Cannabis against chronic musculoskeletal pain: a scoping review on users and their perceptions

Abstract

Background

Chronic musculoskeletal pain (CMP) may lead to reduced physical function and is the most common cause of chronic non-cancer pain. Currently, the pharmacotherapeutic options against CMP are limited and frequently consist of pain management with non-steroidal anti-inflammatories, gabapentinoids, or opioids, which carry major adverse effects. Although the effectiveness of medical cannabis (MC) for CMP still lacks solid evidence, several patients suffering from it are exploring this therapeutic option with their physicians.

Objectives

Little is known about patients’ perceptions of their MC treatment for CMP. We aimed to increase this knowledge, useful for healthcare professionals and patients considering this treatment, by conducting a scoping literature review, following guidance by Arksey and O’Malley, to describe the views and perceptions of adult patients who had consumed MC to relieve chronic CMP.

Methods

Databases (PUBMED, EMBASE, Web of Science) and websites were searched using combinations of controlled and free vocabulary. All studies and study designs reporting on patients’ perceptions regarding MC against CMP were considered. Studies had to include adult patients reporting qualitatively or quantitatively, i.e., through questionnaires, on MC use to treat CMP or other non-cancer pain, since studies reporting exclusively on perceptions regarding CMP were very rare. Study characteristics were extracted and limitations of the study quality were assessed. The review includes patients’ demographic characteristics, patterns of MC use, perceived positive and negative effects, use of alcohol or other drugs, reported barriers to CM use, and funding sources of the studies.

Results

Participants of the 49 included studies reported that MC use helped them to reduce CMP and other chronic non-cancer pain, with only minor adverse effects, and some reported improved psychological well-being.

In the included studies, men represent between 18 and 88% of the subjects. The mean age of participants in these studies (42/49) varied between 28.4 and 62.8 years old. The most common route of administration is inhalation.

Conclusion

MC users suffering from CMP or other chronic non-cancer pain perceived more benefits than harms. However, the information from these studies has several methodological limitations and results are exploratory. These user-reported experiences must thus be examined by well-designed and methodologically sound clinical or observational studies, particularly regarding CMP, where reports are very scarce.

Background

Musculoskeletal pain

Musculoskeletal pain is a condition affecting bones, muscles, ligaments, and joints, resulting from underlying diseases or health problems such as osteoarthritis, inflammatory rheumatic diseases, and fibromyalgia, although in many cases the exact cause cannot be identified (Arthritis Society, 2015a). Musculoskeletal pain is the most common type of severe long-term pain and it impacts on all aspects of life by typically affecting dexterity and mobility, and by limiting work and activities of daily living (Woolf et al., 2012). It has been recently reported that one in two American adults lives with a musculoskeletal disease (Yelin et al., 2016), and in Canada, approximately 17% of the adult population are affected, nearly half of whom (44%) are aged 65 years or older (Arthritis Society, 2015a). Some cases of musculoskeletal pain are of short duration and have no long-term consequences. Chronic musculoskeletal pain (CMP), which persists for more than 3 months (Task Force on Taxonomy of the International Association for the Study of Pain, 1994), however, is associated with a range of problems such as sleep disorders, depression, anxiety, fatigue, reduced quality of life, and inability to work or socialize (Moore et al., 2014). In the USA, the impact of CMP on the economy in terms of healthcare costs and lost productivity is estimated at US $304 billion for the year 2013 (Yelin et al., 2016).

Effective pharmacological therapeutic options for the relief of CMP are limited and the treatment remains suboptimal for many patients (Fitzcharles et al., 2016). Examples for this are the use of non-steroidal anti-inflammatories, gabapentinoids (e.g., pregabalin and gabapentin), or the antidepressants duloxetine and milnacipran, which have shown clinical efficacy in the treatment of fibromyalgia and may have benefit in osteoarthritis and low back pain. However, it is estimated that only about one-third of patients will have at least 50% pain relief with one of these agents used as monotherapy; due to significant adverse effects, patients often fail to achieve recommended doses, further diminishing the medications’ effectiveness (Goldenberg et al., 2011). Opioids are also used to manage CMP, although the effectiveness of this approach remains uncertain (Petzke & Enax-Krumova, 2016; Schaefert et al., 2015) and the clinical management of CMP with opioids is challenging due to adverse effects such as dependence and/or addiction leading to possible overdose and death (Atluri et al., 2014; Ballantyne, 2015; Hauser et al., 2016; Tobin et al., 2016). It is therefore urgent to explore new treatment options to relieve pain in persons affected by CMP and thus improve their quality of life and social participation (Rowe & Caprio, 2013; Gereau et al., 2014; Lynch & Ware, 2015). Many persons for whom CMP is not satisfactorily relieved are turning to alternative therapies. Among these, the products derived from cannabis are perceived as an interesting analgesic option, both by some physicians and some patients (Elikottil et al., 2009; Boehnke et al., 2016), although its use remains controversial (Hosking & Zajicek, 2008; D'Souza & Ranganathan, 2015).

