I’ve been teaching evidence-based medicine (EBM) at the University of Minnesota since 2004. This page contains EBM resources for our Family Medicine Clerkship students. Those visiting from outside the University of Minnesota should use the PubMed and Google Books links, where provided
Data Interpretation
Step 1: Is the difference real? (no major flaws, statistically significant)
Step 2: Does the difference matter to patients? (patient-oriented outcome: POO)
— B —
Bias, Catalogue of [Centre for Evidence-Based Medicine]
— C —
Centre for Evidence-Based Medicine
Clinical Versus Statistical Prediction [textbook]
ClinicalTrials.gov [trials registry]
Cochrane Database of Systematic Reviews [discovery tool]
— D —
Diagnosis [PowerPoint]
Diagnosis: What Doctors Are Missing [essay]
Diagnostic Error
Disease-Oriented Outcomes (DOO) [see also POO]
DynaMed [vs. UpToDate — see Figure 1]
— E —
Effect size [see Cohen’s d]
Effect size [calculator]
Evidence-based medicine, alternatives to
Evidence-Based Medicine Toolkit
Evidence-Based Physical Diagnosis [textbook] (Google Books)
— F —
Funnel plot [publication bias] (Brief Explainer)
— G —
Generally Recognized as Safe (GRAS)
Google Scholar [discovery tool]
GRADE system
Grade recommendations (UpToDate — these are based on the GRADE system)
Grey literature (PubMed)
— H —
Health Sciences Library (University of Minnesota)
Health statistics, understand of [patient resource]
Heterogeneity
How Doctors Think [book]
How Doctors Think [essay]
— I —
I2 (I-squared)
Index Medicus [old-school discovery resource]
Intention-to-treat analyses (PubMed)
— J —
Jadad scale (PubMed)
— L —
Level of Evidence (LOE) [required taxonomy]
— M —
Medically necessary (medical necessity)
Minimal clinically important difference (MCID) (PubMed)
Motivated reasoning
— N —
Narrative review, vs. systematic review [narrative review = literature review]
National Comprehensive Cancer Network (NCCN — requires free registration)
Negative Predictive Value (NPV; a/k/a Negative Predictive Power [NPP])
Network meta-analyses [see Figure 4] [and also here and here]
Non-inferiority studies [see Figure 1] (PubMed)
Number needed to treat (NNT) [calculator]
— O —
Observational studies (PubMed)
— P —
P-values (PubMed) (and Table 1 here)
Patient-Oriented Evidence that Matters (POEM)
Patient-Oriented Outcome (POO) [see also DOO]
Per-protocol analyses (PubMed)
Placebo effect, likely overestimated
Positive Predictive Value (PPV; a/k/a Positive Predictive Power [PPP])
Power, a priori [calculator]
Power, post-hoc [calculator]
PROSPERO [systematic reviews registry]
Publication bias: see Funnel plot
PubMed Clinical Queries [discovery tool]
— R —
Rational Clinical Examination [textbook] (Google Books)
Reliable evidence
Reproducibility [title is ironic in light of the author’s struggles]
— S —
Scopus [discovery tool]
Statistics Notes [long-running feature in BMJ — highly recommend!]
Strength of Recommendation Taxonomy (SORT) [required taxonomy] (and Figure 1 here)
— T —
Test validation summary (PubMed) [PPV and NPV for any base rate]
Therapy [PowerPoint]
Trip Database [discovery tool]
Type I and Type II errors (PubMed) (Wikipedia)
— U —
United States Preventative Services Task Force (USPSTF)
UpToDate [vs. DynaMed — see Figure 1]
Users’ Guides to the Medical Literature [textbook] (Google Books)
— W —
Web of Science [discovery tool]