Researchers have created an online mortality-risk calculator for patients with type 2 diabetes, which stratifies patients into low, medium, or high risk of dying from any cause within 2 years.
By plugging in values for 9 readily available patient characteristics — age, body mass index (BMI), diastolic blood pressure, LDL cholesterol, triglycerides, HDL cholesterol, urine albumin-to-creatinine ratio, antihypertensive treatment, and insulin therapy — a physician can quickly determine whether a patient has a high risk for death.
"The novelty and the importance of this study is that we provide physicians with a simple, free, and validated Web-based risk [calculator] able to predict mortality in their type 2 diabetic patients," lead author Salvatore de Cosmo, MD, from Casa Sollievo della Sofferenza, in San Giovanni Rotondo, and senior author Vincenzo Trischitta, MD, from Sapienza University, in Rome, Italy, told Medscape Medical News in an email.
"We do believe that the implementation of our model may help prioritize the use of available resources for targeting aggressive preventive and treatment strategies in a subset of very high-risk individuals."
Having diabetes doubles a person's risk of dying, the authors write, and it would be useful for physicians to be able to identify which diabetic patients are at greatest risk. The few previously proposed models to predict mortality in diabetes were not validated in an independent cohort and did not use recently available statistical tools, they explain.
The researchers analyzed data from 2 prospective cohorts of patients with type 2 diabetes. They developed their model based on data from 679 patients in the Gargano Mortality Study, who were followed for 7.4 years. Risk predictors were added one by one, until there was no further improvement in prediction of death.
They validated their model using data from 936 patients in the Foggia Mortality trial, who were followed for 4.5 years.
Elevated age, urine-albumin-to-creatinine ratio, and LDL cholesterol, as well as taking insulin therapy or being on antihypertensive treatment all predicted higher mortality risk. Elevated HDL, BMI, and diastolic blood pressure predicted lower mortality risk, however, and triglyceride levels also showed a trend toward an inverse correlation.
"We want to stress that our interest was not a matter of pathophysiology but simply that of unraveling good-enough markers and methodological approaches to predict all-cause mortality," they commented.
The overall mortality risk score ran from 0 to 1, with a mean of 0.53. The researchers partitioned this into 3 risk categories: low (risk score 0.67 or less), intermediate (0.68 to 0.79), and high (0.80 or more). At 2 years, compared with the patients with low risk scores, those with medium scores were 7 times more likely to have died, and those with high scores were 24 times more likely to have died.
"To the best of our knowledge, this is the first study developing a well-performing model for this event that was validated in a second independent sample," the authors write. "Our model is parsimonious, with a few simple-to-measure variables needed to make it very informative," they add.
The calculator will help physicians identify which patients should be monitored more closely and which ones are candidates for more aggressive strategies to treat modifiable risk factors, Drs. De Cosmo and Trischitta note.
However, it should also be tested in other populations of diabetes patients in order to address its generalizability, they point out, since the studies to develop and validate it were conducted in primarily white, Italian populations.
No comments:
Post a Comment