WEDNESDAY, March 18 (HealthDay News) -- A score that uses medical data and does not require laboratory tests accurately estimates 10-year diabetes risk, according to research published March 17 in BMJ Online First.
Julia Hippisley-Cox, M.D., from the University of Nottingham in the United Kingdom, and colleagues developed a score to estimate the 10-year risk of type 2 diabetes (the QDScore) using data from 2.5 million patients (78,081 developed diabetes) and validated the score in 1.2 million patients (37,535 developed diabetes). The model incorporated 10 variables available through medical records, including self-assigned ethnicity and social deprivation.
The researchers found that ethnicity had a significant effect on diabetes risk, with Bangladeshi men having the highest risk (hazard ratio 4.53) compared with whites. In the validation group, the model could account for 51.53 percent of the variation in women and 48.16 percent of the variation in men. The score was well-calibrated and showed good discrimination (D statistic 2.11 in women and 1.97 in men), the report indicates. Predicted and observed risk correlated well (97 percent) and the area under the receiver-operating characteristics curve for the score exceeded 0.80, the authors report.
"The QDScore is the first risk prediction algorithm to estimate the 10-year risk of diabetes on the basis of a prospective cohort study and including both social deprivation and ethnicity," Hippisley-Cox and colleagues conclude. "The algorithm does not need laboratory tests and can be used in clinical settings and also by the public through a simple Web calculator (www.qdscore.org)."
Hippisley-Cox is director of ClinRisk Ltd., which produces software to ensure reliable implementation of clinical risk algorithms within computer systems to improve patient care.
Abstract
Full Text
Editorial
