'Eyeball Test' Is Poor Estimate of Cardiac Surgical Risk
Statistical methods better for estimating mortality risk after cardiac surgery
THURSDAY, Jan. 16, 2014 (HealthDay News) -- Compared with subjective physician estimates, statistical risk estimates are more accurate for predicting operative and long-term mortality in patients undergoing cardiac surgery, according to research published online Jan. 14 in Circulation: Cardiovascular Quality and Outcomes.
Renuka Jain, M.D., of the University of Minnesota in Minneapolis, and colleagues analyzed data from 5,099 consecutive patients who underwent coronary artery bypass grafting or valve surgery at a Veterans Affairs Medical Center. The authors sought to compare physician risk estimates (eyeball test) and statistical risk estimates for mortality.
The researchers found a modest correlation between the physician risk estimate (mean, 5.6 percent) and the statistical risk estimate (mean, 4.3 percent), and both methods overestimated operative mortality risk. Statistical risk estimate was significantly better than physician subjective risk estimate in distinguishing patients who died from those who survived at 30 days, one year, and five years, after surgery. Except for high-risk patients, the physician risk estimate was higher than the statistical risk estimate for all subgroups.
"In patients undergoing cardiac surgery, statistical risk estimate is a better method to predict operative and long-term mortality compared with physician's subjective risk estimate," the authors write. "However, both methods modestly overestimate actual operative mortality risk."
One author disclosed financial ties to Medtronic and Boston Scientific.
Abstract (http://circoutcomes.ahajournals.org/content/early/2014/01/13/CIRCOUTCOMES.113.000329.abstract )Full Text (subscription or payment may be required) (http://circoutcomes.ahajournals.org/content/early/2014/01/13/CIRCOUTCOMES.113.000329.full.pdf+html )Editorial (subscription or payment may be required) (http://circoutcomes.ahajournals.org/content/early/2014/01/13/CIRCOUTCOMES.113.000760.full.pdf+html )