Trend analysis of grading systems for level of evidence and strength of recommendation

When clinicians or healthcare professionals are to make decisions, they can judge the quality of evidence and reliability of recommendations by 'Level of evidence' and 'Grade of recommendation'. Because of this, the step of grading evidence and recommendations is very important in developing clinical practice guidelines. The objective of this study was to identify the various grading systems and criteria of the clinical practice guidelines. We reviewed 101 guidelines from the National Guideline Clearinghouse Database and chose 66 guidelines to analyze in terms of the grading systems for level of evidence and strength of recommendation. The grading systems for 'Level of evidence' were classified into 4 types by criteria such as study design, study quality, consistency, limitations, strength of evidence, and validity. Type II was the most common evidence grading system applied by 12 organizations (37.5%) and 30 guidelines (45.5%). The grading systems for 'Grade of recommendation' were classified into 4 types by criteria such as level/quality of evidence, strength of recommendations, study quality, consistency, applicability, balance between benefit and harm, and effectiveness/usefulness. Type I was the most common recommendation grading system applied by 9 organizations (33.4%) and 23 guidelines (40.4%). A formal grading system based on consistent and clear approaches is needed because the process of grading work can be subjective when clinical practice guideline users are making decisions. It is necessary for clinical practice guideline developers to have a common criterion so that they can judge the grade of evidence and recommendations objectively in the development of clinical practice guidelines.