Research library
Start with the question.
Follow the evidence.
Seven reading paths, selected literature-based briefs, and practical review materials. The library connects the four sites without reproducing their articles.
From the supplied research collection
Focused reading, with boundaries.
THC potency and driving performance
A controlled trial with positive impairment findings, a negative primary outcome, and important limits on individual interpretation.
Read the THC research briefTwo blood samples and alcohol elimination
Study design, apparent decline, sampling-time reliability, and why a short interval does not settle an earlier BAC.
Read the research briefCBD, THC, and derivatization
A method-specific reading path about TFAA-HFIP sample preparation. It is not a claim about every cannabinoid test.
Read the methods article on .comAge and balance evidence
A population-balance study and the limits of extending its findings to a roadside observation or individual impairment.
Read the balance-evidence articleFor a worked example involving an individual prediction interval rather than a simple average conversion, see the vitreous-ethanol guide and its discussion of Pounder and Kuroda (1994).
01 / Research reading path
What did the laboratory actually measure?
Start with the procedure, not the conclusion. Use these guides to frame questions about preparation, separation, identification, and the role of an internal standard.
02 / Research reading path
Which specimen produced the result?
Organize collection, storage, and specimen-validity records before comparing results. These introductions address the sample itself and the records needed to understand it.
03 / Research reading path
How was alcohol measured or estimated?
Separate a measured blood or breath result from a later calculation about an earlier time. Start with the analytical method, then examine the assumptions used for interpretation.
04 / Research reading path
Was the finding screened, confirmed, or quantified?
Choose the resource that matches the stage of testing. Do not use a general drug-testing overview as a substitute for the laboratory’s actual procedure and reported analytes.
05 / Research reading path
What supports the reliability of the number?
Request the performance evidence behind the method. These guides introduce calibration, reporting limits, contamination controls, and external proficiency testing.
06 / Research reading path
Does the proposed inference fit the evidence?
Frame the question precisely: detection, timing, pharmacokinetics, or a clinical comparison. Follow the source references and limitations on the destination page.
Study appraisal / Orriols et al., 2019
Does a prescription record establish driving impairment?
A dispensing record is not a measurement of driving performance. Orriols and colleagues linked French prescription reimbursements, police reports and injury-crash records from July 2005 through December 2015. Their study estimated medication exposure; it did not measure benzodiazepine blood concentrations or determine impairment in each driver.
What was inferred? Exposure began the day after dispensing, with duration estimated from a prescription survey. Medication dispensed on the crash date was excluded to reduce the chance of counting treatment supplied because of the crash. These rules model exposure. They do not establish whether, when or how much medication a person actually took.
Who was compared? A standardized crash-circumstance score classified 97,936 drivers as responsible and 103,522 as non-responsible. Both groups had been involved in injury crashes. The central responsibility analysis was not a comparison with drivers who completed a trip without crashing. A separate general-population sample was used for exposure trends, and its members’ driving status was unknown.
What did the estimates show? Table 3 reports an adjusted odds ratio for benzodiazepine anxiolytics of 1.42 (95% confidence interval 1.24-1.62) in the first study period and 1.27 (1.09-1.47) in the last. These are population associations under the study’s exposure and responsibility definitions. An odds ratio of 1.42 is not a 42% probability that one driver caused a crash.
What limits the inference? Only drivers with linkable records entered the study, and more seriously injured drivers were overrepresented. Actual consumption was unknown, and some drug-use and behavioral information was unavailable. A driver classified as non-responsible could still have been impaired and less able to avoid a collision. Adjustment does not guarantee that every relevant difference between groups has been removed.
The comparisons around France’s medication-warning pictograms were observational, not randomized tests of a label’s effect. The paper provides neither an individual impairment concentration cutoff nor a safe-driving waiting period. Keep dispensing, actual use, impairment and event causation separate.
Read the source: Orriols L, Gbaguidi GN, Contrand B, Gadegbeku B, Lagarde E. French benzodiazepine and z-hypnotic crash-responsibility study. Injury Epidemiology. 2019;6:32. DOI: 10.1186/s40621-019-0209-8.
Review disclosure: AI-assisted appraisal of the complete 11-page paper, including methods, results, limitations, references, figures and tables. No separate supplement was reviewed. This appraisal has not been personally reviewed by Okorie Okorocha. It is historical research context, not current prescribing guidance or a case-specific opinion.
07 / Research reading path
Is this a postmortem interpretation?
Use a death-investigation reading path rather than silently importing assumptions from a living subject. The specimen source and the question being answered belong at the start of the review.
Prepare a review, not just a reading list.
- Case-review checklist
- Laboratory-packet review guide
- Scientific foundations on .com
- Complete evidence library on .net
This is a selected, growing library, not a claim that every supplied paper has been reviewed. A link identifies related reading; it does not certify every statement on the destination page. Read the original publication and the destination’s authorship and disclaimer. Our source and editorial policy.