CNCP conditions, may shed further light.
There were inconsistencies in our findings between what participants reported and our statistical assessment of associations. Although participants who used cannabis reported that the mean effectiveness of cannabis on pain was seven out of a possible score of 10, in unadjusted cross-sectional and longitudinal analysis, people who used cannabis in the past month reported greater pain severity and inference than those who had not used cannabis in the past month. In adjusted longitudinal analyses, there was no association between cannabis and pain severity or interference. This finding is inconsistent with previous studies that have found cannabis reduced pain severity17–19.
In our cohort, CNCP patients who used cannabis reported significantly greater pain severity than those not using cannabis, consistent with surveys of medicinal users who report using cannabis because of a failure of more conventional treatments34,35. Those using cannabis with the intent of relieving their pain may comprise a patient population with more distress and poorer coping mechanisms as evidenced in our study by the lower pain self-efficacy scores for people who used cannabis. It may be that in the absence of cannabis use, pain severity and interference may have been worse. Importantly, however, this study supports recent research which suggests cannabis use is associated with reduced self-efficacy in managing 36depression and anxiety36. Although previous reviews have found moderate support for cannabis use in reducing pain in CNCP5–7,9, they have mainly relied on RCT studies in which people with complex comorbidities have been excluded. In light of the recent findings of Wilson36 and the current study, it is important future research focuses on self-efficacy and the complexity of patients with CNCP in order to better understand which type of CNCP patients might benefit from using cannabinoids.
Previous cross-sectional studies have suggested cannabis may have ‘opioid-sparing’ effects in people with CNCP37,38, though a systematic review revealed a lack of high-quality clinical studies testing potential opioid-sparing effects39. In the current study, using both cross-sectional and longitudinal analytic approaches, we found no evidence that cannabis use was associated with reduced opioid use or opioid cessation. This finding needs to be qualified, as participants had access only to illicit cannabis and were not taking cannabis as part of structured pain management under medical supervision.
Strengths and Limitations
Our study was unique in exploring temporal associations between cannabis use, pain and opioid use in a large cohort with multiple assessment waves and low attrition. There might be concern that we did not recruit a representative sample of people prescribed opioids for CNCP. We collected data from a random sample of 71 pharmacies on the characteristics of all customers obtaining opioids during their six-week recruitment window, which showed important similarities between the cohort we recruited and customers overall. Among all customers recorded purchasing opioids in these pharmacies, 52% were female (vs. 55% in the POINT cohort); and 7% were 18–34 years, 55% 35–64 years and 38% 65+ years (vs. 5%, 62% and 33% respectively, in the POINT cohort). Two thirds (63%) were prescribed oxycodone (vs. 62% in the POINT cohort), 16.5% prescribed morphine (vs. 15% in the POINT cohort), and 24% prescribed buprenorphine patches (vs. 21% in the POINT cohort).
Although data were self-reported, this method of collection is reasonably reliable41, particularly when there are no disincentives for being honest42. All participants were assured of confidentiality and that the data would be de-identified, however, we performed no independent checks of participant reports of cannabis use. Due to the illegality of cannabis during the study period, it is possible that cannabis use has been underreported. We also note however, that other epidemiological studies that have reported that cannabis use reduces opioid use also depend upon self-reported cannabis and opioid use 37,38,43,44.
Additionally, we recorded frequency of cannabis use, rather than quantity and type of cannabis, but there are major complexities in reliably measuring this, given variations in THC content and amounts consumed in a session of use45,46. Finally, although we found no significant association between cannabis use and pain, it is difficult to understand completely the effects of cannabis on pain in an observational study design.
Conclusion
Cannabis use is common in people living with CNCP prescribed opioids, and interest in medicinal use of cannabis is increasing. We found no evidence that cannabis use improved patient outcomes: those who used cannabis had greater pain and lower self-efficacy in managing pain. We found no evidence that cannabis use reduced pain interference or exerted an opioid-sparing effect.
Supplementary Material
Evidence before this study
There has been substantial interest in the potential utility of cannabinoids for use in chronic non-cancer pain (CNCP). We conducted and literature review using MEDLINE, Embase, PsycINFO, CENTRAL and clinicaltrials.gov were searched in July 2017 for and RCT and observational studies relating to all cannabinoid types and specific CNCP conditions and pain-related outcomes. We identified 91 publications, containing 104 studies, which included 47 randomised control trials and 57 observational studies. We found pooled change in pain intensity (standardised mean difference: −0.14, 95%CI −0.20, −0.08) was equivalent to 3mm on a 100mm visual analogue scale greater than placebo. Evidence was graded as moderate. Existing clinical studies of the effects of cannabinoids on CNCP mainly comprised RCTs conducted using a limited range of cannabinoids in a limited range of CNCP conditions and a lack of clarity in reporting of pain outcomes.