Cannabis and cannabinoids

The Cannabis sativa plant contains over 100 cannabinoids (ElSohly & Gul, 2014). The most abundant cannabinoid, delta-9-tetrahydrocannabinol (THC), is responsible for the main psychoactive effect of cannabis, but preclinical studies suggest that THC also has some analgesic and anti-inflammatory effects (Ashton, 2007). The second most abundant cannabinoid, cannabidiol (CBD), has antipsychotic effects and is not intoxicating (Niesink & van Laar, 2013; Zhu et al., 2006). Preclinical studies also support anti-inflammatory and analgesic effects of this compound (Burstein, 2015; Costa et al., 2007; Maione et al., 2011). The quantities and proportions of the different cannabinoids vary between different sources and preparations of cannabis (Ashton, 2001; de Meijer, 2014). Furthermore, there are differences between herbal preparations and consumption methods of cannabis regarding levels of individual cannabinoids, and between patients regarding the pharmacokinetics of these molecules (MacCallum & Russo, 2018). These differences affect treatment experiences (i.e., anxiety compared to relaxation), making it hard to come up with evidence-based information to guide physicians and patients on the most appropriate prescribing and dosing of cannabis for a given case (Beaulieu et al., 2016; Ko et al., 2016). Worldwide, several cannabinoid-based medicines are available in several countries. The first product, nabiximols (tradename Sativex®), contains the cannabinoids THC and CBD. The most common indication for its use is spasticity associated with multiple sclerosis. The second product, nabilone (tradename Cesamet®) contains a synthetic cannabinoid similar to THC and is used to alleviate nausea and vomiting associated with chemotherapy treatments. The third product, dronabinol (tradename Marinol®), is a synthetic cannabinoid chemically identical to THC and its main indications are anorexia associated with weight loss in patients with AIDS, as well as severe nausea and vomiting caused by cancer chemotherapy (Abuhasira et al., 2018). Quite recently, a product containing cannabidiol, Epidiolex®, has been approved by the US Food and Drug Administration for the treatment of Dravet syndrome and Lennox-Gastaut syndrome, which are severe epileptic encephalopathies.

Medical cannabis and musculoskeletal pain: gaps in knowledge

Given the confusion between the terms cannabis, cannabinoids, and cannabis for medical purposes, we will refer to the term “medical cannabis” (MC) in this review, in order to describe cannabis products (plant-based products or pharmaceutical products) used for CMP or other non-cancer chronic pain. Chronic pain in general, including CMP, is the most common reason given for the therapeutic use of MC among adults (Fitzcharles et al., 2016; Swift et al., 2005; Ware et al., 2005; Aggarwal et al., 2009; Arthritis Society, 2015b). The effectiveness of MC in the management of such pain, however, remains controversial. In a systematic review and meta-analysis on cannabinoids for medical use by Whiting et al., only 4 of the 79 trials included were judged at low risk of bias (Whiting et al., 2015). Individual studies suggested improvement in pain intensity, but most of the differences did not reach clinical significance and there was no clear evidence for an effect of the type of cannabinoid or the mode of administration. It is also important to note that different products were used in the individual studies, plant based or pharmaceutical, making comparisons between the studies even more difficult. Moreover, none of the studies assessed the long-term effects of cannabinoids.

In 2015, Lynch et al. published a systematic review of randomized controlled trials published since 2010 and examining cannabinoids for the treatment of chronic non-cancer pain, including CMP. They reported that seven out of the 11 included studies demonstrated a significant analgesic effect. Several trials also demonstrated improvement in secondary outcomes (e.g., sleep, muscle stiffness, and spasticity) (Lynch & Ware, 2015). Adverse effects most frequently reported, such as fatigue and dizziness, were mild to moderate in severity and generally well tolerated.

In 2017, the National Academies for Science, Engineering, and Medicine of the USA published an exhaustive review on the health effects of cannabis and cannabinoids and concluded that “there is conclusive or substantial evidence that cannabis or cannabinoids are effective for the treatment of chronic pain in adults”, based on a review of reviews, following the conclusions of Whiting et al. (Whiting et al., 2015), as well as two primary studies (National Academies of Sciences E, and Medicine, 2017). It should be pointed out, however, that the conclusions reported in the paper of Whiting et al. should be regarded with caution, as most of the studies assessed in this systematic review showed a high risk of bias.

In 2018, Stockings et al. performed another systematic review and meta-analysis of 47 randomized controlled studies and 57 observational studies on cannabinoids for the treatment of chronic non-cancer pain and concluded that the evidence for the effectiveness of MC on chronic non-cancer pain is limited [pooled events rates for 50% reduction in pain were not significant: 18.2% (cannabinoids) vs 14.4% (placebo); moreover, the number needed to treat was high (NNT = 24; 95% CI: 15–61) and the number needed to harm was low (NNH = 6; 95% CI: 5–8)]. From the results of the reviewed studies, the authors considered it unlikely that cannabinoids would become an important treatment option in chronic non-cancer pain (Stockings et al., 2018). Similarly, Nugent et al. reported in their 2017 review that the utilization of MC to alleviate chronic pain might be associated with several harms, including increased risk for motor vehicle accidents, psychotic symptoms, and short-term cognitive impairment, in addition to negative impacts on the respiratory tract (Nugent et al., 2017).

