Cannabis Use in Physicians: A Systematic Review and Meta-Analysis
1Université Clermont Auvergne, CNRS, LaPSCo, CHU Clermont–Ferrand, WittyFit, F-63000 Clermont-Ferrand, France; pierre-louis_naillon13@hotmail.fr (P.-L.N.);
2Université de Nantes, Laboratoire de Psychologie des Pays de la Loire, LPPL, F-44000 Nantes, France
3Université Clermont Auvergne, NPsy-Sydo, CHU Clermont–Ferrand, Addiction, F-63000 Clermont-Ferrand, France
4Université Clermont Auvergne, Clermont Auvergne INP, CNRS, Institut Pascal, CHU Clermont-Ferrand, F-63000 Clermont-Ferrand, France
5Sport and Physical Education, Hong Kong Baptist University, Kowloon CN-99230, Hong Kong
6Université Clermont Auvergne, CHU Clermont-Ferrand, Endocrinology Diabetology and Metabolic Diseases, F-63000 Clermont-Ferrand, France
7INSERM, CIC 1405 CRECHE Unit, CHU Clermont-Ferrand, F-63000 Clermont-Ferrand, France
8Institute of Health Sciences, The Jan Kochanowski University of Kielce, P-25-002 Kielce, Poland
*Correspondence: frederic.dutheil@uca.fr; Tel.: +33-6-88-22-48-48; Fax: +33-4-73-27-46-49Abstract
Background: Cannabis use by physicians can be detrimental for them and their patients. We conducted a systematic review and meta-analysis on the prevalence of cannabis use by medical doctors (MDs)/students. Method: PubMed, Cochrane, Embase, PsycInfo and ScienceDirect were searched for studies reporting cannabis use in MDs/students. For each frequency of use (lifetime/past year/past month/daily), we stratified a random effect meta-analysis depending on specialties, education level, continents, and periods of time, which were further compared using meta-regressions. Results: We included 54 studies with a total of 42,936 MDs/students: 20,267 MDs, 20,063 medical students, and 1976 residents. Overall, 37% had used cannabis at least once over their lifetime, 14% over the past year, 8% over the past month and 1.1 per thousand (‰) had a daily use. Medical students had a greater cannabis use than MDs over their lifetime (38% vs. 35%, p < 0.001), the past year (24% vs. 5%, p < 0.001), and the past month (10% vs. 2%, p < 0.05), without significance for daily use (0.5% vs. 0.05%, NS). Insufficient data precluded comparisons among medical specialties. MDs/students from Asian countries seemed to have the lowest cannabis use: 16% over their lifetime, 10% in the past year, 1% in the past month, and 0.4% daily. Regarding periods of time, cannabis use seems to follow a U-shape, with a high use before 1990, followed by a decrease between 1990 and 2005, and a rebound after 2005. Younger and male MDs/students had the highest cannabis use. Conclusions: If more than a third of MDs tried cannabis at least once in their lifetime, this means its daily use is low but not uncommon (1.1‰). Medical students are the biggest cannabis users. Despite being common worldwide, cannabis use is predominant in the West, with a rebound since 2005 making salient those public health interventions during the early stage of medical studies.
1. Introduction
Cannabis dependence is one of the most common drug use disorders [1]. Several studies have described cannabis use in the general population [2,3]. Even if there are some studies on drug issues in medical doctors [4,5,6], there are very few studies and heterogenous data on cannabis use by physicians. In the general population, the 2012 European annual report on drug use showed that 25% of the 15–64 year-old Europeans tried cannabis at least once in a lifetime, 6.8% used cannabis during the past year, 3.6% during the past month and 1% had a daily use [7]. However, there is no such prevalence of cannabis use synthesized for medical doctors. Despite a very high percentage of cannabis use among students [8], data are scarce in the medical field and there are a lack of comparisons with medical doctors. Some medical specialties may also be more prone to cannabis use because of working conditions such as stress at work and workload, for example [9,10,11]. Moreover, the use of cannabis is widely heterogeneous across the globe, depending on culture and specificities of regions of the world [12]. The trends towards a decrease in the use of cannabis over time is common in most regions [12]. However, no study has focused on the regional and time effect of the use of cannabis in medical doctors or medical students. Lastly, some sociodemographic features such as age and sex are common influencing factors of cannabis use—men traditionally being more frequent users than women [13].
