Risk scores based on self-reported or available clinical data to detect undiagnosed type 2 diabetes: a systematic review

Diabetes Res Clin Pract. 2012 Dec;98(3):369-85. doi: 10.1016/j.diabres.2012.09.005. Epub 2012 Sep 23.


Objective: To systematically review published primary research on the development or validation of risk scores that require only self-reported or available clinical data to identify undiagnosed Type 2 Diabetes Mellitus (T2DM).

Methods: A systematic literature search of Medline and EMBASE was conducted until January 2011. Studies focusing on the development or validation of risk scores to identify undiagnosed T2DM were included. Risk scores to predict future risk of T2DM were excluded.

Results: Thirty-one studies were included; 17 developed a new risk score, 14 validated existing scores. Twenty-six studies were conducted in high-income countries. Age and measures of body mass/fat distribution were the most commonly used predictor variables. Studies developing new scores performed better than validation studies, with 11 reporting an AUC of >0.80 compared to one validation study. Fourteen validation studies reported sensitivities of <80%. The performance of scores did not differ by the number of variables included or the country setting.

Conclusions: There is a proliferation of newly developed risk scores using similar variables, which sometimes perform poorly upon external validation. Future research should explore the recalibration, validation and applicability of existing scores to other settings, particularly in low/middle income countries, and on the utility of scores to improve diabetes-related outcomes.

Publication types

  • Meta-Analysis
  • Review
  • Systematic Review

MeSH terms

  • Adiposity
  • Age Factors
  • Body Mass Index
  • Diabetes Mellitus, Type 2 / diagnosis*
  • Diabetes Mellitus, Type 2 / epidemiology
  • Evidence-Based Medicine
  • Humans
  • Mass Screening / methods*
  • Risk
  • Self Report