Hoopcare ML Surgical Risk Calculator
Welcome to the Surgical Risk Calculator
Version 2.0 · scoring parameters updated August 2026
Enter preoperative information about a patient to see estimated risks of postoperative complications, the preoperative pathway those risks imply, and the resources the case is likely to consume — operating-room time, red cells, ward and critical-care beds.
Approach (if relevant)
Care setting
Emergency case
Sex
Labs & studies
most recent · 90 daysClinical notes & report analysis optional — paste a chart or upload documents; feeds the LLM feature set
Upload medical documents
PDF, CCD or clinic notes. All health records may not be available.
Sent to a Hoopcare endpoint and read by a language model. Synthetic demo — do not paste real patient information.
0 / 50,000Features computed from clinical notes · AI review
How the estimate is produced
A fixed weight table scores the chart first and always produces an answer on its own. A general-purpose language model reads any pasted clinical text into the questionnaire — it never adjusts a number. The complication risks, blood order and surgery-day timeline come from LightGBM models trained on de-identified Mayo Clinic surgical data (research use only): one set of structured features drives all three, unanswered fields follow the models' learned missing-value paths, and when the service does not answer, the deterministic estimates stand — the badge beside the score says which you are looking at.
Disclaimer
Hoopcare is not a medical device and has not been evaluated by the FDA. The patients shown here are synthetic; complication, blood and timeline estimates come from research models trained on de-identified Mayo Clinic surgical data — outputs must not inform patient care. The percentages are estimates from a limited set of inputs; factors not captured here may raise or lower real risk, and a complication can occur even when an estimate is low. Nothing here replaces the judgement of the clinician responsible for the patient.
Not for
Patient care decisions · site-of-service decisions · coverage, payment or prior authorization · any automated use. Text pasted into the calculator leaves the browser and is read by a language model; never paste real patient information.