Thus, we aimed to conduct a systematic review and meta-analysis on the prevalence of cannabis use by physicians, to assess the prevalence and frequency of use, to determine if the cannabis use happens once over a lifetime or more frequently, such as daily cannabis use. The secondary objectives were to report physicians’ cannabis use stratified by specialty, educational level (medical students or medical doctors), continents, periods of time and putative influencing factors such as age and sex.
2. Methods
2.1. Literature Search
The PubMed, Cochrane Library, PsycInfo, ScienceDirect and Embase databases were searched for entries up until 30 March 2022 with the following keywords: addict* AND (physician* OR doctor*) AND (marijuana OR cannabis). Details for the search strategy within each database are available in Supplement S1. To be included, studies needed to describe our primary outcome variable, i.e., the use of cannabis in physicians. We considered any frequency of cannabis use (from once in a lifetime to daily use). The search was not limited to specific years or languages. Reference lists of publications that met our inclusion criteria were manually searched to retrieve further articles. Two authors (Pierre-Louis Naillon and Jean-Baptiste Bouillon) conducted the literature searches, reviewed the abstracts, and, based on the selection criteria, decided the suitability of the articles for inclusion, and extracted the data. When necessary, disagreements were solved with a third author (Frédéric Dutheil) (Figure 1). This systematic review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines (Supplement S2). The study was not registered, as there were delays in procedures during the COVID-19 pandemic [14].
2.2. Data Extraction
The primary outcome was cannabis use in physicians (once over a lifetime/past year/past month/daily). Secondary outcomes were education level (student or doctor), medical specialty, country and continent, period of the study, and sociodemographic (age, gender and family status).
2.3. Quality of Assessment
We used the Newcastle–Ottawa Scale (NOS) to check the quality of included articles [15]. The maximum score was nine for the cohort and ten for the cross-sectional studies. Additionally, we also used the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) for cohort and cross-sectional studies (Supplement S3) [16].
2.4. Statistical Considerations
Statistical analysis was conducted using Stata software (v16, StataCorp., College Station, TX, USA). Extracted data were summarized for each study and reported as mean (standard deviation) and number (%) for continuous and categorical variables, respectively. The prevalence of cannabis use and 95% confidence intervals (95% CI) were estimated using random effects models assuming between- and within-study variability (DerSimonian and Laird approach) [10,17,18]. More specifically, we conducted four meta-analyses on the use of cannabis in physicians (once over a lifetime/past year/past month/daily). Then, for each meta-analysis, we stratified results depending on level of study (student, resident, or doctor), specialties, continents, and periods of time (before 1990, between 1990 and 2005, and after 2005). Statistical heterogeneity between studies was assessed using forest plots, confidence intervals and I2. The I2 statistic is the most common metric for measuring heterogeneity and is easily interpretable: heterogeneity is considered low for I2 < 25%, modest for 25–50%, and high for >50%. We aimed to conduct a sensitivity analysis by excluding studies not evenly distributed around the base of the metafunnel. We also proposed meta-regressions to investigate putative factors influencing the prevalence of cannabis use in physicians, such as level of study, specialties, continents, periods of time, and sociodemographic (age and gender). Results were expressed as regression coefficients and 95 CI. Type I error was fixed at a = 0.05.
3. Results
An initial search produced 3958 possible articles. The removal of duplicates and use of the selection criteria reduced the number of articles reporting cannabis use among physicians to 54 articles in the systematic review and 52 articles in the meta-analysis (Figure 1), because two articles focused only on addicted physicians [19,20]. The main characteristics of the studies are presented in Table 1. We describe below the articles included in the meta-analysis.
