Levels of Evidence and Research Designs

Longitudinal Studies

• follow subjects through time; measurements at prescribed intervals

Prospective Studies

a longitudinal study that looks forward in time

Retrospective Studies

a longitudinal study looking backward in time

Cross-Sectional Studies

• a “snapshot” of a population/ one point in time

Expert Opinion

Scientific views or comments by a group of designated experts (clinical and/or researcher) based on a review of scientific evidence and/or expert opinion.

ex. Delphi Consensus Study, Narrative Review, Textbooks, Editorial, Discussions, Letters to the Editor

Animal Studies

Animal testing has played an important role in the development of medicine:

• insulin

• anti-cancer drugs

• modern anesthetics

• tetanus and other vaccines

In-Vitro Studies

Allow scientists to isolate specific cells, bacteria, and viruses and study them without the distractions of having to look at a whole organism.

• HIV treatment drugs

  • culture of infected cells

  • treatment

Case Report

• tell the story of an individual pt

• rare but important phenomena

• N=1 so difficult to apply to another pt

Case Series

• several case studies put together to illustrate a single point

• N>1 but usually <20

• no group statistical analysis

Case Control Studies

• an observational study

• persons with a condition (cases) are compared to similar (matched) persons

• cross-sectional or retrospective

• best evidence for etiology of rare diseases or conditions with long latency

  • medical history of persons with a rare disease is compared to the history of similar persons without the disease

Cohort Studies

• one group of people or multiple groups that are predetermined or not randomly assigned

• monitored prospectively or retrospectively to assess development of disease or other outcome

• prospective cohort is the strongest way to study prognosis and etiology

Randomized Controlled Trials

• group of similar people randomly assigned to different treatment groups

• most effective way to test the effect of an intervention (treatment)

• allow for examination of cause and effect

• prospective

Systematic Reviews

• combine many studies into one BIG study

  • no new subjects - compiling the work of others

• use systematic methods for gathering, appraising, and combining different articles

• meta-analysis - when data of the studies are compiled and analyzed as one big study

  • statistical technique for combining the results of independent RCTs reported in systematic reviews. Provides a quantitative estimate of the net of effectiveness of interventions aggregated overall studies

Clinical Practice Guidelines

• a systematically developed summary of research

• includes recommendations for clinical practice

• recommendations are usually given grades

  • A - there is strong evidence to support

  • B - there is moderate evidence to support

  • C - there is weak evidence to support

  • D - there is conflicting evidence

• Not given a level because they are a combination of all types of research

• if well done, can be the most useful form of evidence