Added value of this study
To our knowledge, this is one of the longest in-depth prospective studies of a community cohort of people with a variety of different types of CNCP that examined the impacts of cannabis use on pain and prescribed opioid use over four years of follow up. Cannabis use was common in the cohort, patients reported that it reduced their pain, and interest in using cannabis for pain doubled in the cohort over the four-year follow up. Nonetheless, patients who had used cannabis had greater pain severity and interference, lower pain self-efficacy, and greater generalised anxiety disorder severity than patients who had not. Unlike recent reviews which suggest a positive impact of cannabinoids on pain and reduction in opioid use, we found there was no evidence of a temporal relationship between cannabis use and pain severity or pain interference, and no evidence that cannabis use reduced prescribed opioid use or increased rates of opioid discontinuation.
Implications of all the available evidence
Previous systematic reviews have suggested that there is moderate evidence that cannabinoids are effective for certain types of pain. Previous evidence has been limited due to studies with limited duration and excluding participants with complex clinical profiles. In our current 4-year prospective cohort of people prescribed opioids for CNCP, we did not find evidence supporting claims that cannabis and cannabinoids improve outcomes in CNCP, nor that they reduce prescription opioid use. To date, evidence that cannabinoids are effective for CNCP and aid in reducing opioid use is limited. Large, well-designed clinical trials are required to evaluate in which patients, cannabinoids may be effective in reducing pain severity, interference and opioid doses.
Acknowledgements
This study was funded by the Australian Government and the National Health and Medical Research Council. The Australian National Drug and Alcohol Research Centre, UNSW Sydney, is supported by funding from the Australian Government, under the Substance Misuse Prevention and Service Improvements Grant Fund. We thank our associate investigators members of the POINT advisory committee and the POINT research team. Most importantly, we would like to thank the members of our POINT cohort, whose willingness to share their experiences was essential.
Funding: National Health and Medical Research Council and the Australian Government
GC, BL, AP, SN, and LD are supported by National Health and Medical Research Council fellowships (#1119992, 1073858, 1109366, 1132433, 1041472 and 1135991). The National Drug and Alcohol Research Centre at University of New South Wales, Australia, is supported by funding from the Australian Government, under the Substance Misuse Prevention and Service Improvements Grant Fund. Some of the authors have received investigator-initiated untied educational grants from Reckitt Benckiser/Indivior for studies of buprenorphine-naloxone (BL, LD, and NL), buprenorphine depot (BL, LD, MF, and NL), naloxone (LD and MF), the development of an opioid-related behaviour scale (BL, LD, NL, and RB), projects regarding opioid dependence treatment (NL); and a study of opioid substitution therapy uptake among chronic non-cancer pain patients (BL, LD, NL, and RB). LD, BL, AP, and MF have also received an untied educational grant from Seqirus for studies of tapentadol.
Footnotes
Declaration of interests
TD declares no competing interests.
References
- 1.Chou R, Turner JA, Devine EB, et al. The Effectiveness and Risks of Long-Term Opioid Therapy for Chronic Pain: A Systematic Review for a National Institutes of Health Pathways to Prevention WorkshopEffectiveness and Risks of Long-Term Opioid Therapy for Chronic Pain. Annals of Internal Medicine 2015; 162(4): 276–86. [DOI] [PubMed] [Google Scholar]
- 2.Krebs EE, Gravely A, Nugent S, et al. Effect of opioid vs nonopioid medications on pain-related function in patients with chronic back pain or hip or knee osteoarthritis pain: The space randomized clinical trial. JAMA 2018; 319(9): 872–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Fischer B, Argento E. Prescription opioid related misuse, harms, diversion and interventions in Canada: a review. Pain physician 2012; 15(3 Suppl): ES191–203. [PubMed] [Google Scholar]
- 4.Fischer B, Rehm J. Revisiting the ‘paradigm shift’ in opioid use: Developments and implications 10 years later. Drug and alcohol review 2017: n/a-n/a. [DOI] [PubMed] [Google Scholar]