Thus, available evidence on the effectiveness of MC against CMP and other chronic non-cancer pain remains limited and the results of systematic reviews are somewhat inconclusive. It is even more difficult to conclude about the use of cannabis specifically in the management of CMP because, according to three systematic reviews of clinical trials on cannabis (Fitzcharles et al., 2016; Stockings et al., 2018), only two clinical trials have focused exclusively on musculoskeletal conditions. The authors of these clinical trials reported that cannabinoids (nabilone or Sativex®) led to a significant decrease in some aspects of pain in patients with fibromyalgia (Skrabek et al., 2008) or rheumatoid arthritis (Blake et al., 2006). However, only a small number of patients were studied for a short period of time in these trials and further methodological limitations may have affected their quality (Aviram & Samuelly-Leichtag, 2017) (Fitzcharles et al., 2016; Stockings et al., 2018). In conclusion, more high-quality randomized controlled trials comparing herbal cannabis or pharmaceutical cannabinoids with established therapies or placebo are necessary to define their role in the management of CMP or other chronic pain (Fitzcharles et al., 2016).

Although the use of MC remains controversial, it is gaining popularity and legal frameworks for its use are increasingly seen under certain conditions in a growing number of countries, i.e., Australia, France, Israel, the Netherlands, the UK, New Zealand, Spain, Germany, 29 US states, and since 1999 in Canada (Aguilar et al., 2018), where “serious arthritis” was mentioned as one of the main diagnoses justifying a license to obtain cannabis for medical use in 2013 (Arthritis Society, 2015b). Several countries are therefore already confronted with increasing use of MC against CMP, including self-medication, even though its efficacy and safety are still unknown.

Two recent reviews reported on MC use in patients suffering from different diseases, including anxiety, depression, HIV/AIDS, pain, and multiple sclerosis, highlighting that pain is the most frequent reason for MC use and its increasing frequency in general and cannabis self-medication in particular (Kosiba et al., 2019; Park & Wu, 2017). However, we did not identify major reviews on the characteristics, motivations, perceptions, and expectations of patients with regard to the use of medical cannabis against musculoskeletal or other chronic non-cancer pain. Thus, a knowledge gap exists in our understanding of patients’ characteristics and perceptions with regard to this use. Therefore, we conducted a scoping review to explore and describe these characteristics and perceptions of persons using MC against chronic non-cancer pain, including CMP. This review represents a first step towards a larger research program on this topic.

Methods

Eligibility criteria and selection of articles

The study protocol was submitted to the funding organizations and can be accessed through the corresponding author. Included studies had to comprise adults having used cannabis or cannabinoids for therapeutic purposes, including CMP or other chronic pain. Moreover, study samples had to have included at least several participants with chronic musculoskeletal or non-cancer pain. Qualitative, quantitative, and mixed methods studies were considered.

Studies that were specific to only one disease, other than musculoskeletal conditions or chronic non-cancer pain, such as HIV/AIDS, cancer, multiple sclerosis, epilepsy, inflammatory bowel disease, glaucoma, Tourette’s syndrome, neuropathic pain, spinal cord injury, migraine, post-traumatic stress disorder, dementia, or mental illness, as well as palliative care, were excluded. Furthermore, all studies that did not report any patient perceptions or results—including clinical trials on the therapeutic or adverse effects of cannabis—were excluded. Books, meeting abstracts, editorials, letters, policy evaluations, or newspaper articles were also excluded. Initial eligibility was assessed by screening the titles and abstracts of retrieved references by three persons Daniela Furrer, Martine Marcotte, and Norma Perez. Then, full texts of eligible references were reviewed by three persons (Daniela Furrer, Martine Marcotte, and Rosa Martins). Included publications that reported about one study in two or more articles were combined into a single study, with one exception (see below). Thereafter, reference lists of relevant reviews and of included studies were hand searched for additional references following the same procedure.

Information sources

Three large databases (MEDLINE, EMBASE, and Web of Science) were searched using keywords from the controlled vocabulary and free text, and combined to identify publications on users of cannabis for therapeutic purposes (see search strategies in Appendix 1). The searches were conducted during the second half of 2016, updated in June 2019, and were restricted to publications in English, French, or German with no other time limit.

