CERTAINTY OF NET BENEFIT

When making a decision how sure can we be that we are making the right decision? When making a recommendation (a suggestion to someone else for the decision they are making) how sure can be that we are making the right recommendation, and how sure can they be that we are making the right recommendation?

For many decisions or recommendations our desire to reduce UNCERTAINTY and our desire to have an impossible state of CERTAINTY leads us to methods to understand and convey the degree of CERTAINTY around the concepts contributing to our decision-making.

There are many ways to do this -- 95% confidence intervals using statistical techniques for mathematical concepts, vote counts to report the ratings of others, seeking a trusted advisor for specific feedback.

But what are we trying the determine the CERTAINTY about? CERTAINTY in the trustworthiness of the facts contributing to our decision (often called evidence), CERTAINTY in the prediction of what will happen in the future, CERTAINTY in judgment that we are making the "right" choice?

Half a year ago I posted on CERTAINTY OF NET EFFECT as a most important focal point for making recommendations. I am traveling to Rome this week to discuss this with the Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group - an international collaborative guiding how clinical practice guidelines and evidence-based medicine is expressed. GRADE recently clarified that "certainty of evidence" (estimating effects) means certainty that the true effect lies within a specified range or on one side of a threshold. That is important as the contextualization makes the results more meaningful.

A contextualized CERTAINTY OF NET EFFECT for decision-making is the CERTAINTY OF NET BENEFIT -- the certainty that the decision will lead to more good than harm.

If we figure out simple ways to convey the CERTAINTY OF NET BENEFIT we could make it easier to communicate the degree of CERTAINTY around recommendations.