A Danish Diabetes Risk Score for Targeted Screening

Abstract

OBJECTIVE -To develop a simple self-administered questionnaire identifying individuals with undiagnosed diabetes with a sensitivity of 75% and minimizing the high-risk group needing subsequent testing. RESULTS -The final risk score included age, sex, BMI, known hypertension, physical activity at leisure time, and family history of diabetes, items independently and significantly (P Ͻ 0.05) associated with the presence of previously undiagnosed diabetes. The area under the receiver operating curve was 0.804 (95% CI 0.765-0.838) for the first half of the Inter99 population, 0.761 (0.720 -0.803) for the second half of the Inter99 population, and 0.803 (0.721-0.876) for the ADDITION pilot study. The sensitivity, specificity, and percentage that needed subsequent testing were 76, 72, and 29%, respectively. The false-negative individuals in the risk score had a lower absolute risk of ischemic heart disease compared with the true-positive individuals (11.3 vs. 20.4%; P Ͻ 0.0001). RESEARCH DESIGN AND METHODSCONCLUSIONS -We developed a questionnaire to be used in a stepwise screening strategy for type 2 diabetes, decreasing the numbers of subsequent tests and thereby possibly minimizing the economical and personal costs of the screening strategy.

A prediction model for type 2 diabetes risk among Chinese people

“…In cross-sectional studies conducted in the USA and Europe, prediction models based on clinical information and lifestyle-related factors have appeared to be useful for identifying undiagnosed diabetes cases and high HbA 1c levels in screening populations [7,8,38,39]. For example, the Cambridge model has been applied successfully to identify individuals with high HbA 1c levels [9,10].…”

“…The prediction models for the risk of diabetes can help to guide screening and interventions and to predict diabetes occurrence [3][4][5]. Routinely available and easily collected clinical and lifestyle-related information has been found to be effective for identifying diabetes cases [6][7][8][9][10][11][12]. In addition, a prediction model has been developed in Mexican Americans [13] and further tested in Japanese Americans [14].…”

Appropriate cutoff values of anthropometric variables to predict cardiovascular outcomes: 7.6 years follow-up in an Iranian population

“…Furthermore, this may not be considered as the optimal point for a test in the clinic, as sensitivity versus specificity should be weighted against many other factors such as seriousness of the illness (that is, CVD events in our study), the test used for first evaluation, how invasive or feasible it might be and finally how often the test should be performed. 43 Furthermore, as a public health priority, it is necessary to cover a reasonable proportion of the population at risk; hence, we preferred not to miss more than 20% cases of incident CVD (that is, predefined sensitivity at the level of at least 80%) in identifying the appropriate cutoff point of each anthropometrics. The results, as expected, produced lower cutoff points of anthropometric variables with lower specificities.…”

Determinants of progression from impaired fasting glucose and impaired glucose tolerance to diabetes in a high-risk screened population: 3 year follow-up in the ADDITION study, Denmark

“…The The screening process The screening process was a stepwise, decision-tree model. Persons aged 40 to 69 years, registered with the participating practices in five counties in Denmark, received an invitation with a risk score questionnaire [18,19]. Persons with a risk score of 5 points or more were recommended to contact their general practitioners for further tests.…”

“…Measurements Three self-administered questionnaires were used: (1) the Danish Risk Score [19]; (2) a baseline questionnaire on medical history, family history of diabetes, smoking status (daily smoker or not) and alcohol consumption; and (3) a follow-up questionnaire on changes since screening. Physical activity was not part of the baseline questionnaire.…”