Telemedicine Use Predictors Among Cardiology Professionals via XGBoost Model

112 healthcare professionals from a German cross-sectional survey provide data on telemedicine use determinants. The dataset contains the performance metrics and predictor importance results from a final XGBoost model developed by Pascal Petit, last updated in April 2026. The model achieved an AUROC of 0.88 and 79% accuracy in predicting telemedicine adoption.

Use Cases

• Predict telemedicine adoption likelihood among healthcare professionals based on demographic and attitudinal factors.
• Analyze the relative importance of promoting factors like telemedicine knowledge and specialty using SHAP values.
• Benchmark machine learning model performance for binary classification in small-sample healthcare survey data.
• Identify target groups for telemedicine intervention programs based on predictor patterns like age and perceived suitability.

Strengths

• Model performance is quantitatively reported with an AUROC of 0.88 and a 95% confidence interval.
• Predictor importance was evaluated using Shapley additive explanations (SHAP), providing interpretability.
• The modeling approach used nested cross-validation to prevent overfitting.
• Dataset is openly licensed under CC-BY-4.0.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is very small at 5.5 KB, indicating limited scope, likely containing only model results.
• Data is based on a specific sample of 112 German healthcare professionals, which may limit generalizability.

Provenance

Source

Secondary analysis of a German cross-sectional survey.

Collection Method

Bayesian Model Averaging and XGBoost machine learning applied to survey data.

Time Range

Survey collection date is unspecified.

Freshness

Last updated 2026-04-20 17:49:16; freshness should be verified.

Geography

Germany