A tool designed to predict the risk of MI and cardiac arrest (MICA) after noncardiac surgery in older patients works significantly better than two widely used models that aren’t age-specific, according to a study published Nov. 16 in the Journal of the American Heart Association.

The Geriatric-Sensitive Cardiac Risk Index (GSCRI) outperformed the Gupta MICA calculator and the Lee Revised Cardiac Risk Index (RCRI) by 6 percent and 13 percent, respectively, when predicting the 30-day MICA risk of 172,905 patients age 65 and older.

Accuracy was measured using a technique called area under the curve, with the GSCRI scoring 0.76, the Gupta MICA scoring 0.70 and the RCRI scoring 0.63. A score of 1 represents a perfect predictive tool, while a 0.5 represents a worthless test—no better than a coin flip.

“This study demonstrates that currently used risk models … have moderate performance in older patients and tend to underestimate the actual cardiac risk in this age group,” wrote the UCLA-based research group, led by Rami Alrezk, MD. “With a significant change in the demographics and an increasing older patient population undergoing surgery, this Geriatric‐Sensitive Perioperative Cardiac Risk Index could be an effective tool for cardiac risk evaluation.”

There are 40 million people 65 or older living in the U.S., accounting for one-third of all inpatient surgeries, according to the authors. Those numbers are projected to increase to 72 million geriatric American by 2030, with a corresponding increase in the number of surgeries in this population.

Compared to younger patients in the study, the odds of MICA were about five times greater in patients 65 or over (about 1 percent versus 0.2 percent). However, neither of the previously used risk-estimating tools are specifically designed with geriatric patients in mind.

Older individuals are sometimes excluded completely from clinical trials based on age-related comorbidities, or, when included, they are pooled together with younger participants with much lower risk, Alrezk and colleagues noted.

“Developing predictive models on these pooled data that ignore age categories can lead to models that are dominated by variables and coefficients not optimized for performance in geriatric patients and hence provide decreased predictive accuracy and lower sensitivity to certain geriatric characteristics,” the authors wrote.

Alrezk et al. developed their new tool using the National Surgical Quality Improvement Program (NSQIP) 2013 geriatric cohort and validated it with the 2012 NSQIP cohort. They considered 17 variables for the GSCRI but settled on seven: functional status, American Society of Anesthesiologists classification, creatinine level greater than or equal to 1.5 milligrams per deciliter, diabetes status, stroke history, heart failure and type of surgery.

“Although additional relevant variables could have been added, the increased model complexity would not meaningfully improve the model's predictive ability based on our examination of the upper limit of model performance from the candidate variables,” the researchers wrote. “Creating a parsimonious model was essential to ensure the ease of use that physicians working in clinical settings require.”

The researchers also plan to develop an online calculator—composed of seven questions—to help physicians quickly use the GSCRI to estimate risk in everyday practice. The authors cautioned the calculator should be a supplemental tool, “because no risk model can substitute for the clinical judgment of physicians.”

Alrezk and colleagues added there were limited geriatric-specific data available in the study group, so future work could focus on testing whether additional variables could improve the GSCRI’s predictive power. Some possibilities, they wrote, include inflammatory factors, nutritional status, functional status, depression, cognition and frailty indices.

In addition, the model will likely have to be updated every few years to account for advances in medical practice, including new surgical techniques.