Development of a screening tool using electronic health records for undiagnosed Type 2 diabetes mellitus and impaired fasting glucose detection in the Slovenian population

Štiglic, Gregor; Kocbek, Primož; Cilar, Leona; Fijačko, Nino; Stožer, Andraž; Zaletel, Jelka; Sheikh, Aziz; Bržan, Petra Povalej

doi:10.1111/dme.13605

Cited by 18 publications

(23 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similar results were found in the previous research conducted in Slovenia 17 , where the most important variables were hyperglycemia history, gender, age, physical activity, waist circumference, (BMI), diabetes in family, fruit and vegetables consumption and antihypertensive drugs. Variables as age, parental history of diabetes and BMI were statistically significant predictors of T2DM already in The Framingham Offspring Study 39 .…”

Section: Discussionsupporting

confidence: 91%

“…Initially, the dataset consisted of 111 variables including a group of variables related to the FINDRISC (FR) questionnaire 16 such as variables representing physical activity (at least 30 min during the day), fruit and vegetable consumption as well as keeping a track of medical history including the history of antihypertensive drug treatment, history of high blood glucose levels and family history of diabetes. Consequently, all cases with any missing FR variable were removed from the dataset to allow comparison to the FR-based model that was developed for Slovenian population by Stiglic et al 17 .…”

mentioning

confidence: 99%

“…It is important to note that we used lm as a baseline model with a fixed set of variables. More specifically, lm was always built using only seven FR variables representing different questions, which were previously used in development of a simplified screening tool for undiagnosed T2DM and IFG in the Slovenian population 17 .…”

mentioning

confidence: 99%

See 2 more Smart Citations

Early detection of type 2 diabetes mellitus using machine learning-based prediction models

Kopitar

Kocbek

Cilar

et al. 2020

Sci Rep

233

114

View full text Add to dashboard Cite

Most screening tests for T2DM in use today were developed using multivariate regression methods that are often further simplified to allow transformation into a scoring formula. The increasing volume of electronically collected data opened the opportunity to develop more complex, accurate prediction models that can be continuously updated using machine learning approaches. This study compares machine learning-based prediction models (i.e. Glmnet, RF, XGBoost, LightGBM) to commonly used regression models for prediction of undiagnosed T2DM. The performance in prediction of fasting plasma glucose level was measured using 100 bootstrap iterations in different subsets of data simulating new incoming data in 6-month batches. With 6 months of data available, simple regression model performed with the lowest average RMSE of 0.838, followed by RF (0.842), LightGBM (0.846), Glmnet (0.859) and XGBoost (0.881). When more data were added, Glmnet improved with the highest rate (+ 3.4%). The highest level of variable selection stability over time was observed with LightGBM models. Our results show no clinically relevant improvement when more sophisticated prediction models were used. Since higher stability of selected variables over time contributes to simpler interpretation of the models, interpretability and model calibration should also be considered in development of clinical prediction models. Type 2 diabetes mellitus (T2DM) is very common and is responsible for very considerable morbidity, mortality. Furthermore, it is a substantial financial drain both on individuals/families, health systems and societies. Of major concern is that the incidence and prevalence of T2DM are increasing rapidly-globally. In 2017, it was estimated that 425 million people had any type of diabetes (approx. 5.5% of worldwide population) of which 90% had T2DM and according to projection estimations the prevalence is going to increase substantially in the coming years; by 2,045, for example, a 48% increase of prevalence from the above numbers is expected or in absolute numbers an estimated 629 million people (approx. 6.6% of the worldwide population) are expected to be suffering from any type of diabetes 1. T2DM can lead to substantially increased risk of macrovascular and microvascular disease, especially in those with inadequate glycaemic control 2. Progression of T2DM from impaired fasting glucose is typically slow and more importantly, its symptoms may remain undetected for many years. Delays in diagnosis are an important contributory factor to poor control and risk of complications 3. Data mining is nowadays applied to various fields of science, including healthcare and medicine. Often applied are pattern recognition, disease prediction and classification using various data mining techniques 4. Due to the increased prevalence of T2DM, various techniques have been used to build predictive models and models for early disease diagnosis, such as logistic and Cox proportional hazard regression models 5-7 , Random Forest 8,9 , boosted ensembles 10,1...