3.1. Quality of Articles
Using the NOS criteria, the studies demonstrated a low risk of bias, except for response bias and insufficient description of statistical tests in the cross-sectional studies and for ascertainment and adequacy of follow up for the longitudinal study (Figure 2). Results were similar using STROBE (Supplementary S4). Details for each study are available in Table 1.
3.2. Study Designs and Objectives
All studies were cross-sectional [8,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69], except for two cohorts [70,71]. All the 52 studies included described cannabis-use prevalence among physicians or medical students. The main objective was to assess the prevalence of multiple-substance use in physicians in most studies [8,20,21,22,23,25,26,27,28,29,30,31,33,34,35,36,39,40,41,42,43,44,45,47,48,49,50,52,53,55,56,57,58,59,60,61,62,63,64,65,66,67,69,70,71]. Six studies assessed health status and psychological wellbeing of physicians [19,30,37,40,51,68], four focused on cannabis-use prevalence and belief about cannabis among physicians [24,32,46,54], and one on physicians’ attitudes toward drug testing [38].
3.3. Recruitment of Physicians
Medical doctors were recruited randomly using the quota method and stratification, using a national or state physician database before mailing or a phone call [27,33,35,37,47,48,58], from lists of diplomas delivered from faculties [29,38,39,45,70], by mailing all physicians from a country/state using an inter-university or national database [31,34], and at a meeting in a medical school [46]. Students and residents were recruited in medical schools—either monocentric [8,21,22,23,25,26,28,40,42,49,50,51,52,53,54,57,59,61,62,63,64,65,66,67,69,70,71] or multicentric [24,30,32,36,41,44,55,60,68]—and at a convention or a festival [43,46].
3.4. Populations Studied
Sample size ranged from 46 [25] to 5426 [33,48]. In total, 42,297 physicians or medical students were included in this meta-analysis: 20,267 medical doctors (15 studies) [27,29,31,33,34,35,37,38,39,45,46,47,48,58,70], 20,063 medical students (37 studies) [8,21,22,23,24,25,26,28,30,32,36,40,42,43,44,46,50,51,52,53,54,55,56,57,59,60,61,62,63,64,65,66,67,68,69,70,71], and 1976 residents (two studies) [41,49].
Age was reported in 26 studies. Overall, the mean age was 25.74 years old (95 CI 22.67 to 28.81), ranging from 20 [68] to 53 [37] years old.
Gender was reported in 84% of the studies (n = 44). The mean percentage of men was 59% (95 CI 55 to 64%), ranging from 91% [45] to 29% [40]. Cannabis use by gender was described in 13 studies [26,28,30,41,42,44,51,52,56,57,62,68,70] (Supplementary S5).
Specialty was mostly not reported in 86% of the studies (n = 45), followed by anesthesiologists (four studies, n = 5075) [33,34,39,58], and general practitioners (three studies, n = 1269) [27,31,33].
The location of studies was always reported. Most studies were conducted in North America (twenty-five studies, n = 23,903), followed by Europe (twelve studies, n = 9963), South America (nine studies, n = 4685), South Asia (four studies, n = 3095) and Oceania (two studies, n = 651).
Other variables were less well described. Family status was reported in 12 studies [25,34,35,36,37,38,39,41,45,48,55,66]. The study level when first use of cannabis occurred was recorded in four studies [8,28,41,44].
3.5. Cannabis Use Assessment
Most studies used a self-administered questionnaire (postal or email, after a brief explanation of the study goal) (35 studies) [8,21,22,24,25,26,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,53,54,55,58,59,63,67,69]. Other studies collected data from an interview or class interview [23,28,51,52,56,57,60,61,62,64,66,68,71], from an interview and by post [70], and from a phone call [27]. The data collection method was unclear in two studies [50,65]. The definition of cannabis use was never described in included articles; however, we can assume a smoking use.
