Grading systems often consider the type of study, the strength of the study, the size of the effect, and the cost. They enable readers to immediately assess the value of the evidence and make key decisions on choice of test or treatment and funding priorities, or on fitness for work.
An internationally recognised system, the Grading of Recommendations Assessment, Development and Evaluation forms the basis of many approaches to policy development. The GRADE system specifically assesses:
GRADE scores A high, B moderate, C low or D very low. It forms the basis of the English NICE and Scottish SIGN systems of evaluation.
|Code||Quality of evidence||Definition|
|A||High||Further research very unlikely to change confidence in result|
|B||Moderate||Further research likely to have important impact on confidence, may change estimate of effect|
|C||Low||Further research very likely to have important impact on confidence in estimate of effect, may change estimate of effect|
|D||Very low||Any estimate of effect is uncertain (includes expert opinion, no direct research evidence)|
The Scottish Intercollegiate Guidelines Network (SIGN) grades levels of evidence using both digits and + or -. 1++ is the highest quality of evidence through 1+, 1-, then 2++, 2+ and 2-, 3 and 4. Quality of recommendations is graded A to D depending on the quality of evidence available.
The National Institute for Health and Care Excellence (NICE) follows a complex staged process starting with scoping to develop review questions, a systematic literature search and call for evidence, a formal evidence review with economic analysis, stakeholder comments with revision and final guideline production. The nature of each study is determined and an appraisal checklist used to assess quality. The GRADE system is then used to categorise recommendations.
The Royal College of General Practitioners (RCGP) modified three star grading system is as follows:
|***||strong evidence||Generally consistent findings in multiple, high quality scientific studies|
|**||moderate evidence||Generally consistent findings in fewer, smaller or lower quality scientific studies|
|*||limited or contradictory evidence||One scientific study or inconsistent findings in multiple scientific studies|
|-||no scientific evidence||Clinical studies, theoretical considerations or consensus|