Table 6.

Studies related to the legalization of recreational and medicinal cannabis and its decriminalization: design, main results, and limitations.

Article	Statistical Analysis	Main Results	Limitations
Benedetti et al. (2021) [42]	Multiple logistic regression model.	States with MCL have a higher number of drivers who have driven under the influence of marijuana versus states that have not legalized MC and/or RC (OR 1.29; 95% CI 0.98, 1.70; p = 0.075). THC threshold laws: less likely to drive after consumption (OR 0.74; 95% CI 0.57, 0.95; p = 0.018).	Biases associated with self-reports. Quasi-experimental design that does not allow inferring causal relationships between marijuana use among drivers and states’ policies on marijuana use.
Kruse et al. (2021) [44]	Retrospective analysis of data. Percentages.	There seems to be no relationship between legalization and the probability of finding THC in patients admitted after an accident.	Discrepancies in urine THC detection limits by institution. A lack of standardized laws by the state does not allow the detection of real THC prevalence.
Woo et al. (2019) [8]	Series of logistic regressions.	Being a young man, driving a motorcycle, and testing positive for alcohol, delta 9-THC, or carboxy-THC and other drugs (p < 0.001) are risk factors for speeding. Cannabis predicts risky driving behavior.	Only fatal accidents are examined. Washington State data only. Not all crashes tested for drugs. Measurement errors in drug rates.
Keric et al. (2018) [45]	Time frame. Percentages.	A total of 90% of surgeons report no increase in cases of traumatic injuries in traffic accidents after cannabis legalization.	Not specified.
Lee et al. (2018) [43]	Series estimation of crash modification factors.	Increase in fatal accidents in which the driver tests positive for cannabis, mainly with decriminalization (p < 0.001) and/or legalization of RC but not MC (p < 0.001) Other effects are between decriminalization and decriminalization and MCL (p = 0.020) and between MCL and full legalization (p = 0.010).	Short post-legalization periods. Differences between states in drug testing protocols and trends after legalization. Difficulties in selecting a control group. Cannot assert causality.
Hamzeie et al. (2017) [46]	Logistic regression models	Higher probability of testing positive for THC in an accident in states with cannabis decriminalization (17%) and/or legalization laws (48%) (p < 0.001). Being young, male, positive for alcohol, and exhibiting more risky driving behaviors increased the probability of THC+ (p < 0.001).	They only test for CRL in two states and for a short period of time. Not all drivers take the drug test. Differences between states in drug testing protocols and trends after legalization.
Pollini et al. (2015) [47]	Multiple logistic regression analyses	Significant increase in the prevalence of cannabis positives among drivers involved in fatal crashes after decriminalization (17.8%; 95% CI: 14.6, 20.9). No change in THC positives among weekend nighttime drivers after decriminalization (9.2%; 95% CI: 6.3, 12.2).	Differences in drug testing protocols. Changes in consumption trends after legalization. Small and restricted sample. THC+ does not imply recent use.
Couper and Peterson (2014) [48]	Chi-squared tests	After a stable trend, there is a significant increase in the percentage of positive cases of THC consumption in drivers after legalization (p < 0.05).	THC concentration can be altered causing problems in the cut-off point for considering a subject positive. Delays in blood collection can influence the concentration of metabolites.