Table 1: Details of included studies.

Study IDMethodologyMethodological qualityPopulation/settingPrevalence resultsAdditional results

Al-Dubai and Rampal, 2012 [40](i) Sampling frame: four main government hospitals ( = not reported)  
(ii) Sampling method: universal sampling (all doctors)  
(iii) Recruitment method: in person  
(iv) Administration method: in person, self-administered
(i) Sample size calculation: no  
(ii) Sampling type: nonprobability sampling (broad sampling, covered all doctors)  
(iii) Validity of tool: for khat/tobacco: self-developed tool, no validation reported  
(iv) Pilot testing done: no  
(v) Response rate: 70.4%
(i) Population: Yemeni doctors  
(ii) Country: Yemen  
(iii) Setting: hospitals  
(iv) (sampled) = 800  
(v) (participated) = 563  
(vi) (included in analysis) = 563
(i) Prevalence of khat, (%): 248 (44.0%) (sometimes 21.7%, frequently 9.95%, and daily 12%)  
(ii) Prevalence of tobacco use (specify tobacco type and pattern of use), (%): 99 (17.6%) (cigarettes, pattern not specified)  
(iii) Prevalence of tobacco among khat chewers (specify tobacco type and pattern of use), cigarette, (%): 84/248 (33.9%)
(i) Level of dependence among khat chewers who use tobacco: not reported  
(ii) Associated factors: cigarette smoking among khat chewers is significantly associated with male gender (AOR 3.77 (1.10–12.92)), working in government and private compared to government only (0.40 (0.21–0.75)), and for working > 10 years compared to less than 10 years (AOR 3.54 (1.44–8.67)). Model also adjusted for age, marital status, residence, and income

Alem et al., 1999 [41](i) Sampling frame: nine peasant associations and one urban dwellers association ( = 15,000)  
(ii) Sampling method: cluster random sampling  
(iii) Recruitment method: in person (iv) Administration method: in person, self-administered (with assistance)
(i) Sample size calculation: no  
(ii) Sampling type: probability  
(iii) Validity of tool: self-developed tool, validation reported (questions rephrased after showing them to pilot group of forty adults)  
(iv) Pilot testing done: yes  
(v) Response rate: 85.1%
(i) Population: persons aged over 15 years resident in Butaijra town  
(ii) Country: Ethiopia  
(iii) Setting: house-house (community-based)  
(iv) (sampled) = 12,531  
(v) (participated) = 10,658  
(vi) (included in analysis) = 10,468
(i) Prevalence of khat, (%): 911 (8.7%) daily, 5234 (50%) current (pattern not specified)  
(ii) Prevalence of tobacco use (specify tobacco type), (%): 186 (1.8%) were mild cigarette smokers, 132 (1.3%) were moderate cigarette smokers, and 141 (1.3%) were heavy cigarette smokers  
(iii) Prevalence of tobacco among khat chewers (specific tobacco type), (%): among daily khat chewers, 46 (5.0%) were mild (1–3 daily) cigarette smokers, 47 (5.2%) were moderate (4–9 daily) cigarette smokers, and 59 (6.5%) were heavy (>9 daily) cigarette smokers  
(iv) Pattern of tobacco use among khat chewers: we only have data for daily cigarette smokers  
(v) Daily khat and tobacco (cigarette): 46 + 47 + 59 = 152 100/911 = 17%
(i) Level of dependence among khat chewers who use tobacco: not reported  
(ii) Associated factors: not reported

Al-Sanosy, 2009 [21] (i) Sampling frame: total number of students enrolled in 11 colleges in the region ( = 18243) and secondary schools ( = 46760), total 102 schools  
(ii) Sampling method: systematic random sampling  
(iii) Recruitment method: in person  
(iv) Administration method: in person, self-administered
(i) Sample size calculation: yes  
(ii) Sampling type: probability sampling  
(iii) Validity of tool: previously reported validated tool  
(iv) Pilot testing done: yes  
(v) Response rate: 89.7%
(i) Population: secondary school and college students, 51.7% male, mean age 18.9 (2.58) years  
(ii) Country: Saudi Arabia  
(iii) Setting: secondary schools and colleges, May 2006  
(iv) (sampled) = 10,000  
(v) (participated) = 8965  
(vi) (included in analysis) = 8965
(i) Prevalence of khat, (%) (specify pattern of use): past 30-day use: 1795/8965 (20.0%), daily khat chewers: 250/8965 (2.8%), most of week days: 322/8965 (3.6%), weekends: 765/8965 (8.5%), and occasionally: 468/8965 (5.2%)  
(ii) Prevalence of tobacco use (specify tobacco type and pattern of use), (%): smoking (not defined) 863/8965 (9.6%)  
(iii) Prevalence of tobacco among khat chewers (specific tobacco type and pattern of use), (%): 863/1101 (78.4%) of khat chewers were also cigarette smokers  
(iv) Mentioned only cigarettes
(i) Level of dependence among khat chewers who use tobacco:  
Not reported  
(ii) Associated factors: not reported