Search strategy

This scoping review followed guidance by Arksey and O’Malley, Levac et al., and Colquhoun et al. (Arksey & O'Malley, 2005; Levac et al., 2010; Colquhoun et al., 2014) and examined the published knowledge regarding perceptions and experiences of MC users suffering from CMP or chronic non-cancer pain. Early search results revealed the scarcity of publications studying MC users for CMP specifically, and since CMP represents the most common etiology for chronic non-cancer pain, we expanded our search to all studies including patients using MC for chronic non-cancer pain (Podichetty et al., 2003). Moreover, given the scarcity of studies on the perceptions of users of MC, we decided to include both plant-based products and pharmaceutical products such as nabilone or nabiximols in the present review, similarly to some of the included studies (Hazekamp et al., 2013). As such, in the remainder of the manuscript, the abbreviation MC refers to both plant-based products and cannabis-derived medicine.

Data collection and quality appraisal

For this narrative synthesis, the following data were extracted by three persons into pre-determined Word files (Daniela Furrer, Martine Marcotte, and Rosa Martins) from the included studies: study design and setting, period of data collection, sample size, participants’ age and sex, indications for MC consumption, patterns of MC use, perceived benefits and adverse effects of use, and financial support for the study. When available, MC consumption as a substitute for other drugs, as well as barriers to MC use, were also documented. No individual quality appraisal was performed, according to the guidance used (Arksey & O'Malley, 2005; Levac et al., 2010; Colquhoun et al., 2014), but multiple limitations of the included study designs are outlined in the discussion.

Results

A total of 3639 references were first identified, and the full-text was screened for 201 articles, of which 52 publications reporting on 49 studies met the inclusion criteria (Fig. 1). In one publication (Perron et al., 2015), a sub-sample from a previous study (Ilgen et al., 2013) was used but, since study objectives and measures were different, they were treated as two different studies.

Fig. 1

Flowchart of the scoping review

Characteristics of the included studies

The main characteristics of all included studies are summarized in Fig. 2.

Fig. 2

Medical cannabis and musculoskeletal pain: scoping review key data

Among all included studies, only two examined the prevalence of cannabis use exclusively among patients suffering from CMP (Ste-Marie et al., 2016). Most of the studies focused on mixed samples that included patients with CMP (between 2 and 91% of participants) (31 studies) (Swift et al., 2005; Aggarwal et al., 2009; Hazekamp et al., 2013; Ilgen et al., 2013; Aggarwal et al., 2013a; Aggarwal et al., 2013b; Belle-Isle et al., 2014; Bottorff et al., 2011; Bruce et al., 2018; Coomber et al., 2003; Degenhardt et al., 2015; Erkens et al., 2005; Gorter et al., 2005; Haroutounian et al., 2016; Harris et al., 2000; Hoffman et al., 2017; Kilcher et al., 2017; Lucas & Walsh, 2017; Lynch et al., 2006; Nunberg et al., 2011; Ogborne et al., 2000; Pedersen & Sandberg, 2013; Piper et al., 2017; Reinarman et al., 2011; Schnelle et al., 1999; Sexton et al., 2016; Shiplo et al., 2016; Ste-Marie et al., 2012; Troutt & DiDonato, 2015; Walsh et al., 2013; Ware et al., 2003) or experiencing unspecified chronic non-cancer pain (between 24 and 97% of participants) (17 studies) (Boehnke et al., 2016; Perron et al., 2015; Alexandre, 2011; Bonn-Miller et al., 2014; Brunt et al., 2014; Corroon Jr. et al., 2017; Cranford et al., 2016; Crowell, 2017; Fanelli et al., 2017; Grella et al., 2014; Grotenhermen & Schnelle, 2003; Hazekamp & Heerdink, 2013; Reiman, 2009; Reiman et al., 2017; Shah et al., 2017; Webb & Webb, 2014; Zaller et al., 2015).

Funding

Funding information was reported in 28 of the 49 (57%) studies (Table 1); 23 studies were funded by research grants or governmental scholarships (Aggarwal et al., 2009; Perron et al., 2015; Ste-Marie et al., 2016; Aggarwal et al., 2013a; Aggarwal et al., 2013b; Belle-Isle et al., 2014; Bruce et al., 2018; Degenhardt et al., 2015; Erkens et al., 2005; Haroutounian et al., 2016; Harris et al., 2000; Hoffman et al., 2017; Lucas & Walsh, 2017; Pedersen & Sandberg, 2013; Sexton et al., 2016; Shiplo et al., 2016; Ste-Marie et al., 2012; Walsh et al., 2013; Brunt et al., 2014; Corroon Jr. et al., 2017; Cranford et al., 2016; Grella et al., 2014; Lavie-Ajayi & Shvartzman, 2018). Two studies were supported by non-governmental organizations (Hazekamp et al., 2013; Gorter et al., 2005). Five studies received mixed funding from research grants, non-governmental organizations, dispensaries or private foundations (Nunberg et al., 2011; Piper et al., 2017; Reinarman et al., 2011; Ware et al., 2003; Bonn-Miller et al., 2014; Lintzeris et al., 2018). Those five studies also had received funding from commercial cannabis interest or cannabis patient groups (Hazekamp et al., 2013; Gorter et al., 2005; Nunberg et al., 2011; Reinarman et al., 2011; Bonn-Miller et al., 2014; Lintzeris et al., 2018).