3.6. Frequency of Use and Period of Data Collection
A total of 32 studies (n = 29,521 physicians/medical students) reported cannabis use at least once over a lifetime [8,21,23,24,25,26,27,28,29,30,32,35,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,57,58,60,61,62,64,65,66,67,68,69,70], 26 studies (n = 26,500) over the past year [8,21,22,26,28,29,31,32,33,34,35,36,37,41,44,48,50,52,55,56,57,59,63,69,70,71], 20 studies (n = 17,341) over the past month [8,24,26,27,28,29,35,36,41,43,44,48,52,55,57,59,60,66,69,70], and 12 studies (n = 11,172) daily [8,24,26,28,30,35,41,48,55,59,63,66]. Publication occurred within two years of data collection for 54% of studies [21,22,24,25,27,28,29,32,35,36,37,39,40,42,45,47,48,50,51,52,56,58,62,67,68,69,70], within 2 to 5 years for 31% [26,30,31,34,38,41,43,44,46,49,53,54,55,59,61,66,71], and more than 5 years for 8% [23,33,57,60], and was not reported for 7% of studies [63,64,65]. Studies ranged from 1971 [24,25,32] to 2021 [58], with 15 studies before 1990 (n = 9137) [8,21,24,25,28,29,32,42,46,47,50,55,61,62,65], 18 between 1990 and 2005 (n = 21,756) [23,27,33,34,35,38,39,41,44,45,48,49,51,64,66,68,70,71], and 19 after 2005 (n = 11,404) [22,26,30,31,36,37,40,43,52,53,54,56,57,58,59,60,63,67,69] (Table 1).
4. Discussion
The main findings were that the prevalence of cannabis use in medical doctors/students over a lifetime is high, at around 37%. Daily use was rare but not uncommon, with 1.1‰ of medical doctors/students smoking daily. Medical students have the greatest use. Despite the fact that cannabis use is common both in developed and developing countries, there are some cultural differences, with a predominant use in the West. After a decrease in cannabis use after the 1990s, there has been a rebound since 2005. Young and male physicians seem to have higher cannabis use.
4.1. Cannabis Use by Physicians: A Public Health Issue
Cannabis is not a benign substance, and inhalation of cannabis smoke is more harmful than tobacco smoke, delivering 50 to 70% more carcinogens [72]. Cannabis also decreases the immune function, promotes cardiac arrythmias and anxiety, and can lead to schizophrenia for genetically predisposed people [72,73,74]. Cannabis can also exacerbate pre-existing psychosis [75]. Medical doctors are exposed to many stressors, from long working hours, sometimes at night, to life-and-death emergencies [76]. Stress may lead to addictive behavior [77,78,79], and consequently the medical profession seems more subjected to drug abuse and psychiatric disorders [80]. Cannabis may be used by physicians to decrease their stress, such as in post-traumatic stress syndrome [81]. Interestingly, consumers using cannabis as a stress-coping strategy are those with the greatest risk of addiction [82,83]. Unfortunately, no studies included in our meta-analysis reported the workload. Despite not studied for cannabis, medical doctors who smoke tobacco promote less cessation advice to patients [84,85,86]. The impact of the use of cannabis by medicals doctors on their practice warrants further studies. Quitting cannabis is quite hard, with less than 10% success at 6 months [87]. The predominant psychotropic component is Δ9 tetrahydrocannabinol (THC), and the major non-psycho-active ingredient is cannabidiol (CBD). Both THC and CBD are a partial agonist or antagonist of prototypical cannabinoid receptors CB1 and CB2 [88]. No pharmacotherapy treatments demonstrated efficacy—from nicotine replacement therapy to psychotropic drugs [89]—but there are effective psychosocial interventions [89], such as completing self-determined goals [87,90]. Very interestingly, targeting the microbiome as a therapeutic and diagnostic tool may also be a promising avenue of exploration in the forthcoming years, considering the role of the gut–brain axis in a wide range of substance-use disorders [91].
