0ur reviews impact of #TransitionInterventions in #GenderDysphoria: uncertainty remains. Used by National Post to justify Alberta government forbidding interventions. NP didn’t publish my letter: our work informs appropriate autonomous judgements by patients. Low integrity.

#BMJ article documented disastrous performance of #RCT authors making inappropriate strong claims for #SubgroupEffects when application of appropriate criteria would have concluded low credibility. Fortunately, the #ICEMAN instrument now provides definitive guidance.

#RealWorldData advocates facing challenges addressing sources of bias suggest causal modelling as a solution. Doesn’t work. Comparison of 19 modelling studies with #RCTs showed 42% differed in direction and 47% of confidence intervals didn’t include RCT estimate.

Randomized trials in individual patients – N of 1 #RCTs - remain underutilized. This @AnnalsofIM paper describes our initial experience with 57 N of 1 RCTs. In 15, results prompted patients and clinicians to alter their treatment plans.

For clarity, simplicity, accessibility and user-friendliness the original #JAMA users guides – including this one addressing harm – remain the best for clinicians seeking guidance in how to interpret studies to guide their patient care.

In a previous post I noted that Guidelines International Network (GIN), in rejecting a workshop on Core GRADE, were underestimating the likely interest in the new #BMJ Core #GRADE series. We were thinking of conducting the workshop in an alternative venue but it has proved.

#GRADE evidence to decision (EtD) frameworks present a useful, transparent and comprehensive approach to moving from evidence to recommendations. This #BMJ paper focuses on EtD frameworks for clinical practice #guidelines.

Advocates suggest administrative real-world data can substitute for #RCTs and offer propensity matching to reduce bias. But results often fail to correspond: #BMJ study showed direction of effect differed in 31% & CI failed to include #RCT estimate in 56%.

Discontinued randomized control trials (#RCTs) a problem. This #JAMA article presents the prevalence (almost 25%), characteristics (most frequently due to poor recruitment), & publication status (odds ratio>3 of not being published) of discontinued RCTs.

On recent trip to Helsinki, biking in the nearby forest with @KariTikkinen leading urological and surgical investigator and medicine overuse expert; @GSCollins #BMJ stats editor; and @LauriLavikainen recent PhD grad who created AI Gordon.

Two fundamental strategies for establishing #MinimalImportantDifferences (#MIDs) in #PatientReportedOutcomes (#PROMs). The more popular is looking at changes within person over time. This paper provides an alternative: looking at differences between persons.

PhD candidate created “AI Gordon”, sent his thesis dominated by low certainty evidence, asked for advice how to defend. Presumably, LLM reviewed my writings, interviews, videos etc. Provided extraordinary response that even captured my voice. Wow, I’ve just achieved immortality!.

This #BMJ paper provides a consensus view and a checklist on when and how to update #SystematicReviews (#SRs) .#EBM #Medicine .

Participants and tutors at our recent Core #GRADE workshop, offering education on optimally using guidance presented in the 7 part #BMJ series published April to June this year. We are anticipating this #workshop as a yearly event.

#CompositeEndpoints in #RCTs: big problem. #BMJ survey of use in #cardiovascular trials: they make results look more important than they are through impression of benefits in death when only real effect in much less important outcomes (e.g. stent placement).

Just published commentary shows how expanding definitions of diseases such as #hypertension, gestational #diabetes, #osteoporosis, attention-deficit/hyperactivity disorder & social #anxiety disorder can result in harmful #medicalization & inefficient resource use.

Incorporating cost/resource use into clinical practice #guidelines is tricky. This #BMJ paper designed for clinician users of #GRADE guidelines outlines GRADE’s approach to the problem.

Do technological advances require rigorous testing as with drugs? Or should we continue to give them a pass? Just published study showing increased bile duct injury with #robotic versus #laparoscopic #cholecystectomy the latest evidence we should get more stringent.

Brian Haynes, PJ Devereux and I present the rationale behind the third key principle of #EBM: evidence alone never is sufficient for decisions, #patient values and preferences always crucial. Or as this #BMJ article put it, evidence doesn’t make decisions, people do.

You are trying to help someone who is ambivalent between mutually exclusive actions A and B. You think A clearly preferable and say so. This is the best way to drive them toward B. To help: listen, empathize and help clarify relative merits. They will then find their way.