show abstract

Section: Discussionsupporting

confidence: 91%

mentioning

confidence: 99%

See 1 more Smart Citation

Early detection of type 2 diabetes mellitus using machine learning-based prediction models

Kopitar

Kocbek

Cilar

et al. 2020

Sci Rep

233

114

View full text Add to dashboard Cite

show abstract

“…For example, for the age component of the tool, several Caucasian studies have reported a mean age greater than that of our study [12,19,21,36]. This increase in age could lead to an increase in the risk score in this population [12,21,47,49]. Several studies conducted in South Africa have suggested threshold values of waist circumference different from those currently used for African populations [50][51][52].…”

Section: Optimal Threshold For the Prediction Of Type2 Diabetes With contrasting

confidence: 51%

“…In addition, the threshold used in Caucasians studies, which ranges from 11 to 15 in most FINDRISC questionnaire validation studies is higher than that of our study. Optimal thresholds used were higher or equal to 11 [20,21,[47][48][49].…”

Section: Optimal Threshold For the Prediction Of Type2 Diabetes With mentioning

confidence: 99%

Optimal Threshold of the Finnish Diabetes Risk Score (Findrisc) for Screening at-Risk Adults in an African Population in Southern Benin

Metonnou-Adanhoume¹,

Aguèh²,

Azandjèmé³

et al. 2019

ujph

View full text Add to dashboard Cite

This study aims to determine the performance of the FINDRISC tool for screening adults at risk for type2 diabetes in an African population in southern Benin. This retrospective study included 536 subjects aged 25 to 65 years. The FINDRISC questionnaire score was calculated using seven components: age, body mass index, waist circumference, regular practice of 30 minutes of physical activity, regular consumption of five portions of fruits and vegetables, the family history of diabetes and the presence of high blood pressure. The performance of the tool was determined from the FINDRISC score and fasting blood glucose The "Receiver Operator Characteristic" (ROC) Curve of the FINDRISC score was used to define the optimal cut-off value of the tool for. The FINDRISC threshold-optimal value for detecting T2D risk in southern Benin was 8.5 with 77% sensitivity, 89% specificity, 45% positive predictive value and 71% negative predictive value. The area under the ROC curve was 0.86 (95% CI: 0.81-0.90). This study revealed that the optimal threshold value for detecting the medium or high risk of T2D observed in Benin is lower than Finland's one. Further studies including larger and more representative samples need to be carried out to confirm this finding.

show abstract

Explaining artificial intelligence with visual analytics in healthcare

Ooge

Štiglic

Verbert

2021

WIREs Data Min & Knowl

Self Cite

View full text Add to dashboard Cite

To make predictions and explore large datasets, healthcare is increasingly applying advanced algorithms of artificial intelligence. However, to make wellconsidered and trustworthy decisions, healthcare professionals require ways to gain insights in these algorithms' outputs. One approach is visual analytics, which integrates humans in decision-making through visualizations that facilitate interaction with algorithms. Although many visual analytics systems have been developed for healthcare, a clear overview of their explanation techniques is lacking. Therefore, we review 71 visual analytics systems for healthcare, and analyze how they explain advanced algorithms through visualization, interaction, shepherding, and direct explanation. Based on our analysis, we outline research opportunities and challenges to further guide the exciting rapprochement of visual analytics and healthcare.

show abstract

Development of a screening tool using electronic health records for undiagnosed Type 2 diabetes mellitus and impaired fasting glucose detection in the Slovenian population

Cited by 18 publications

References 36 publications

Early detection of type 2 diabetes mellitus using machine learning-based prediction models

Early detection of type 2 diabetes mellitus using machine learning-based prediction models

Optimal Threshold of the Finnish Diabetes Risk Score (Findrisc) for Screening at-Risk Adults in an African Population in Southern Benin

Explaining artificial intelligence with visual analytics in healthcare

Contact Info

Product

Resources

About