New calculator predicts two-year mortality risk in type 2 diabetes

Researchers have developed a new Web-based calculator that predicts short-term mortality risk in patients with type 2 diabetes.

The researchers investigated two cohorts of patients. The first “training set” comprised 679 white Italian patients with type 2 diabetes who were followed for more than seven years. The second “validation set” comprised 936 white Italian patients with diabetes who were followed for approximately 4.5 years. In both cases, the only exclusion criterion was poor life expectancy due to malignancy. Both sets were tracked for all-cause mortality via check-ins with city registry offices, or phone calls to the patient or relatives. Results were published online May 1 in Diabetes Care.

Nine variables were significantly associated with two-year all-cause mortality and were used to make a final predictive model (C statistic, 0.88 in training set and 0.82 in validation set). Those variables were age, body mass index, diastolic blood pressure, LDL cholesterol, triglycerides, HDL cholesterol, urine albumin-to-creatinine ratio, antihypertensive therapy, and insulin therapy. Researchers then used an algorithm to identify patients at intermediate and high risk of death (hazard ratios, 7.0 and 24.4, respectively, compared with those at low risk) and developed a Web-based risk calculator (available at operapadrepio.it/rcalc/rcalc.php) .

This is the first known study to develop a well-performing model for predicting death in type 2 diabetes that was then validated in a second independent sample, the authors wrote. Study limitations include lack of information on causes of death and on previous cardiovascular events (a major risk factor for all-cause death) and unknown generalizability to non-Italian patients, they wrote. Still, the initial results seem promising and may help clinicians prioritize resources “for targeting aggressive preventive and treatment strategies in a subset of very high-risk individuals,” the authors concluded.