4.2. From Daily to Lifetime Use
Our results showed a logical decrease from lifetime, to year, to month, to daily use, in accordance with the literature [8,24,26,28,35,41,48,55,57,63,66]. In the general population, 25% of the 15–64 year-old Europeans tried cannabis at least once in a lifetime, 6.8% used cannabis over the past year, and 3.6% over the past month [7]. Results from our meta-analysis suggest a higher prevalence of cannabis use by physicians for lifetime and monthly use, which may be linked with work-related stress [76]. Cannabis use has already been studied in other stressful jobs, such as US military veterans who reported a 12% use over the past 6 months [92]—our results for physicians were still higher. Cannabis use may also influence cognitive performance. For example, cannabis multiples by two the risk of a fatal road accident [93]. We note that no study assessed whether cannabis was used at work or during the rest periods of physicians. The moment of use may be relevant when considering the putative side effects of cannabis on medical errors. Cannabis use by physicians is relevant to fitness-for-duty concerns, similar to those for other drugs and alcohol [94,95,96]. We also demonstrated that physicians with a daily use of cannabis were very rare and uncommon (0.1%), compared with the 1% of daily users in the general population [7], which may reflect a low percentage of addicted physicians. Moreover, cannabis addiction may be more linked to the amount of cannabis than to the frequency of use [97]. The amount of cannabis smoked has not been studied in any of our included studies, and could be a salient indicator.
4.3. Medical Students as the Heaviest Consumers
We demonstrated a greater use of cannabis in medical students compared to medical doctors, which could be in line with the desire for new experiences in youngers [98]. Medical students may also use cannabis at university to cope with stress or depressive episodes [99]. However, cannabis decreases memory function in students (and cannabis abstinence leads to improved memory), and is associated with the poorest academic performance [99], which could be related to the cannabis-induced hypodopaminergic anhedonia [100]. Early first use of cannabis is also a risk factor for schizophrenia and bipolar disease [101,102]. Following the example of successful alcohol prevention university [103,104], efficient preventive strategies should take place in universities. Nonetheless, all physicians should benefit from targeted preventive strategies. We did not show differences in cannabis use between specialties, but specialties were seldomly reported, which precluded robust conclusions. The most frequently reported specialties were anesthesiologists [33,34,39,58] and general practitioners [27,31,33]. Anesthesiologists are known to be at greater risk of use of psychoactive substances, probably due to overwork and easier access to drugs [105]. If general practitioners are the heaviest tobacco smokers [10], data are lacking for cannabis. To our knowledge, there are no data on co-addiction in medical doctors, i.e., the combination of smoking, alcohol, cannabis or other psychoactive drugs. As alcohol use may predict cannabis use, particularly in the youngest [106], a longitudinal follow-up may be of particular interest.
4.4. Cannabis Use Worldwide and through Time
We showed huge disparities among continents in the use of cannabis by physicians, following the trends of cannabis use in the general population [107]. Differences between continents may be explained by a complex interplay between laws, cultural and religion beliefs [108]. In the USA, cannabis use is more frequent in states that legalized cannabis [109], but the legal status has not been reported in our included studies. In Europe, cannabis legislation also differs widely among countries [110]. South America seems to occupy the middle ground of cannabis use in physicians, with huge differences in cannabis policies. Chile and Uruguay have legalized recreational use, while Peru and Bolivia have harsh laws restricting both medical and recreational use [26,111]. Most of our included studies from South America were from Brazil [52,53,56,57,58,60,63], which authorizes cannabis for both personal use and medical use [112]. The continent of Asia continent very repressive policies, explaining the low prevalence of cannabis use [111]. Paradoxically, cannabis was used in Central Asia from 12.000 years ago, and Ayurvedic medicine first used it in India 3.000 years ago [113]. The decline in cannabis started all around the world in the early 20th century, after the second International Opium Convention. The opium and drugs trade were restrained and regulated [113]. Several Western countries then “tolerated” cannabis and reintroduced permissive laws from the 1960s [113]. In Oceania, Australia legalized recreational cannabis use in 2016, and New Zealand is thinking about legalization [50,114]. We included only two studies from Oceania [40,50], but physicians seem to follow the general population of Oceania [107]. A UNO worldwide report showed the same trend in cannabis use by continents as our results, in particular with America and then Europe being the biggest users, far above Asia [107]. Even if there is no study from Africa in our meta-analysis, the same UNO report showed a high cannabis use in Africa, at levels between America and Europe [107]. Despite 1.4 billion people, Africa has less than 1% of all scientific articles on addiction [115]. Regarding periods of time, we demonstrated that cannabis use seems to follow a U-shape. Cannabis use was high before 1990, and so the decrease from the 1990s is coincident with the laws and regulations on cannabis and other drugs [116], followed by a rebound after 2005 that could be a consequence of more permissive laws [117]. The perception of cannabis evolved as a “non-risky” recreational use [117]. In the USA, the states that legalized medical usage of cannabis saw an increase in illicit cannabis use [118]. The consequences of legalization of cannabis are still under debate. As medical use of cannabinoids has become more available, and the need for an evidence-based evaluation of safety and efficacy is necessary [119].
