Table 2. Four levels of evidence.

Recommendation level/description	Assessment of benefits and risks	Methodological quality of the supporting evidences	Note
1A: Strongly recommended, based on high-quality evidences	Remarkably higher benefits than harms	RCTs without major defect or observational studies with strong evidences	Strongly recommend; apply for most patients
1B: Strongly recommended, based on intermediate-quality evidences	Remarkably higher benefits than harms	RCTs with obvious limitations (defects in methods and results) or observational studies with relatively strong evidences	Strongly recommend; apply for most patients
1C: Strongly recommended, based on low-quality evidences	Remarkably higher benefits than harms	Observational studies or case studies	Strongly recommend; may be adjusted if higher-quality evidence occurs
2A: Weakly recommended, based on high-quality evidences	The benefits and harms are very close	RCTs without major defects or observational studies with strong evidences	Weakly recommend; adjusted based on the patients’ specific conditions
2B: Weakly recommended, based on intermediate-quality evidences	The benefits and harms are very close	RCTs with obvious limitations (defects in methods and results) or observational studies with relatively strong evidences	Weakly recommend; adjusted based on the patients’ specific conditions
2C: Weakly recommended, based on low-quality evidences	The benefits and harms are very close	Observational studies or case studies	Do not recommend; other alternative methods are available

RCT, randomized controlled trial.