Medical decision analysis can be challenging to understand and apply in clinical practice or systems decision-making. This recent Users’ Guide does a great job of explaining decision analysis and guiding its practical use. pubmed.ncbi.nlm.nih.gov/34226025/

Serving grad students most efficiently. Instructions about booking appointments. 'Please go straight to my assistant and choose one: a) I need Gordon’s help asap, within a couple of days. b) I need Gordon’s help soon please, within a week. c) I need Gordon’s help, no rush.'

Considering patient and public values and preferences is a crucial part of every clinical practice guideline.  Guideline panels should search relevant evidence. GRADE summary of findings tables enhanced understanding of values and preferences evidence. jclinepi.com/article/S0895-…

New editorial exposes big problems with the clin epi/EBM methods community losing our bearings in balancing simplicity vs methods sophistication. Excessive complexity includes Cochrane RoB 2, ROBINS-I, and GRADE methods. Solutions are on the way! pubmed.ncbi.nlm.nih.gov/38453419/

Recently published step by step approach to choosing the optimal MID when literature includes multiple published estimates for the same instrument. doi.org/10.1136/bmj-20…

Two articles highlight big problems when platform trials include non-contemporaneous controls. In addressing A vs no A, analysis should only use randomizations that included possibility of A. Should never use controls of B vs non-B.
nejm.org/doi/full/10.10…
sciencedirect.com/science/articl…

Latest evidence of utter failure of non-inferiority trials to set appropriate non-inferiority margins. Clinicians should ignore these margins & statistical claims of non-inferiority & instead think how their patients would balance benefits/harms/burdens. pubmed.ncbi.nlm.nih.gov/37421996/

Just published survey of oncology trials demonstrates seriously flawed subgroup inferences: doi.org/10.1001/jamane…
Enlightened commentary highlights failure to consider prior knowledge, not reporting key interaction statistics, and ignoring multiplicity. doi.org/10.1001/jamane…

Recent new concepts in assessing certainty of evidence on prognostic models' calibration performance: need to decide whether rating certainty that model's performance is satisfactory or unsatisfactory; inconsistency critical GRADE domain in this decision. jclinepi.com/article/S0895-…

After revisions for regulators, we're just about to start recruitment for large, pragmatic #ARTStrial that recruits uro, gyne & abdominal surgery patients in which net benefit of pharmacologic prophylaxis uncertain. If interested, please comment or message #EBM #EAU24 European Association of Urology (EAU)

Public ignorance about making wise health choices a catastrophe (COVID!). Solution: school-age education. International group doing amazing work. Latest teaches teenagers to think critically about health choices. More including April 25th webinar on blog. 3ieimpact.org/blogs/teaching…

A recent innovative guideline addressing chronic temporomandibular pain demonstrates how to deal with a large number of competing interventions in a network meta-analysis and use GRADE to generate trustworthy recommendations. doi.org/10.1136/bmj-20…

Four mentorship archetypes. Traditional (formal reciprocal relationship in work setting); coach (improves specific skill); sponsor (uses influence to increase mentee visibility, recommends mentee); connector (pair mentors, coaches, & sponsors with mentees) doi.org/10.1001/jamain…

The standard gamble is beloved by health economists. Its measurement properties are, sadly, consistently inferior to the feeling thermometer as this application of generalizability theory shows. pubmed.ncbi.nlm.nih.gov/17998080/

Two infinite human capacities. One, the capacity for self-deception. Sometimes helpful, sometimes disaster. Second, the capacity to feel hard done by. Never helpful, always self-defeating and destructive.

For teachers of evidence-based medicine: how to clearly & simply explain adjusted/stratified analysis so that learners understand the fundamental principle of how it’s done. That is, create prognostically homogeneous groups & then pool across those groups. doi.org/10.1007/s11606…

Teaching EBM in Buenos Aires - Thursday, March 14, 2024.

Previously established that relative effects almost always similar constant across high and low risk subgroups. Now compelling new evidence the same is true for comorbidity: relative effects constant across high and low comorbidity subgroups. doi.org/10.1371/journa…

Meta-epidemiological studies have repeatedly demonstrated the superior responsiveness of disease-specific to generic instruments. jclinepi.com/article/S0895-…

For folks using other platforms, you will now find my posts on both Threads (threads.net/@guyattgh) and LinkedIn (linkedin.com/in/guyattgh/).