Clinical predictions can have a major impact on decisions doctors make for patients, so it’s important to have a firm grasp on how predictive prognostics actually are, according to a new study.
These types of predictions can especially impact decisions for critically ill patients in intensive care, so a research team from the University of Pennsylvania looked at 10 different predictive measures, including survival within six months, ability to return home and going to the bathroom independently to examine how accurate clinicians' predictions were, according to findings published in the Journal of the American Medical Association.
When doctors were confident about their predictions, they were also significantly more accurate. The results were similar for nurses, according to the study, who were confident in their predictions slightly more often than physicians.
Both most accurately predicted a patient’s likely six-month survival chances, and least accurately predicted likely cognitive function.
One area where clinicians in the study struggled was applying population-level outcomes data to predict individual patient outcomes for conditions such as sepsis and acute respiratory distress syndrome, according to the study. The researchers add that physicians should be open about this when offering predictions at a patient’s bedside.
“When relaying such predictions at the bedside, it is important that clinicians acknowledge their uncertainty,” according to the study. “Most family members not only desire prognostic guidance from ICU clinicians but also appreciate that some uncertainty is inevitable.”
The findings show that much of the work in improving communication between physicians and patients and their families has centered on bettering shared decision-making and not so much on ensuring that prognostic information is accurate, Jesse Hall, M.D., professor emeritus of medicine and anesthesia and critical care, wrote in an accompanying editorial.
ICU clinicians will likely always operate in a “gray zone” when making recommendations based on prognostics, but the goal should be to offer as clear and accurate information as possible to patients and their families within that constraint, according to Hall, which makes studies into predictive accuracy all the more important.
“It is of paramount importance to understand how well calibrated clinicians are in generating prognostications, and how they may improve, because patients and families deserve the most realistic assessment of outcome, properly framed by the reality that ICU clinicians can never offer perfectly accurate predictions,” Hall wrote.