PROVE-HF Surgical Score Calculator

Clinical calculation support · Verify results before clinical use

Surgical mortality risk score

Enter the nine published predictors. The score estimates in-hospital or 30-day mortality after valvular surgery in patients with suspected heart failure.

0Total points

Age ≥60 years

Patient age at surgery is 60 years or older.

+2 points

NYHA functional class IV

Severe symptoms at rest or with minimal activity, documented as NYHA class IV.

+1 point

LVEF <35%

Most recent clinically appropriate preoperative left ventricular ejection fraction.

+2 points

eGFR <50 mL/min/1.73 m²

Use the estimated glomerular filtration rate in mL/min/1.73 m².

+2 points

Preoperative dialysis

Patient receives dialysis before the operation.

+4 points

Left main coronary artery stenosis

Left main artery stenosis documented in the preoperative assessment.

+1 point

Non-elective surgery

Operation is classified as non-elective.

+3 points

Cardiopulmonary bypass time >200 minutes

Actual CPB duration; this is an intraoperative variable.

+3 points

Perioperative transfusion

Transfusion during the perioperative period, as defined in the source study.

+1 point
0Score
Published model estimate
Score 0
Predicted mortality: 0.8% in the derivation cohort.
Clinical interpretationThis score is a risk estimate from the original study cohort, not an individualized guarantee. Consider the patient's full clinical context and locally validated surgical risk tools.
Important scope note: This is the simplified score published by Lin et al. for patients with suspected heart failure undergoing valvular surgery. It was developed using a Chinese cardiac surgery registry and requires external validation before assuming calibration in other populations. CPB duration and perioperative transfusion are intraoperative/perioperative predictors, so the complete score cannot be known from preoperative data alone. The former four-variable formula using eGFR × LVEF × hemoglobin ÷ BNP and cutoffs of 22.51/68.10 are not supported by the cited publication and have been removed.

About the score

The study defined postoperative mortality as death before hospital discharge or within 30 days after surgery. The final simplified model assigns weighted points to nine predictors, for a total score from 0 to 19.

Published score weights

PredictorCriterionPoints
Age≥60 years2
NYHA classIV1
LVEF<35%2
eGFR<50 mL/min/1.73 m²2
DialysisPreoperative dialysis4
Left main stenosisPresent1
Surgery statusNon-elective3
CPB duration>200 minutes3
TransfusionPerioperative transfusion1

Predicted mortality in the derivation cohort

ScorePredicted mortalityObserved mortality
00.8%0%
11.3%1.5%
22.1%1.8%
33.4%3.6%
45.4%7.2%
58.5%11.8%
612.8%11.5%
719.3%15.5%
827.9%23.1%
938.5%32.0%
≥1050.3%45.5%

These are cohort-level estimates reported in the original study, not universal risk categories. The study reported a testing-group AUC of 0.776; external validation in other populations remains necessary.

Reference

Lin H, Gong J, An K, Wu Y, Zheng Z, Hou J.A New Risk Score for Predicting Postoperative Mortality in Suspected Heart Failure Patients Undergoing Valvular Surgery. Reviews in Cardiovascular Medicine. 2023;24(2):38.https://doi.org/10.31083/j.rcm2402038

The published cohort included 14,645 patients from the Chinese Cardiac Surgery Registry; the derivation cohort had 11,292 patients and the testing cohort had 3,353 patients.

Limitations

  • Designed for suspected heart failure patients undergoing valvular surgery, not all non-cardiac or cardiac operations.
  • The model is region-specific; performance may differ in other health systems and patient populations.
  • CPB time and perioperative transfusion mean that the final score incorporates perioperative information.
  • Use alongside clinical assessment and established institutional risk assessment, not as the sole basis for decisions.
Clinical disclaimer: This calculator is an aid to risk assessment and does not replace clinician judgement, shared decision-making, or validated local protocols. Do not use this score outside the population and outcome for which it was developed without appropriate validation.