Table 1 Brief summary of included studies

Participants’ characteristics

Participants’ characteristics are described for each study in Table 1 and summarized in Table 4.

Patterns of MC use

Reported patterns of MC use for each study are presented in Table 2 and user experiences relating to the pattern or mode of use are shown in Table 3. The mode of cannabis administration was described in 36 studies. The most common form of MC consumption was inhalation (reported in 35 studies), either via smoking (joint or blunt, joint with tobacco, pipe, water pipe) or vaping (vaporizer) (Swift et al., 2005; Aggarwal et al., 2009; Hazekamp et al., 2013; Ste-Marie et al., 2016; Bottorff et al., 2011; Bruce et al., 2018; Coomber et al., 2003; Erkens et al., 2005; Haroutounian et al., 2016; Harris et al., 2000; Hoffman et al., 2017; Lucas & Walsh, 2017; Lynch et al., 2006; Ogborne et al., 2000; Piper et al., 2017; Reinarman et al., 2011; Schnelle et al., 1999; Sexton et al., 2016; Shiplo et al., 2016; Ste-Marie et al., 2012; Troutt & DiDonato, 2015; Walsh et al., 2013; Ware et al., 2003; Brunt et al., 2014; Cranford et al., 2016; Crowell, 2017; Fanelli et al., 2017; Grella et al., 2014; Grotenhermen & Schnelle, 2003; Shah et al., 2017; Zaller et al., 2015; Lavie-Ajayi & Shvartzman, 2018; Lintzeris et al., 2018; Reiman, 2007; Sagy et al., 2019). Reported smoking prevalence ranged from 20 (Erkens et al., 2005) to 91% (Cranford et al., 2016) and vaping prevalence from 7 (Crowell, 2017) to 53% (Shiplo et al., 2016). Ingested (cannabis tea, baked goods, oils, tinctures, tablets and capsules) (Hazekamp et al., 2013; Ste-Marie et al., 2016; Bruce et al., 2018; Erkens et al., 2005; Haroutounian et al., 2016; Hoffman et al., 2017; Lucas & Walsh, 2017; Lynch et al., 2006; Piper et al., 2017; Reinarman et al., 2011; Sexton et al., 2016; Troutt & DiDonato, 2015; Walsh et al., 2013; Ware et al., 2003; Brunt et al., 2014; Cranford et al., 2016; Crowell, 2017; Fanelli et al., 2017; Grella et al., 2014; Grotenhermen & Schnelle, 2003; Reiman et al., 2017; Shah et al., 2017; Zaller et al., 2015; Sagy et al., 2019) and topical administration (Ste-Marie et al., 2016; Bruce et al., 2018; Hoffman et al., 2017; Lucas & Walsh, 2017; Sexton et al., 2016; Cranford et al., 2016) were less common forms of MC use (reported in 25 and 6 studies, respectively). The reported prevalence of ingested MC varied from 0.5 (Sexton et al., 2016) to 70% (Erkens et al., 2005) and the prevalence of topical administration varied from 0.6 (Sexton et al., 2016) to 11% (Cranford et al., 2016). A combined mode of cannabis consumption (e.g., both smoked MC and edible MC products) was also reported (Haroutounian et al., 2016; Shiplo et al., 2016; Ste-Marie et al., 2012; Grotenhermen & Schnelle, 2003). Frequency and quantity of MC consumption was described in 23 (Swift et al., 2005; Aggarwal et al., 2009; Hazekamp et al., 2013; Coomber et al., 2003; Erkens et al., 2005; Harris et al., 2000; Lucas & Walsh, 2017; Lynch et al., 2006; Ogborne et al., 2000; Reinarman et al., 2011; Sexton et al., 2016; Shiplo et al., 2016; Troutt & DiDonato, 2015; Walsh et al., 2013; Ware et al., 2003; Bonn-Miller et al., 2014; Brunt et al., 2014; Cranford et al., 2016; Crowell, 2017; Grella et al., 2014; Shah et al., 2017; Zaller et al., 2015; Lintzeris et al., 2018) and 22 studies (Aggarwal et al., 2009; Hazekamp et al., 2013; Ste-Marie et al., 2016; Haroutounian et al., 2016; Harris et al., 2000; Lucas & Walsh, 2017; Lynch et al., 2006; Nunberg et al., 2011; Ogborne et al., 2000; Reinarman et al., 2011; Sexton et al., 2016; Shiplo et al., 2016; Ste-Marie et al., 2012; Troutt & DiDonato, 2015; Walsh et al., 2013; Bonn-Miller et al., 2014; Brunt et al., 2014; Cranford et al., 2016; Fanelli et al., 2017; Grotenhermen & Schnelle, 2003; Zaller et al., 2015; Lavie-Ajayi & Shvartzman, 2018; Sagy et al., 2019), respectively. Between 38 (Ware et al., 2003) and 90% (Brunt et al., 2014) of participants reported daily MC consumption. Consumed quantity of MC varied from 0.05 (Fanelli et al., 2017) to 1 gram per day (Harris et al., 2000).

