Evidence-based Decision Making: Introduction and Formulating Good Clinical Questions
Course Number: 311
Levels of Evidence
Sources regarded as strong evidence include clinical practice guidelines, systematic reviews with meta-analyses, systematic reviews alone, individual randomized controlled trials (RCT), and well-designed non-randomized control studies (Figure 2). The hierarchy of evidence for treatment questions is based on the notion of causation and the need to control bias.13
Figure 2. Hierarchy of Study Types and Levels of Clinical Evidence.
2016 Modified Evidence Pyramid. Copyright permission granted by SUNY Downstate Medical Center, Medical Research Library at Brooklyn
Although each level may contribute to the total body of knowledge, "...not all levels are equally useful for making patient care decisions."13 As you progress up the pyramid, the number of studies and correspondingly, the amount of available literature decreases, while at the same time their relevance to answering clinical questions increases (Figure 3).13
Figure 3. Available Literature and its Relevance.
Knowing which segment of the literature is appropriate for clinical decision-making and how to quickly retrieve this information is important to evidence-based practice. For example, the study methodology and level of evidence will differ based on the type of question asked, such as those derived from issues of therapy/prevention, diagnosis, etiology, and prognosis. Table 2 reviews the type of question and the highest levels of evidence-based on the study methodology. For example, for questions associated with therapy and prevention, the highest level of evidence will be from meta-analyses or systematic reviews of randomized controlled trials (RCTs), since the objective of these studies is to test interventions demonstrating cause and effect and to select treatments that improve the condition/disease and avoid adverse events.5
Correctly identifying the type of study to answer the question is an important skill to develop to access the appropriate evidence when searching the healthcare literature. For example, identifying the best implant technique for replacing a single maxillary molar is a treatment question. Ideally, a meta-analysis or systematic review of RCTs would be available on the treatment being considered. If one were not available, then the next best evidence would be from a well-conducted individual RCT. However, when the focus of the question is on long-term outcomes of treatment, then it is a question of prognosis where the highest level of evidence would be provided by a systematic review of inception cohort studies, which are studies that follow patients from when a disease or condition first manifests itself clinically. And again, if a meta-analysis or systematic review were not available, the next highest level would be an individual inception cohort study, and so on down the hierarchy (Table 2). Two important concepts to keep in mind are that: 1) for any type of question, having a well-conducted meta-analysis or systematic review typically provides stronger evidence than a single study, and 2) a meta-analysis or systematic review is only as good as the individual studies that comprise it.
Table 2. Type of Question Related to Levels of Evidence and Study Methodology.
| Type of Question | Type of Study or Methodology of Choice14 | Question Focus15,16 | Why Study?15,16 |
|---|---|---|---|
| Therapy/ Prevention |
Meta-analysis [MA] or Systematic Review [SR] of RCT’s Single Randomized Controlled Trial SR of Cohort Studies | Study effect of therapy or test on real patients; allows for comparison between intervention group and control groups for a particular condition. Largest volume of EB Literature. | To select treatments, if any, that do more good than harm (improve function, avoid adverse events) that are worth the effort and cost. |
| Diagnosis | MA or SR of Controlled Trials (Prospective cohort study) Single Controlled Trial (Prospective-compare tests with a reference or “gold” standard test.) | Measures reliability of a particular diagnostic measure/test for a disease against the “gold standard” diagnostic measure for the same disease. Sensitivity and specificity of the measures are compared. | To select and interpret diagnostic methods or tests. To determine the degree to which a test is reliable and useful; establish the power of an intervention to differentiate between those with and without a target condition or disease. |
| Etiology, Causation, Harm | MA or SR of RCTs Single RCT SR/MA of Cohort Studies Single Cohort Study (Prospective data collection with formal control group.) | Compares a group exposed to a particular agent with an unexposed group. Important for understanding prevention and control of disease. | To identify causes of a disease or condition including iatrogenic forms. To determine relationships between risk factors, potentially harmful agents, and possible causes of a disease or condition. |
| Prognosis | MA or SR of Inception Cohort Studies (Follow patients from when disease 1st becomes clinically manifest) Cohort Study | Follows progression of a group with particular disease and compares with a group without the disease. Groups must be as similar as possible and must have good follow-up>80% of each group. | To estimate clinical course or progression of a disease or condition over time and anticipate likely complications (and prevent them). |
An excellent website that graphically displays the different types of research methods and designs can be found at the SUNY Downstate Medical Research Library of Brooklyn: Evidence-Based Medicine Course.