Welcome to the Spearhead Antibiotic Resistance Calculator.
This interactive tool displays predicted antibiotic resistance probabilities using routinely available patient and laboratory data. By entering patient characteristics in the left-hand panel, users can compare resistance risk across multiple antibiotics and visualise the key factors influencing each prediction.
Compare All Antibiotics
Feature Importance
About This Tool
The predictions shown in this interface are generated using machine-learning models developed in Subproject 2 of the SPEARHEAD Innosuisse project. These models were trained on real-world electronic health record data from two Basel hospitals and rely exclusively on information available at the time of clinical assessment, including demographics, routine laboratory markers, urine analysis results, and prior resistance history. A detailed description of the modelling framework, datasets, and full performance evaluation is provided in our accompanying scientific manuscript, currently being prepared for journal submission (a link will be added here once available).
The upper panel offers an overview of predicted resistance probabilities for multiple antibiotics, allowing rapid visual comparison of empiric treatment options. The feature-importance view displays SHAP-based explanations for the selected antibiotic, illustrating which variables contribute positively or negatively to the model's predicted risk.
This web application was developed as a research and demonstration platform within SPEARHEAD. It is intended to facilitate exploration of the models, support communication with clinical, scientific, and governance stakeholders, and provide an accessible interface for future validation activities. It is not a medical device and is intended solely for research and evaluation purposes.