Alsanosy et al., 2013 [15](i) Sampling frame: intermediate educational level and high school students in the Jazan region  
(ii) Sampling method: cluster multistage sampling
(iii) Recruitment method: in person
(iv) Administration method: in person, self-administered
(i) Sample size calculation: yes  
(ii) Sampling type: probability sampling  
(iii) Validity of tool: previously reported validated tool  
(iv) Pilot testing done: yes  
(v) Response rate: 95.68%
(i) Population: full-time intermediate and high school students aged 13–21 years (75.1% aged 15–19 years, 56.3% male, and 61.3% urban residence)  
(ii) Country: Saudi Arabia  
(iii) Setting: schools in academic year 2011/2012  
(iv) (sampled) = 4100  
(v) (participated) = 3923  
(vi) (included in analysis) = 3923
(i) Prevalence of khat, (%): current (past 30 days): 806/3923 (20.5%, 95% CI 19.27–21.79); ever: 952/3923 (24.2%, 95% CI 22.9–25.57)  
(ii) Prevalence of tobacco use (specify tobacco type), (%): smoking status yes/no: 627 (17.3%)  
(iii) Prevalence of tobacco among khat chewers (specific tobacco type), (%): cigarettes 489 (54.3%).
(i) Level of dependence among khat chewers who use tobacco: not reported  
(ii) Associated factors: not reported

Ayana and Mekonen,
2004 [42]
(i) Sampling frame: all registered university students in 2001/2002 ( = 2073)  
(ii) Sampling method: systematic random sampling  
(iii) Recruitment method: not reported  
(iv) Administration method: in person self-administered
(i) Sample size calculation: yes  
(ii) Sampling type: probability sampling  
(iii) Validity of tool: previously reported validated tool  
(iv) Pilot testing done: yes  
(v) Response rate: 94.4%
(i) Population: university students, mean age 24 (range 16–46), 76.91% male, 59.5% orthodox, and 49.2% Amhara ethnicity  
(ii) Country: Ethiopia  
(iii) Setting: university, January 2002  
(iv) (sampled) = 500  
(v) (participated) = 472  
(vi) (included in analysis) = 472
(i) Prevalence of khat, (%) (specify pattern of use): current (daily, weekly, or occasionally) prevalence 117/472 (24.79%), divided into 52/472 (11.0%) every day, 35/472 (7.4%) once a week, and 30/472 (6.4%) occasionally  
(ii) Prevalence of tobacco use (specify tobacco type and pattern of use), (%): 64/472 (13.56%) were cigarette smokers (pattern not defined)  
(iii) Prevalence of tobacco among khat chewers (specific tobacco type and pattern of use), (%): among khat chewers, 53/117 (45.3%) were cigarette smokers. 16/117 (13.7%) of students smoke cigarettes while chewing khat; 8/117 (6.8%) of students smoke cigarettes after chewing khat
(i) Level of dependence among khat chewers who use tobacco: not reported  
(ii) Associated factors: not reported

Gorsky et al., 2004 [43](i) Sampling frame: one city in Israel (not mentioned)  
(ii) Sampling method: not reported  
(iii) Recruitment method: not reported  
(iv) Administration method: not reported
(i) Sample size calculation: no  
(ii) Sampling type: not reported  
(iii) Validity of tool: not reported  
(iv) Pilot testing done: no  
(v) Response rate: not reported
(i) Population: Yemenite Jews (all > 30 years old, male, and parents born in Yemen)  
(ii) Country: Israel  
(iii) Setting: unnamed city  
(iv) (sampled) = 1500  
(v) (participated) = 1500  
(vi) (included in analysis) = 47
(i) Prevalence of khat, (%) (specify pattern of use): 102/1500 (6.8%) (all used khat > 3 years)  
(ii) Prevalence of tobacco use (specify tobacco type and pattern of use), (%): N/A  
(iii) Prevalence of tobacco among khat chewers (specific tobacco type and pattern of use), (%): 32/47 (68%) were cigarette smokers and smoked more (29.5/day) than nonchewers (22.3/day)
(i) Level of dependence among khat chewers who use tobacco: not reported  
(ii) Associated factors: not reported

