You are a primary care practitioner considering the possibility of the use of aspirin for primary prevention of serious cardiovascular events and cancer in a 60-year-old man. He has hypertension, adequately controlled with thiazides, but he is otherwise healthy; he does not have diabetes or dyslipidemia, does not smoke, and has no family history of heart disease.
To inform your decision, you search first for an evidence-based recommendation and find the following: “In many adults, the benefits of aspirin exceed the risks (principally bleeding). For individuals 50 years or older without excess bleeding risk, we suggest low-dose daily aspirin (75-100 mg) (weak recommendation based on moderate confidence in effect estimates).”1
After reading the corresponding Users' Guide (see Chapter 26, How to Use a Patient Management Recommendation: Clinical Practice Guidelines and Decision Analyses), you know that weak recommendations reflect the judgment of a specific group (eg, a guideline panel) that an individualized decision is necessary. You are curious, however, about the rationale for a weak recommendation for aspirin use and decide that this is a good opportunity to understand more about the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach to move from evidence to recommendations.
Direction and Strength of Recommendations
Like recommendations based on other evidence-rating systems, recommendations developed with the GRADE approach specify the direction of the recommendation (in favor or against the intervention) and are classified as strong or weak.
Strong recommendations apply to all or almost all patients and indicate that clinicians do not (if they are ready to trust the judgment of the panel) require thorough (or even cursory) review of the underlying evidence or a discussion of benefits and risks with the patient.
Weak recommendations, in contrast, apply to most patients but not to everyone. To effectively use weak recommendations, clinicians need to understand and consider the key factors that drive the direction and strength of the recommendation. Given that weak recommendations are usually sensitive to patients' values and preferences, they require that one be prepared to engage the patient in shared decision making (see Chapter 27, Decision Making and the Patient).
Developing Grade Recommendations
GRADE provides a structure for assessing confidence in estimates from evidence summaries (eg, systematic reviews and meta-analyses) and, in the context of guidelines, also for moving from evidence to recommendations. Figure 28.1-1 presents the steps involved in developing a GRADE recommendation, including evidence synthesis and moving from evidence to recommendations. In the first step, guideline panelists formulate a clinical question, which involves specifying the target population, the intervention of interest, and an appropriate comparator. Conceptually, the final recommendation represents the answer to this question. Having formulated the question, guideline panelists select the relevant outcomes and rate their importance as critical, important, or not important for decision making; ...