Table 2 Patterns of medical cannabis use and utilization of medical cannabis as a substitute for prescription medications

Table 3 Perceived positive and negative effects

Medical cannabis used as a substitute for prescription medications

Of the 20 studies that examined the impact of MC use on the utilization of other prescribed medications (Boehnke et al., 2016; Swift et al., 2005; Bruce et al., 2018; Haroutounian et al., 2016; Lucas & Walsh, 2017; Lynch et al., 2006; Nunberg et al., 2011; Piper et al., 2017; Reinarman et al., 2011; Sexton et al., 2016; Troutt & DiDonato, 2015; Corroon Jr. et al., 2017; Crowell, 2017; Grella et al., 2014; Reiman, 2009; Reiman et al., 2017; Shah et al., 2017; Webb & Webb, 2014; Zaller et al., 2015; Lavie-Ajayi & Shvartzman, 2018; Sagy et al., 2019), 19 reported that MC consumption was accompanied by a decrease in the number and amount of prescribed drugs used, including opioids, antidepressants, anxiolytics and benzodiazepines, and non-opioid-based pain medication (Boehnke et al., 2016; Swift et al., 2005; Bruce et al., 2018; Haroutounian et al., 2016; Lucas & Walsh, 2017; Lynch et al., 2006; Nunberg et al., 2011; Piper et al., 2017; Reinarman et al., 2011; Sexton et al., 2016; Troutt & DiDonato, 2015; Corroon Jr. et al., 2017; Crowell, 2017; Grella et al., 2014; Reiman, 2009; Reiman et al., 2017; Webb & Webb, 2014; Zaller et al., 2015; Lavie-Ajayi & Shvartzman, 2018; Sagy et al., 2019) (Table 2). In twelve studies, it had been observed that participants discontinued their use of opioids or other prescription drugs following the start of MC consumption (Swift et al., 2005; Bruce et al., 2018; Haroutounian et al., 2016; Lucas & Walsh, 2017; Nunberg et al., 2011; Reinarman et al., 2011; Sexton et al., 2016; Corroon Jr. et al., 2017; Reiman, 2009; Webb & Webb, 2014; Zaller et al., 2015; Lavie-Ajayi & Shvartzman, 2018; Sagy et al., 2019), in a proportion varying from 6% (Webb & Webb, 2014) to 63% of participants (Lucas & Walsh, 2017). Participants also reported preferring the use of MC to prescription medication (Grella et al., 2014), mainly because of the adverse effects of their prescription drugs (Lynch et al., 2006).

Past and current use of cannabis and other licit and illicit substances

In 18 studies, 20 (Ste-Marie et al., 2016) to 90% (Harris et al., 2000) of participants reported that they had previously consumed cannabis recreationally or that they consumed it simultaneously to their therapeutic cannabis use (Swift et al., 2005; Hazekamp et al., 2013; Ste-Marie et al., 2016; Belle-Isle et al., 2014; Degenhardt et al., 2015; Erkens et al., 2005; Harris et al., 2000; Hoffman et al., 2017; Lucas & Walsh, 2017; Nunberg et al., 2011; Ogborne et al., 2000; Reinarman et al., 2011; Schnelle et al., 1999; Walsh et al., 2013; Ware et al., 2003; Grella et al., 2014; Grotenhermen & Schnelle, 2003; Shah et al., 2017; Lintzeris et al., 2018; Sagy et al., 2019) (Supplemental Table S1). One study reported that 29% of participants discovered the therapeutic effects of cannabis while using it recreationally (Swift et al., 2005). Six studies suggested that there might be a link between current MC use and past consumption of licit and illicit substances, as a proportion of MC users (3 to 89%) reported a past history of substance abuse, including alcohol, cocaine, amphetamines, hallucinogens, or other prescription drugs (Perron et al., 2015; Ilgen et al., 2013; Harris et al., 2000; Bonn-Miller et al., 2014; Grella et al., 2014; Zaller et al., 2015). Moreover, some MC users considered cannabis a substitute for alcohol (up to 26% of participants) (Lucas & Walsh, 2017) or illicit drugs (up to 16% of participants) (Zaller et al., 2015).

