Author (Year)	Study Type	Confounding	Selection of Participants	Classification of Interventions	Deviations from Intended Interventions	Missing Data	Measurement of Outcomes	Selection of Reported Result	Overall Judgment
Altintas et al. (2016) [27]	Single-center cross-sectional analysis	Serious—Confounding by psychiatric diagnosis and polysubstance use	Moderate—Participants were psychiatric patients, which limits representativeness	Low—Exposure classification via structured interview	Low—No deviations relevant	Moderate—Some missing records	Serious—Outcomes not blinded, limited validity	Moderate—Some selective emphasis	Serious risk of bias
Bassir et al. (2016) [29]	Retrospective review	Serious—Multiple confounders, no adjustment	Moderate—Hospitalized psychiatric patients only	Low—Exposure classification based on records	Low—No deviations relevant	Moderate—Some missing records	Serious—Outcomes not systematically validated	Moderate—Some selective emphasis	Serious risk of bias
Every-Palmer (2011) [37]	Cohort study	Serious—Multiple unmeasured confounders	Serious—Small forensic psychiatric sample, very selective	Low—Exposure classification consistent (self-report of JWH-018)	Moderate—No intervention deviations applicable	Moderate—Incomplete data from interviews	Serious—Outcomes subjective, not standardized	Serious—Results selectively described	Serious risk of bias
Glue et al. (2013) [39]	Retrospective observational study	Serious—Confounding by severity of illness	Moderate—All hospitalizations reviewed, small sample	Low—Exposure classification clear (SC identified)	Low—No deviations relevant	Moderate—Some missing documentation	Serious—Outcomes variable, not standardized	Moderate—Some selective reporting	Serious risk of bias
Hermanns-Clausen et al. (2018) [42]	Prospective observational study	Serious—Multiple confounding variables not controlled	Low—Consecutive ED patients included	Low—Exposure confirmed analytically	Low—No deviations, observational design	Moderate—Some missing clinical and lab data	Serious—Heterogeneous outcome measures across sites	Moderate—Selective reporting possible	Serious risk of bias
Hoyte et al. (2012) [43]	Observational study	Serious—Confounding by reporting patterns	Low—All NPDS cases included	Low—Exposure classification robust for poison center data	Low—Observational, no deviations relevant	Moderate—Some missing data in registry	Serious—Outcomes inconsistently categorized	Moderate—Registry limitations	Serious risk of bias
Kekelidze et al. (2019) [45]	Interventional study	Serious—No confounder adjustment	Moderate—Psychiatric hospital patients, limited representativeness	Low—Exposure classification by clinical record	Low—Observational, no deviations relevant	Moderate—Some incomplete records	Serious—Outcomes not validated	Moderate—Selective reporting likely	Serious risk of bias
Monte et al. (2017) [47]	Cohort study	Serious—Confounding by indication and severity, no comparator	Low—All cases from registry included consecutively	Low—Exposure classification robust (clinical registry)	Low—No deviations relevant, observational design	Moderate—Some incomplete clinical details	Serious—Outcomes heterogeneous across centers	Moderate—Registry structure limits selective reporting	Serious risk of bias
Skryabin et al. (2019) [55]	Observational study	Serious—No adjustment for confounding variables	Moderate—Participants were consecutive SC users but not population-based	Low—Exposure classification clear (self-report + clinical diagnosis)	Moderate—Deviations possible, treatment not standardized	Moderate—Some missing follow-up data	Serious—Outcomes not systematically validated	Moderate—Some selective reporting likely	Serious risk of bias
Skryabin et al. (2018) [56]	Longitudinal, observational cohort study	Serious—No confounder adjustment, polysubstance use not controlled	Moderate—Consecutive inpatients but not representative of general population	Low—Exposure classification clinical + confirmed SC use	Low—Observational, no intervention deviations	Moderate—Missing follow-up details for some patients	Serious—Psychiatric outcomes not systematically validated	Moderate—Potential selective emphasis in reporting	Serious risk of bias