- 5.Nugent SM, Morasco BJ, O’neil ME, et al. The effects of cannabis among adults with chronic pain and an overview of general harms: A systematic review. Annals of internal medicine 2017; 167(5): 319–31. [DOI] [PubMed] [Google Scholar]
- 6.Whiting PF, Wolff RF, Deshpande S, et al. Cannabinoids for Medical Use: A Systematic Review and Meta-analysis. Jama 2015; 313(24): 2456–73. [DOI] [PubMed] [Google Scholar]
- 7.Stockings E, Campbell G, Hall W, et al. Cannabis and cannabinoids for the treatment of people with chronic non-cancer pain conditions: A systematic review and meta-analysis of controlled and observational studies. Pain accepted. [DOI] [PubMed] [Google Scholar]
- 8.Lintzeris N, Driels J, Elias N, Arnold J, McGregor I, Allsop DJ. Findings from the Cannabis as Medicine Survey (CAMS-16): an online survey of medical cannabis use in Australia . Medical Journal of Australia Accepted March 2018. [DOI] [PubMed] [Google Scholar]
- 9.National Academies of Sciences E, Medicine. The Health Effects of Cannabis and Cannabinoids: The Current State of Evidence and Recommendations for Research. Washington, DC: The National Academies Press; 2017. [PubMed] [Google Scholar]
- 10.Walsh Z, Callaway R, Belle-Isle L, et al. Cannabis for therapeutic purposes: patient characteristics, access, and reasons for use. The International journal on drug policy 2013; 24(6): 511–6. [DOI] [PubMed] [Google Scholar]
- 11.Hazekamp A, Heerdink ER. The prevalence and incidence of medicinal cannabis on prescription in The Netherlands. European journal of clinical pharmacology 2013; 69(8): 1575–80. [DOI] [PubMed] [Google Scholar]
- 12.Nielsen S, Sabioni P, Trigo JM, et al. Opioid-sparing effect of cannabinoids: a systematic review and meta-analysis. Neuropsychopharmacology 2017; 42(9): 1752. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Therapeutics Goods Administration. Guidance for the use of medicinal cannabis in the treatment of chronic non-cancer pain. In: Health Do, editor. Canberra; 2017. [Google Scholar]
- 14.Vos T, Barber RM, Bell B, et al. Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. The Lancet 2016; 386(9995): 743–800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lucas P Moral regulation and the presumption of guilt in Health Canada’s medical cannabis policy and practice. The International journal on drug policy 2009; 20(4): 296–303. [DOI] [PubMed] [Google Scholar]
- 16.Campbell G, Nielsen S, Bruno R, et al. The Pain and Opioids IN Treatment study: characteristics of a cohort using opioids to manage chronic non-cancer pain. Pain 2015; 156(2): 231–42. [DOI] [PubMed] [Google Scholar]
- 17.Haroutounian S, Ratz Y, Ginosar Y, et al. The Effect of Medicinal Cannabis on Pain and Quality-of-Life Outcomes in Chronic Pain: A Prospective Open-label Study. The Clinical Journal of Pain 2016; 32(12): 1036–43. [DOI] [PubMed] [Google Scholar]
- 18.Hoggart B, Ratcliffe S, Ehler E, et al. A multicentre, open-label, follow-on study to assess the long-term maintenance of effect, tolerance and safety of THC/CBD oromucosal spray in the management of neuropathic pain. Journal of Neurology 2015; 262(1): 27–40. [DOI] [PubMed] [Google Scholar]
- 19.Ware MA, Wang T, Shapiro S, Collet JP. Cannabis for the Management of Pain: Assessment of Safety Study (COMPASS). The journal of pain : official journal of the American Pain Society 2015; 16(12): 1233–42. [DOI] [PubMed] [Google Scholar]
- 20.Campbell G, Mattick R, Bruno R, et al. Cohort protocol paper: the Pain and Opioids In Treatment (POINT) study. BMC pharmacology & toxicology 2014; 15(1): 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Dworkin RH, Turk DC, Farrar JT, et al. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain 2005; 113(1–2): 9–19. [DOI] [PubMed] [Google Scholar]
- 22.Turk DC, Dworkin RH, Allen RR, et al. Core outcome domains for chronic pain clinical trials: IMMPACT recommendations. Pain 2003; 106(3): 337–45. [DOI] [PubMed] [Google Scholar]
- 23.Cleeland C The Brief Pain Inventory (BPI). 1991. [Google Scholar]
- 24.Nicholas MK. The pain self‐efficacy questionnaire: Taking pain into account. European journal of pain 2007; 11(2): 153–63. [DOI] [PubMed] [Google Scholar]
- 25.Nielsen S, Degenhardt L, Hoban B, Gisev N. A synthesis of oral morphine equivalents (OME) for opioid utilisation studies. Pharmacoepidemiology and drug safety 2016; 25(6): 733–7. [DOI] [PubMed] [Google Scholar]