Reported barriers to the medical use of cannabis

Obstacles to the medical use of cannabis have been reported at several levels (Supplemental Table S2), including stigmatization from others (Ogborne et al., 2000; Piper et al., 2017), fear of discrimination (Belle-Isle et al., 2014), and physicians’ unwillingness to recommend, certify, or authorize MC (Aggarwal et al., 2009; Belle-Isle et al., 2014; Lucas & Walsh, 2017; Pedersen & Sandberg, 2013). Some MC users expressed health concerns such as pulmonary health or fear of addiction (Swift et al., 2005; Hoffman et al., 2017; Piper et al., 2017; Grella et al., 2014), but no study explicitly investigated perceived addiction to cannabis as a treatment consequence. Difficulties in finding a consistent and affordable MC supply and fear of legal problems associated with MC consumption, e.g., driving after consumption, represent further obstacles to MC utilization (Swift et al., 2005; Aggarwal et al., 2009; Aggarwal et al., 2013a; Aggarwal et al., 2013b; Belle-Isle et al., 2014; Coomber et al., 2003; Hoffman et al., 2017; Lucas & Walsh, 2017; Ogborne et al., 2000; Piper et al., 2017; Alexandre, 2011; Grotenhermen & Schnelle, 2003; Lintzeris et al., 2018).

Discussion

Main findings

In the included studies, a majority of participants who used cannabis for therapeutic purposes to relieve pain were aged 28.4 to 62.8 years in average with a proportion of men varying between 18 and 88% (Table 4). The most frequent mode of cannabis administration was smoking. The majority of MC users consumed cannabis daily, in a quantity ranging between 0.05 and 1 gram/day. MC users from reviewed studies reported positive effects on symptoms alleviation in addition to “secondary outcomes” such as psychological well-being. Reported adverse effects associated with MC utilization were few and of minor intensity and were mainly associated with cannabis smoking, such as negative impacts on pulmonary health. MC users also reported a reduction in the use of prescription drugs for the management of chronic pain (Boehnke et al., 2016; Swift et al., 2005; Bruce et al., 2018; Haroutounian et al., 2016; Lucas & Walsh, 2017; Lynch et al., 2006; Nunberg et al., 2011; Piper et al., 2017; Reinarman et al., 2011; Sexton et al., 2016; Troutt & DiDonato, 2015; Corroon Jr. et al., 2017; Crowell, 2017; Grella et al., 2014; Reiman, 2009; Reiman et al., 2017; Webb & Webb, 2014; Zaller et al., 2015; Lavie-Ajayi & Shvartzman, 2018; Sagy et al., 2019).

Table 4 Patients’ characteristics

Strengths and limitations of the review

To the best of our knowledge, this is the first comprehensive literature review on the perceptions of persons suffering from CMP or other chronic non-cancer pain, who used cannabis for therapeutic reasons. The information gathered in this review represents an opportunity to better understand the perspective of different types of MC users on the multiple dimensions of its consumption, in particular barriers, advantages, and drawbacks.

However, this review has several limitations, related principally to methodological weaknesses in an important proportion of the included studies.

Selection and recruitment of participants

For 41% of participants, they have been recruited at MC dispensaries, MC associations, or MC advocacy groups, including four studies performed in countries without a legal framework for access to MC (Swift et al., 2005; Coomber et al., 2003; Lintzeris et al., 2018; Pedersen et al., 2016). This might have introduced selection and information biases, as it has been reported that people who are already familiar with cannabis through recreational use, may use cannabis for medical reasons (Bigand et al., 2019; Lum et al., 2019). Indeed, among the about 30% of studies reporting on prior cannabis use, many MC users reported recreational cannabis use prior or simultaneously to MC use. Some MC users reported that it was during the recreational use of cannabis that they discovered its therapeutic effects. Moreover, people who are attending these centers may not use cannabis exclusively for medical reasons. In addition, MC users who had stopped MC consumption participated only marginally in these studies. Prevalence of adverse effects might therefore be underestimated. Furthermore, a subgroup of those studies, for which the source of funding was reported, was financially supported by cannabis interest or patient groups. This may have introduced a positive bias toward the use of cannabis against chronic pain. Therefore, we can argue that study participants were likely not representative of the general population with CMP or other chronic non-cancer pain, since a relevant subgroup of persons suffering from CMP or chronic non-cancer pain, but not considering MC as a therapeutic option, are not represented in the included studies. For instance, the mean age of MC users in the included studies (28.4-62.8 years) was lower than that of patients suffering from CMP, the incidence of which increases with age (Yelin et al., 2016). In addition, overall, the proportion of men in the included studies was higher than that of women, although CMP affects more often women than men (Yelin et al., 2016), suggesting a possible “gender effect”: with cannabis consumption being more popular among men than women (Carliner et al., 2017) and considering that individuals who already have consumed cannabis seem to be more disposed to use it as a therapeutic agent, it is possible that men are more likely to use cannabis for therapeutic purpose than women (Swift et al., 2005).

Use of MC with a medical prescription, in a dispensary or in self-medication

In addition, in many studies, it was difficult to distinguish between qualified and self-identified MC users, as it was not specified whether MC use was endorsed by a physician-confirmed diagnosis. It was impossible to estimate the prevalence of each type of user in all selected studies and it was thus not possible to estimate the overall prevalence of self-medication in these studies. Prevalence of self-medication is an important aspect, as it is increasing (Park & Wu, 2017), but self-identified MC users may have different characteristics than qualified MC users. It may become important for physicians to consider the possibility of self-medication with cannabis among their patients with CMP or other chronic pain.

