Røgeberg & Elvik have published a critique and re‐analysis of previous meta‐analyses, as well as an updated meta‐analysis of studies that investigated the odds ratio (OR) for involvement in a road traffic crash (RTC) after using cannabis 1. They concluded that there was only a moderately increased OR of 1.22 or 1.36 for crash involvement during acute cannabis intoxication, depending on the method for calculation. There are, however, two significant problems with the updated meta‐analysis: first, most of the studies did not investigate the association between RTC and acute cannabis intoxication, but rather the association between RTC and the presence of traces of cannabinoids in samples of blood or urine indicating cannabis use during the last hours, days or weeks before sample collection; secondly, some of the included studies did not use the same or equivalent definitions of cannabis exposure among cases and control, creating significant bias in the calculated ORs. Cannabis exposure was determined by analysing biological samples in 20 of 21 studies included in the meta‐analysis (the authors disaggregated two studies into four and six separate estimates; we have listed those as two studies in Table 1). Analysing biological samples is, in general, an objective method that confirms cannabis exposure and may be regarded as less prone to errors than self‐reported use; however, incorrect classification of cannabis exposure may still occur due to variations in tetrahydrocannabinol (THC) concentrations and type of biological sample analysed and thereby also detection periods after cannabis smoking ... Keywords: Cannabis impaired driving; Cannabis; case-control; culpability; driving; driving under the influence; DUI; impairment
Language: en
", language="en", issn="0965-2140", doi="10.1111/add.13435", url="http://dx.doi.org/10.1111/add.13435" }