4.5. Other Influencing Variables
Male physicians tend to have higher rate of cannabis consumption than women, in our study. In the general population, men use cannabis more frequently and in a higher quantity than women [13]. The effect of cannabis use is different, according to sex. Male users report improved memory, enthusiasm, musicality, and increased appetite, whereas women have a desire to clean and a loss of appetite [13]. We also demonstrated that a younger age is linked with cannabis use, in line with the high consumption of medical students.
Unfortunately, body mass index, physical activity, marriage, or co-addictions were lacking in the studies included in our meta-analysis. However, cannabis users are less likely to suffer from obesity [120] or to have low levels of physical activity [121,122]. Marriage was associated with a reduction in drug use—including cannabis [123]. Smoking tobacco increases the risk of illicit-drug use such as cannabis [124], and early cannabis use has been strongly associated with other illicit-drug use [125].
4.6. Limitations
Our study has some limitations. We inherited the limitations of all meta-analyses and the limitations of the individual studies of which they were comprised: the varying quality of studies, and multiple variations in study protocols and evaluation [126]. We conducted our meta-analyses only on published articles, so our results were theoretically exposed to a publication bias. We included only studies reporting physicians’ cannabis use, so our results were theoretically exposed to a selection bias. Most cross-sectional studies included in our meta-analysis described a self-report bias. Data were collected by self-administered questionnaire, not always anonymously, which could lead to errors, as it appears in other kinds of self-report questionnaires, where the self-report questionnaire and interview show differences in the answers [127]. Thus, the reporting of cannabis might have been underestimated by physicians. Our meta-analysis also had limitations regarding the definition of cannabis use, although we can assume it is smoking cannabis. The lack of data by specialty precluded further analyses. Comparisons among continents or time periods might also suffer from a different number of studies within each continent or each period. Further studies should assess the prevalence of cannabis use in the general population, which may also permit comparisons with cannabis use in medical doctors.
5. Conclusions
Despite a high prevalence (37%) of cannabis use among physicians over a lifetime, daily use is rare but not uncommon, with 1.1‰ medical doctors/students smoking daily. Medical students are the biggest cannabis users. Insufficient data precluded comparisons among medical specialties. Despite the fact that cannabis use is common both in developed and developing countries, there are some cultural differences, with a predominant use in the West. After a decrease in cannabis use since 1990, there has been a rebound since 2005 that should benefit from targeted preventive strategies. Young and male physicians seem to have higher cannabis use, making salient those public health interventions during the early stage of medical studies.
