Lecture: Levels of Evidence and Research Design for Evidence-Based Practice

Flashcards covering major concepts from the lecture notes on levels of evidence, bias, research design, NHMRC, and evidence-based practice.

What are the major clinical questions in healthcare?

Clinicians’ daily questions answered by research evidence; not all evidence is equally strong; bias threatens validity; higher levels of evidence correspond to lower risk of bias.

What is bias in medical research?

A systematic error in results that leads to false conclusions and reduces trustworthiness (validity) of the evidence.

How do levels of evidence relate to bias?

The higher the level of evidence, the lower the risk of bias in the study findings.

List the NHMRC-referenced hierarchy of evidence from strongest to weakest.

Systematic reviews; Randomized Controlled Trials; Cohort Studies; Case-Control Studies; Case Series/Case Reports; Editorials/Expert Opinion.

What does NHMRC stand for and what is its role?

National Health and Medical Research Council; funds high-quality health research and guides evaluation of quality and relevance; defines levels of evidence and grades of recommendations.

What is a clinical question in research terms?

A clearly defined question answered by research using a specific design and data from real people.

What is Level I evidence?

High-quality evidence, typically a systematic review of Level II studies or high-quality RCTs, with low risk of bias.

What is Level II evidence?

Randomized Controlled Trials; single original studies with a treatment and a control/placebo; randomization minimizes bias.

What is Level III-1 evidence?

Non-random allocation to groups; fully random method was not used, increasing allocation bias.

What is Level III-2 evidence?

Non-random allocation with higher allocation bias than III-1.

What is Level III-3 evidence?

Treatment and control are not randomly allocated; outcomes are not evaluated concurrently; higher bias.

What is Level IV evidence?

Case series; single-arm studies with pre/post testing; no control; cannot definitively attribute outcomes to treatment.

What are clinical practice guidelines?

Documents that provide evidence-based recommendations for managing a specific condition and assist clinical decision making; NHMRC uses GRADE to evaluate evidence quality and recommendation strength.

What is the GRADE system?

A framework to evaluate the quality of evidence and strength of recommendations: strong, weak, or conditional/discretionary.

What should be considered when evaluating the quality of evidence?

Strengths/limitations of design, level of evidence, consistency of results, treatment effects, bias, and relevance to the population.

What biases may affect intervention studies?

Allocation bias, detection bias, performance bias, attrition bias, and measurement bias.

What biases may affect systematic reviews?

Publication bias, time lag bias, duplicated publication bias, outcome reporting bias, citation bias, database inclusion bias, language bias, and reviewers’ personal biases.

What are the two aspects of research NHMRC emphasizes?

A clearly defined clinical question and a well-defined level of evidence (minimizing bias and maximizing trustworthiness).

How does NHMRC categorize evidence levels?

Level I – high quality (systematic reviews of Level II or high-quality RCTs); Level II – RCTs; Level III – non-randomised; Level IV – case series.

How are clinical questions matched to research designs and NHMRC levels of evidence?

Each question has a study design (intervention, diagnostic, aetiology, prognostic, screening); Level indicates risk of bias; Level I is the strongest.

What is primary vs secondary evidence?

Primary evidence comes from original studies (Levels II–IV); secondary evidence comes from systematic reviews (Level I) or reviews of Level II studies; meta-analysis is used in some reviews.

What databases are used to access best evidence?

CINAHL, MEDLINE, Cochrane Library, PsycINFO.

What is meta-analysis in systematic reviews?

Statistical combination of results from multiple quantitative studies to form a single pooled estimate.

What is a diagnostic accuracy study?

Measures how well an assessment identifies people with a health condition and separates them from those without.

What is an aetiological (etiology) study?

Epidemiological research measuring associations between exposure to a suspected hazard and rates of a health condition.

What is a prognostic study?

Records what happens to people with a specified health condition, especially when untreated.

What is a screening intervention study?

Tests a large number of people without signs to determine whether follow-up procedures lead to a benefit or harm.