- 26.Kroenke K, Spitzer R, Williams JBW. The PHQ-9: Validity of a Brief Depression Severity Measure. Journal of General Internal Medicine 2001; 16(9): 606–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Spitzer RL, Kroenke K, Williams JBW, Lowe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med 2006; 166(10): 1092–7. [DOI] [PubMed] [Google Scholar]
- 28.World Health Organization. Composite International Diagnostic Interview, Version 3.0. . Geneva: World Health Organization; 2001. [Google Scholar]
- 29.Degenhardt L, Lintzeris N, Campbell G, et al. Experience of adjunctive cannabis use for chronic non-cancer pain: findings from the Pain and Opioids IN Treatment (POINT) study. Drug and alcohol dependence 2015; 147: 144–50. [DOI] [PubMed] [Google Scholar]
- 30.Peters SA, Bots ML, Den Ruijter HM, et al. Multiple imputation of missing repeated outcome measurements did not add to linear mixed-effects models. Journal of clinical epidemiology 2012; 65(6): 686–95. [DOI] [PubMed] [Google Scholar]
- 31.Mücke M, Phillips T, Radbruch L, Petzke F, Häuser W. Cannabis‐based medicines for chronic neuropathic pain in adults. The Cochrane Library 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Australian Institute of Health and Welfare. National Drug Strategy Household Survey 2016: Detailed Findings In: Australian Institute of Health and Welfare, editor. Canberra: AIHW; 2017. [Google Scholar]
- 33.Troutt WD, DiDonato MD. Medical Cannabis in Arizona: Patient Characteristics, Perceptions, and Impressions of Medical Cannabis Legalization. Journal of psychoactive drugs 2015; 47(4): 259–66. [DOI] [PubMed] [Google Scholar]
- 34.Swift W, Gates P, Dillon P. Survey of Australians using cannabis for medical purposes. Harm Reduction Journal 2005; 2(1): 18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ware MA, Doyle CR, Woods R, Lynch ME, Clark AJ. Cannabis use for chronic non-cancer pain: results of a prospective survey. Pain 2003; 102(1): 211–6. [DOI] [PubMed] [Google Scholar]
- 36.Wilson M, Gogulski HY, Cuttler C, et al. Cannabis use moderates the relationship between pain and negative affect in adults with opioid use disorder. Addictive Behaviors 2018; 77: 225–31. [DOI] [PubMed] [Google Scholar]
- 37.Corroon JM, Mischley LK, Sexton M. Cannabis as a substitute for prescription drugs – a cross-sectional study. Journal of Pain Research 2017; 10: 989–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Reiman A, Welty M, Solomon P. Cannabis as a Substitute for Opioid-Based Pain Medication: Patient Self-Report. Cannabis and cannabinoid research 2017; 2(1): 160–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Nielsen S, Sabioni P, Trigo JM, et al. Opioid-Sparing Effect of Cannabinoids: A Systematic Review and Meta-Analysis. Neuropsychopharmacology 2017; 42(9): 1752–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Hasin DS, Sarvet AL, Cerdá M, et al. Us adult illicit cannabis use, cannabis use disorder, and medical marijuana laws: 1991–1992 to 2012–2013. JAMA Psychiatry 2017; 74(6): 579–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Darke S Self-report among injecting drug users: a review. . Drug and alcohol dependence 1998; 51(3): 10. [DOI] [PubMed] [Google Scholar]
- 42.Lance CE, Vandenberg RJ. Statistical and methodological myths and urban legends: Doctrine, verity and fable in the organizational and social sciences: Taylor & Francis; 2009. [Google Scholar]
- 43.Abuhasira R, Schleider LB, Mechoulam R, Novack V. Epidemiological characteristics, safety and efficacy of medical cannabis in the elderly. European journal of internal medicine 2018; 49: 44–50. [DOI] [PubMed] [Google Scholar]
- 44.Boehnke KF, Litinas E, Clauw DJ. Medical Cannabis Use Is Associated With Decreased Opiate Medication Use in a Retrospective Cross-Sectional Survey of Patients With Chronic Pain. The journal of pain : official journal of the American Pain Society 2016; 17(6): 739–44. [DOI] [PubMed] [Google Scholar]
- 45.Norberg MM, Mackenzie J, Copeland J. Quantifying cannabis use with the timeline followback approach: a psychometric evaluation. Drug and alcohol dependence 2012; 121(3): 247–52. [DOI] [PubMed] [Google Scholar]
- 46.Pol P, Liebregts N, Graaf R, Korf DJ, Brink W, Laar M. Validation of self‐reported cannabis dose and potency: an ecological study. Addiction 2013; 108(10): 1801–8. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.