Appendix Group
Supplementary Materials
The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/medicines10050029/s1, S1. Details of the search strategy used within each database. S2. PRISMA statements. S3. Checklists used for quality assessment of included studies. S4. Methodological assessment of studies using STROBE criteria. S5. Methodological assessment of studies using NOS criteria for (a) ross sectional studies and (b) for cohort studies. S6. Prevalence of cannabis use by sex among physicians. S7. Meta-regressions and factors influencing prevalence of smoking among physicians over a lifetime. S8. Meta-regressions and factors influencing prevalence of smoking among physicians over the past month. S9. Meta-regressions and factors influencing prevalence of smoking among physicians during the past day.
Data Availability Statement
All relevant data are within the paper.
Conflicts of Interest
The authors declare no conflict of interest.
| Study | Country | Data Collection | Physicians | Specialty | Prevalence | (%) of | Cannabis | Users | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | % Men | Age | Overall | By Specialty | Residents | Students | Lifetime | Past Year | Past 6 Months | Last Month | Last Week | Daily Use | |||
| Ashton 1995 [51] | England | 1994 | 185 | 41.3 | 20.4 | X | 49.2 | ||||||||
| Ayala 2017 [36] | USA | 2015–2016 | 855 | 35.5 | 25.6 | X | 26.2 | 11.7 | |||||||
| Baldwin 1991 [44] | USA | 1987 | 2046 | 62.7 | 27.7 | X | 66.4 | 22.5 | 10 | ||||||
| Baptista 1993 [49] | Venezuela | 1990 | 191 | 49.7 | 31 | X | 7.3 | ||||||||
| Bazargan 2009 [37] | USA | 763 | 75.1 | 53 | X | 4 | |||||||||
| Beaujouan 2005 [34] | France | 2001 | 3453 | 63.3 | Anesthesiologists | 2.6 | |||||||||
| Boniatti 2007 [52] | Brasil | 2006 | 183 | 45.9 | 22.5 | X | 31.1 | 13.7 | 7.7 | ||||||
| Carvalho 2008 [53] | Brasil | 2005 | 465 | 57 | 21.5 | X | 14.4 | ||||||||
| Chan 2017 [54] | USA | 2014 | 236 | 52 | 30 | X | 53.8 | ||||||||
| Coleman 1997 [23] | USA | 1989–1991 | 152 | X | 32 | 3 | |||||||||
| Conard 1988 [55] | USA | 1975–1985 | 589 | 65 | 27.6 | X | 73.7 | 31.6 | 17.3 | 1.2 | |||||
| Cottler 2013 * [20] | USA | 2008–2009 | 99 | 76.7 | 45.6 | Addicted | X | 29.2 | |||||||
| Croen 1997 [71] | USA | 1991–1993 | 170 | 54.1 | X | 29.4 | |||||||||
| Da Silveira 2008 [56] | Brasil | 2007 | 456 | 54.2 | 21 | X | 16.4 | ||||||||
| De Oliveira 2009 [57] | Brasil | 1996–2001 | 248 | 52 | X | 27.3 | 20.2 | 13.8 | |||||||
| De Sousa2021 [58] | Brasil | 2020 | 978 | 65.6 | Anesthesiologists | 43.2 | |||||||||
| Engs 1980 [50] | Australia | 1980 | 431 | X | 18.3 | 7.5 | |||||||||
| Farrell 2019 [40] | New Zealand | 2019 | 220 | 29.5 | X | 35 | |||||||||
