# A nomogram model incorporating blood biomarkers predicts 3-week functional outcomes in stroke patients

## TL;DR

A new model using blood biomarkers helps predict stroke patients' recovery after three weeks, aiding early care planning.

## Contribution

A nomogram integrating blood biomarkers for predicting stroke outcomes is developed and validated.

## Key findings

- Four blood biomarkers (PT, FIB, RBG, UA) independently predict 3-week stroke outcomes.
- The nomogram shows moderate discrimination (AUC = 0.714) and good calibration.
- The model provides positive net benefit for clinical decision-making.

## Abstract

Accurate prediction of functional outcomes of stroke remains clinically challenging. The present study was designed to identify baseline biomarkers in demographic, clinical data, and blood biomarkers to predict 3-week outcomes in stroke patients.

A prospective cohort of two hundred patients with stroke was recruited at the hospital and followed for 3 weeks. We applied the Barthel Index (BI) to measure the activities of daily living functions in stroke patients. The good outcome or poor outcome groups were classified based on the BI scores. A logistic regression analysis was performed to identify independent predictors, which were subsequently integrated into a nomogram. Discrimination and calibration values of the nomogram were analyzed, and its utility was assessed using decision curve analysis.

Four blood biomarkers, including PT (OR = 1.45, 95% CI: 1.05–2.12), FIB (OR = 1.49, 95% CI: 1.14–2.00), RBG (OR = 1.20, 95% CI: 1.03–1.40), and UA (OR = 1.00, 95% CI: 0.99–1.00) were independent predictors of the 3-week functional outcomes after a stroke. The nomogram incorporating these biomarkers demonstrated moderate discriminative ability (AUC values = 0.714, 95%CI: 0.641–0.786), with satisfactory calibration and positive net benefit on DCA across clinically relevant threshold probabilities.

We developed a pragmatic nomogram integrating readily available blood biomarkers to predict 3-week functional outcomes in stroke patients. While validation in larger cohorts is warranted, our findings provide new evidence in early risk stratification and personalized rehabilitation planning, potentially improving post-stroke care efficiency.