Survey of diabetes risk assessment tools: concepts, structure and performance

Diabetes Metab Res Rev. 2012 Sep;28(6):485-98. doi: 10.1002/dmrr.2296.

Abstract

The objective of this study is to review the effectiveness and limitations of existing diabetes risk screening tools to assess the need for further developing of such tools. An electronic search of the EMBASE, MEDLINE, and Cochrane library supplemented by a manual search was performed from 1995-2010. The search retrieved a total of 2168 articles reporting diabetes risk assessment tools which, after culling, produced 41 tools developed in 22 countries, with the majority (n = 26) developed in North America and Europe. All are short questionnaires of 2-16 questions incorporating common variables including age, gender, waist circumference, BMI, family history of diabetes, history of hypertension or antihypertensive medications. While scoring format and cut-offs point are diverse between questionnaires, overall accuracy value range of 40-97%, 24-86% and 62-87% were reported for sensitivity, specificity and receiver operating characteristic curve respectively. In summary, there is a trend of increasing availability of diabetes prediction tools with the existing risk assessment tools being generally a short questionnaire aiming for ease of use in clinical practice. The overall performance of existing tools showed moderate to high accuracy in their predictive performance. However, further detailed comparison of existing questionnaires is needed to evaluate whether they can serve adequately as diabetes risk assessment tool in clinical practice.

Publication types

  • Research Support, Non-U.S. Gov't
  • Review

MeSH terms

  • Asia
  • Australia
  • Diabetes Mellitus / diagnosis*
  • Diabetes Mellitus / etiology*
  • Europe
  • Humans
  • Mass Screening
  • North America
  • ROC Curve
  • Risk Assessment*
  • Sensitivity and Specificity
  • Surveys and Questionnaires
  • United States
  • Waist Circumference