Other methodological concerns

The included studies also varied greatly in terms of objectives, methodology, and participants’ populations, with 13 studies out of 49 (27%) having less than 100 participants. Data obtained during interviews or from questionnaires were self-reported and may suffer from recall or social desirability bias, while chart reviews may not have allowed to capture patient perceptions. The different legal frameworks regarding MC use across the different countries and periods of time might have influenced the availability and quality of MC, the sample size of the studies, and the availability of information on MC users. The conditions permitting to be registered as a MC user as well as access to MC vary between countries, states, and over time. For example, MC can be obtained from pharmacies in the Netherlands (Erkens et al., 2005; Brunt et al., 2014; Hazekamp & Heerdink, 2013), from special dispensaries in some states of the USA (Aggarwal et al., 2009; Aggarwal et al., 2013a; Piper et al., 2017; Troutt & DiDonato, 2015; Bonn-Miller et al., 2014; Grella et al., 2014; Zaller et al., 2015), and since 2013 from registered producers in Canada (Ste-Marie et al., 2016; Lucas & Walsh, 2017; Shiplo et al., 2016), as reflected in the included studies with participants recruited at dispensaries, registration clinics, or through online advertisement.

Chronic musculoskeletal pain

Although our scoping review aimed to report on MC users dealing with CMP, we identified only two studies that specifically assessed this type of chronic pain (Ste-Marie et al., 2016). The remaining studies comprised various proportions of participants suffering from CMP or non-specified chronic non-cancer pain. This heterogeneity among MC users may have influenced the reported information on MC consumption and its effects, since no distinction has been made relative to participants’ disease. Considering that the pathophysiology of pain varies depending on the syndrome (McMahon et al., 2013), clinical characteristics of participants should be as homogeneous as possible in order to conclude on the effects of MC on participants’ pain perception. It is thus somewhat reassuring that the two articles reporting specifically on patients suffering from CMP observed similar results as the other studies reporting on more heterogeneous populations. Indeed, among 1000 consecutive rheumatology patients, Ste-Marie et al. observed that 28 patients consumed MC. In agreement with the other studies, the authors observed that MC users were younger than the other patients of this clinic (52.8 vs. 62.8 years) and were more likely to be male (P = 0.051). In addition, MC users had previously consumed cannabis recreationally and 39.3% of the MC users reported to consume cannabis recreationally, in addition to MC (Ste-Marie et al., 2016).

Gaps in the literature

We identified some gaps in the literature that need to be addressed to better understand patients’ utilization of MC against MCP and unspecified chronic non-cancer pain. First, future studies should include participants who have stopped MC consumption or do not want to consider it, in order to understand the reasons that lead to discontinuation or rejection of MC, such as stigmatization of cannabis users or onset of adverse effects associated with MC use. As an example, Zolotov et al. reported that among participants who consumed cannabis for medical reasons, including chronic non-cancer pain (47.5%), those who abandoned MC (20%) experienced more frequent adverse effects (dizziness, dehydrated mouth, fatigue, mild anxiety, and feeling “weird”) than those who continued MC use (P < 0.05) (Zolotov et al., 2016).

Supported by a recent literature review, it would be interesting to better understand the point of view of physicians to identify the major factors which impact the decision of prescribing or not medical cannabis for patients who suffered from chronic pain (Gardiner et al., 2019). This would bring new knowledge on whether prescribers need support during the informed decision-making regarding the use of MC to treat CMP. The debate among physicians whether or not to prescribe MC is ongoing and has recently been presented in the literature (Caulley et al., 2018). Moreover, a changing legal framework for recreational cannabis may influence the perception of physicians regarding treatment with MC.

The use of MC as a substitute for other drugs, including opioids and other prescription medications, will need to be investigated for improved decision-making regarding MC prescribing, since opioids present serious, well-documented adverse effects. Currently, clinical guidelines in some countries, e.g., Canada, only support the use of MC for specific medical conditions, including neuropathic pain, palliative cancer pain, chemotherapy-induced nausea and vomiting, and spasticity related to multiple sclerosis or spinal cord injury, especially for those patients who do not respond to conventional therapies (Allan et al., 2018). Further randomized clinical trials that evaluate the efficacy and safety of MC in the management of CMP, other chronic pain or as substitute for opioids are urgently needed, but methodological challenges remain, including difficulties in participants’ recruitment and follow-up, and the surveillance of adverse effects.

Conclusion

Although the included studies are frequently exploratory and might be biased by several factors, they describe the perspective of MC users and allow a better understanding of their attitudes and experiences regarding MC use against CMP and other chronic non-cancer pain. These users perceive MC to have more benefits than drawbacks regarding quality of life and adverse effects, and several report on the possibility that MC might decrease the use of some prescription drugs, particularly opioids. However, these user reported experiences must be examined by well-designed and methodologically sound clinical or observational studies before any conclusions can be drawn.