| Gignon 2015 [59] | France | 2011 | 171 | 43.8 | 22.1 | X | 77 | 14 | 7 | ||||||
| Hughes 1991 [41] | USA | 1987 | 1785 | 70 | 30 | X | 65.1 | 17 | 7 | 0.3 | |||||
| Hughes 1992 [48] | USA | 1990 | 5426 | 82.2 | X | X | 35.6 | 4.6 to 10.5 | 2.1 to 7 | 0.1 to 0.3 | |||||
| Kenna 2004 [35] | USA | 2002 | 104 | 74 | 49 | X | 51.9 | 3.8 | 1.9 | 0 | |||||
| Kory 1984 [8] | USA | 1980 | 165 | 71.5 | X | 75.8 | 43.6 | 21.8 | 13.4 | 8.5 | |||||
| Lambert Passos 2006 [60] | Brasil | 1998 | 1054 | 47.4 | 21.1 | X | 20.9 | 5.6 | |||||||
| Laporte 1977 [61] | Spain | 1974 | 808 | 63.7 | X | 9.6 | |||||||||
| Laure 2003 [27] | France | 202 | 75.7 | 45.6 | General practioner | 20 | 2 | ||||||||
| Linn 1990 [38] | USA | 1987 | 303 | 87.1 | 47.6 | X | 55 | 9 | |||||||
| Lipp 1971 [32] | USA | 1970 | 1063 | X | 49.8 | 29.9 | |||||||||
| Lipp 1972 [47] | USA | 1971 | 1314 | X | 25 | ||||||||||
| Lutsky 1993 [39] | Canada | 183 | 84.5 | X | Anesthesiologists | 30 | |||||||||
| Lutsky 1994 [45] | Canada | 824 | 91.9 | X | 16.2 | ||||||||||
| Mansky 1999 * [19] | USA | 576 | Addicted | 1.6 | |||||||||||
| McAuliffe 1984 [21] | USA | 1981–1982 | 134 | 76.1 | X | 61 | 28 | ||||||||
| McKay 1973 [62] | Scotland | 1971 | 749 | 68.7 | X | 13.3 | |||||||||
| Merlo 2017 [30] | USA | 2014 | 862 | 42.8 | X | 46.8 | 4.1 | 1.5 | |||||||
| Newburry-Birch 2001 [70] | UK | 1995–1999 | 122 | 34.4 | 58.3 | X | X | 46.8–65.5 | 21.9–23.6 | 11.4–11.8 | 2.7–7 | ||||
| Petroianu 2010 [63] | Brasil | 332 | 48.2 | X | 15.6 | 0.6 | 0.3 | ||||||||
| Pickard 2000 [64] | UK | 46 | 33.8 | X | 33.5 | ||||||||||
| Polakoff 1972 [46] | USA | 1969 | 395 | X | X | 13 to 42 | |||||||||
| Rai 2008 [43] | India | 2003 | 2135 | 70.8 | 20.5 | X | 6.6 | 1.5 | |||||||
| Rochford 1977 [65] | USA | 134 | X | 68.7 | |||||||||||
| Rodriguez 1986 [42] | Spain | 1984 | 2308 | 48.1 | X | 20.7 | 0.9 | ||||||||
| Romero 2009 [26] | Chile | 2005 | 569 | 55 | 21.5 | X | 33 | 19.7 | 5.1 | 0.17 | |||||
| Saeys 2014 [31] | Belgium | 2011 | 626 | 57.3 | 45 | General practioner | 4 | ||||||||
| Schwartz 1990 [66] | USA | 1987 | 263 | 64.6 | X | 43 | 5 | 3.5 | 1.5 | ||||||
| Shyangwa 2007 [22] | Nepal | 193 | 67.3 | 22.8 | X | 15 | |||||||||
| Singh 1979 [28] | India | 1976–1977 | 672 | 79.9 | X | 23.2 | 11 | 1.3 | 0 | ||||||
| Singh 1980 [29] | India | 1977–1978 | 95 | 75.7 | X | 20 | 3.2 | 0 | |||||||
| Slaby 1971 [25] | USA | 1970 | 46 | 89.1 | 25.7 | X | 52 | ||||||||
| Solursh 1971 [24] | Canada and USA | 234 | X | 54.3 | 32.1 | 22.7 | 8.5 | 1.3 | |||||||
| Vujcic 2017 [67] | Serbia | 2015 | 418 | 37.3 | 22.5 | X | 34.9 | ||||||||
| Webb 1998 [68] | UK | 1996 | 785 | 44 | 20 | X | 43.9 | 9.6 | |||||||
| Zhou 2015 [69] | USA | 2014 | 431 | 50.3 | 25 | X | 31.1 | 12.1 | 8.9 |