This article is more than 5 years old. Information may no longer be current.
Sex-specific risk scores predict mortality in patients undergoing exercise testing
Click Here to Manage Email Alerts
Sex-specific risk scores were better predictors of mortality than conventional risk assessment tools in patients undergoing exercise treadmill testing, according to findings published in JAMA Cardiology.
Researchers conducted a retrospective cohort study of patients aged 18 years or older who underwent exercise treadmill testing to determine which variables contribute to mortality in women and men and whether sex-specific risk scores better predict mortality.
“The Duke Treadmill Score, which has been used forever, was developed in middle-aged men using the VA database,” Leslie Cho, MD, section head for preventive cardiology and rehabilitation and director of the Women’s Cardiovascular Center at Cleveland Clinic, told Cardiology Today. “Some studies have shown it’s not particularly helpful in predicting outcomes in women. We wanted to look at patient characteristics as well as other exercise markers such as heart rate recovery to see if we could come up with a better predictor for CV outcomes in both men and women. We came up with a score that’s a much better predictor for men and women in terms of cardiac mortality.”
Paul C. Cremer, MD, and colleagues analyzed 59,877 patients treated at the Cleveland Clinic Foundation between 2000 and 2010 and 49,278 patients treated at Henry Ford Hospital, Detroit, from 1991 to 2009.
The Cleveland Clinic cohort was randomly divided into derivation and validation samples to assess two risk scores, one for men and one for women. The scores were compared with other assessment tools via net reclassification, C statistics and integrated discrimination improvement. The sex-specific risk scores were also evaluated for discrimination and calibration in the Henry Ford Hospital cohort.
In the Cleveland Clinic cohort (median age, 54 years; interquartile range, 45-63; 59% men), 4.2% of patients died during a median follow-up of 7 years, and in the Henry Ford Hospital cohort (median age, 54 years; interquartile range, 46-64; 53% men), 13.5% of patients died during a median follow-up of 10.2 years, Cremer, from the Heart and Vascular Institute, Cleveland Clinic, and colleagues wrote.
Components of scores
The risk score for women included 4 to 10 points based on peak treadmill metabolic equivalent of task (METs), 5 points for end-stage renal disease, 2 to 4 points based on weight, 1 point for age older than 65 years, 2 points for abnormal heart rate recovery, 2 points for former or current smoking and 1 point for diabetes.
The risk score for men included 4 to 10 points based on peak treadmill METs, 7 points for end-stage renal disease, 1 point for weight < 80 kg, 1 to 7 points based on age, 3 points for abnormal heart rate recovery, 1 point for former or current smoking, 2 points for history of hypertension and 5 points for history of HF.
“Heart rate recovery was never considered for the Duke Treadmill Score because we didn’t know about it then,” Cho told Cardiology Today. “For women, we now know that diabetes is a big deal for risk prediction, and many databases show that if a woman has diabetes, it negates the 10-year age difference [vs. men] in onset of heart disease, and that women with diabetes are at higher risk for HF than men with diabetes. In men, having HF and hypertension was a bigger weight. It’s nice to start with gender-specific medicine, which can be a step toward personalized medicine.”
The researchers reported that in the validation sample of the Cleveland Clinic cohort, for women, C statistics for the sex-specific risk score (0.79) were higher than the Duke Treadmill Score (0.7) and the Lauer nomogram (0.74), and the same was true for men (sex-specific score, 0.81; Duke Treadmill Score, 0.72; Lauer nomogram, 0.75).
Mainly due to correct reclassification of events, net reclassification and integrated discrimination improvement were superior in the sex-specific risk scores vs. the Duke Treadmill Score and the Lauer nomogram, Cremer and colleagues wrote.
The sex-specific risk scores had similar discrimination in the Henry Ford Hospital cohort (C statistic for women, 0.78; C statistic for men, 0.79), with reasonable calibration, according to the researchers.
“The scores risk-stratify patients better,” Cho said in an interview. “You can put patients in a higher-risk category which warrants more aggressive control of hypertension, diabetes and hyperlipidemia, or you can put them in a lower-risk category where you don’t need to treat them as aggressively. The hard work of cardiology is how to modify treatment based on the patient’s risk.”
The researchers developed an online calculator using the sex-specific risk scores to predict 10-year mortality. It can be found at www.clevelandclinic.org/lp/hvi-tools/10YearMortality.html.
“Even when accounting for multiple comorbidities, exercise capacity was still the predominant risk factor in men and women,” Cremer and colleagues wrote. “This online calculator can be used by physicians and patients not only to assess prognosis but also emphasize the importance of exercise, even in the presence of other [CV] risk factors.”
Decision aids valuable
In a related editorial, Ashok Krishnaswami, MD, MAS, and colleagues wrote that “clinical prediction models are superior to eyeball testing, are here to stay and should be used routinely. The challenge is to improve patient outcomes while streamlining the workload of busy clinicians using the best possible available outcomes. The use of decision aids as an integral part of modern electronic health records may be an answer.” – by Erik Swain
For more information:
Leslie Cho, MD, can be reached at Heart and Vascular Institute, Cleveland Clinic, 9500 Euclid Ave, Ste JB-1, Cleveland, OH 44124; email: chol@ccf.org.
Disclosure: The researchers and editorial authors report no relevant financial disclosures.
Click Here to Manage Email Alerts