Evidence pyramid
The evidence-based pyramid provides a visual depiction of forms of research from least reliable (base of pyramid) to most reliable (top of pyramid). Ingham-Broomfield, R. (2016). A nurses' guide to the hierarchy of research designs and evidence. The Australian Journal of Advanced Nursing, 33 (3) pp 38-43.
Figure 1: Level of evidence. From "What is hierarchy of evidence" by Dr Arun Pal Singh, 2015,
https://boneandspine.com/what-is-hierarchy-of-evidence/
| Level 1a/1b Apex | Systematic Reviews and Meta-analysis |
The most reliable of all. Systematic reviews and Meta analyses of primary research into human health care and health policy are recognised internationally as the highest standard in evidence-based care. |
| Level 2 | Critically Appraised Topics (Evidence Syntheses) | Very high reliability. Synthesising research publications entails the categorising of a series of related studies, analysing and interpreting their findings and then summarising those findings in to unified statements. The potential lack of standardisation can undermine the validity. |
| Level 3 | Critically-Appraised Individual Articles (Article Synopses) | Increasing reliability of findings. A synopses is the evidence of an individual article with an expert telling you its strengths. This is less reliable than Critically Appraised Topics as there is less evidence on single articles than in a synthesis of the topic using several papers. |
| Level 4 Middle | Randomized Controlled Trials | Very Reliable/Gold Standard. Randomized Controlled Trials eliminate bias in quantifying the efficacy of an intervention. They are the highest level of evidence for Primary Research. |
| Level 5 | Cohort Studies | Becoming more reliable. Observational studies are good at answering questions about prognosis, diagnosis, frequency and aetiology but not questions regarding the effect of an intervention |
| Level 6 | Case Series and Case Reports | Slightly more reliable but there is a potential for bias in recalling information and the quality may be affected if the information is collected retrospectively |
| Level 7 Base | Ideas, Opinions, Editorials, Anecdotes | Least reliable. Basically anecdotal. Unscientific reports and observations |
Source: Ingham-Broomfield, R. (2016). A nurses' guide to the hierarchy of research designs and evidence. The Australian Journal of Advanced Nursing, 33 (3) pp 38-43.
The Centre for Evidence-Based Medicine is an excellent resource for further definitions and a glossary of key terms.
PICO table
The PICO framework is frequently used, in the health professions, to create a strategy for systematically searching databases for evidence.