Kebede, 2002 [44](i) Sampling frame: all university instructors in four northwestern colleges ()  
(ii) Sampling method: universal sampling  
(iii) Recruitment method: not reported  
(iv) Administration method: self-administered
(i) Sample size calculation: no  
(ii) Sampling type: nonprobability (broad sampling, covered all instructors in four colleges)  
(iii) Validity of tool: not reported  
(iv) Pilot testing done: yes  
(v) Response rate: 75.1%
(i) Population: university instructors, 93.9% male, mean age 35.2 (8.36) years, 73.5% Christian, 12.7% Protestant, and 91.7% Ethiopians  
(ii) Country: Ethiopia  
(iii) Setting: colleges, January 2001  
(iv) (sampled) = 241  
(v) (participated) = 181  
(vi) (included in analysis) = 181
(i) Prevalence of khat, (%) (specify pattern of use): lifetime khat use 59/181 (32.6%), current (past 30 days) use 38/181 (21.0%)  
(ii) Prevalence of tobacco use (specify tobacco type and pattern of use), (%): lifetime cigarette use 51/181 (28.2%), current (past 30 days) use 24/181 (13.3%)  
(iii) Prevalence of tobacco among khat chewers (specific tobacco type and pattern of use), (%): 34/59 (57.6%) of lifetime khat chewers were lifetime cigarette smokers, and 14/38 (36.8%) of current khat chewers were current cigarette smokers
(i) Level of dependence among khat chewers who use tobacco: not reported  
(ii) Associated factors: not reported

Lemma et al., 2012 [14](i) Sampling frame: all students enrolled at two universities ( = 29823)  
(ii) Sampling method: multistage sampling  
(iii) Recruitment method: not reported  
(iv) Administration method: self-administered
(i) Sample size calculation: no  
(ii) Sampling type: probability sampling  
(iii) Validity of tool: not reported  
(iv) Pilot testing done: no  
(v) Response rate: not reported
(i) Population: students, mean age 21.6 (1.7) years, and 77.3% male  
(ii) Country: Ethiopia  
(iii) Setting: universities  
(iv) (sampled) = not reported  
(v) (participated) = 2817  
(vi) (included in analysis) = 2230
(i) Prevalence of khat, (%) (specify pattern of use): 209/1960 (10.7%) (1-2 and ≥ 3 times per week)  
(ii) Prevalence of tobacco use (specify tobacco type and pattern of use), (%): 84/2230 (3.8%) ever cigarette smokers  
(iii) Prevalence of tobacco among khat chewers (specific tobacco type and pattern of use), (%): 28.7% of khat chewers were current cigarette smokers, 2.4% of khat chewers were former cigarette smokers
(i) Level of dependence among khat chewers who use tobacco: not reported  
(ii) Associated factors: not reported

Reda et al., 2012 [36](i) Sampling frame: all students enrolled in high schools in Harar town (capital city of one of the nine regions of Ethiopia) ( = 6523)  
(ii) Sampling method: stratified random sampling  
(iii) Recruitment method: not reported  
(iv) Administration method: Self-administered
(i) Sample size calculation: no  
(ii) Sampling type: probability sampling  
(iii) Validity of tool: not reported  
(iv) Pilot testing done: yes  
(v) Response rate: 91.1%
(i) Population: high school students, 50.1% male, mean age 16.4 (1.6) years, 48.9% 9th grade, 52.8% Orthodox, 32.2% Catholic, and 11.2% Muslim  
(ii) Country: Ethiopia  
(iii) Setting: high schools, April 2010  
(iv) (sampled) = 1890  
(v) (participated) = 1721  
(vi) (included in analysis) = 1721
(i) Prevalence of khat, (%) (specify pattern of use): 427 (24.2% 95% CI 22.2–26.2) had ever chewed khat, 89/1721 (5.2%) chewed khat daily  
(ii) Prevalence of tobacco use (specify tobacco type and pattern of use), (%): 4.2% were current (past 30 days) smokers, and 12.4% were ever smokers  
(iii) Prevalence of tobacco among khat chewers (specific tobacco type and pattern of use), (%): among ever chewers, 128/427 (29.9%) used waterpipe when they chewed khat
(i) Level of dependence among khat chewers who use tobacco: not reported  
(ii) Associated factors: not reported