Week 7: You won't believe it (more stuff)

PROM categories

Patient-reported outcome measures

• Global/generic

• Body part-specific/regional

• Diagnosis-specific

Clinical Prediction Rule categories

• Diagnostic

• Prognostic

• Treatment efficacy

Prisma Flow Diagram

Like a consort flow chart in a single study that tracks status of all participants from the beginning, a prisma flow diagram tracks the process of article review, retainment, and/or exclusion.

I² Statistic

Describes the variation across studies in a meta-analysis/systematic review due to heterogeneity (true difference) as opposed to chance. Chance difference is OK, true difference is not. Want I² to be low, implying that more difference is just chance.

>75% is very high and bad

Funnel Plot

Helps detect publication bias or small-study error.

• X-axis: Effect size estimate (NOT usually centered around 0)

• Y-axis: Measure of precision (high standard error is low precision) (AKA larger studies are higher, smaller studies are lower)

Want the funnel to be symmetrical. Asymmetry implies publication bias or heterogeneity. You can infer from the empty space in a funnel that there may be some missing studies there.

PROM, performance measure, clinician-reported measure, biological measure

Patient-reported outcome measure: Patient’s perceived function or symptoms

Performance measure: Patient ability under standardized, objective conditions

Clinician-reported measure: Like performance measure, but interpreted by the clinician and not objective measurement (aka MMTs)

Biological measure: Physiological or structural findings (A1Cs, imaging, VO2max) that reports physiologic function rather than performance results

Efficacy vs Effectiveness

Efficacy: Reliability under ideal conditions

Effectiveness: Reliability in clinical practice

PEDro

evaluates primary RCTs