According to this system, there are two components to any treatment recommendation:

1. Designation of the strength of the recommendation (strong: 1; or weak: 2) depending on the risk and burden of a particular diagnostic test or a therapeutic procedure to the patient vs the expected benefit.
   • Grade 1 recommendations are those in which the benefits of an intervention clearly outweigh its risk and burdens. All well-informed patients would choose such a treatment, and the physician can securely recommend it without a detailed knowledge of the underlying data.
   • Grade 2 recommendations are weaker and reflect therapies where the benefits and risks are uncertain or are more closely balanced. For such interventions, patients may choose different options based on their underlying values.

The words “we recommend” are used for GRADE 1—strong recommendations—if the benefits clearly outweigh risks and burdens, or vice versa; the words “we suggest” are used for GRADE 2—weak recommendations—when the benefits are closely balanced with risks and burdens.

2. An evaluation of the level of evidence (A to C) based on the confidence that the estimate of effect is correct.
   • Grade A, or high-quality evidence,usually comes from well-executed randomized trials yielding consistent results, and occasionally, obs ervational studies with large effects.
   • Grade B, or moderate-quality evidence, comes from randomized clinical trials with important limitations, inconsistent randomized trials, and strong observational studies.
   • Grade C, or low or very low quality evidence, includes flawed randomized trials and most observational studies as well as data from case reports, descriptive studies, and expert opinion.