# Frailty Index-laboratory and lymphocyte subset patterns in predicting 28-day mortality among elderly sepsis patients: a multicenter observational cohort study

## TL;DR

This study finds that a lab-based frailty index and lymphocyte patterns can predict 28-day mortality in elderly sepsis patients.

## Contribution

The study introduces a new predictive model combining a lab-based frailty index and lymphocyte trajectory patterns for mortality risk in elderly sepsis patients.

## Key findings

- High FI-lab risk is independently associated with increased 28-day mortality in elderly sepsis patients.
- Lymphocyte subset counts, especially natural killer cells, are significantly lower in non-survivors.
- A combined model of FI-lab, APACHE-II scores, and lymphocyte data shows strong predictive accuracy (AUC=0.788).

## Abstract

Frailty is associated with poor outcomes in elderly sepsis patients. This study investigated the relationship between Frailty Index-laboratory (FI-lab) and lymphocyte patterns in predicting 28-day mortality among elderly sepsis patients.

We conducted a multicenter prospective observational study in four tertiary hospitals in Beijing, China. FI-lab was calculated using 24 laboratory parameters. Peripheral blood lymphocyte subsets were measured at ICU admission. Lymphocyte count trajectories were classified into four phenotypes based on patterns during the first 72 hours. The primary outcome was 28-day mortality.

Among 1,197 patients (mean age 74.6 ± 7.4 years), those with high FI-lab risk showed higher mortality (22.2%) than intermediate (12.0%) and low-risk groups (6.1%). Age-stratified analysis demonstrated consistent FI-lab prognostic value in both 65–79 years (OR 2.18) and ≥80 years (OR 2.47) groups. All lymphocyte subset counts were lower in non-survivors, particularly natural killer cells. In multivariable analysis, high FI-lab risk (OR 2.31), APACHE-II scores (OR 1.08), heart rate (OR 1.01), NK cell count (OR 0.994), and pulmonary infection (OR 1.96) independently predicted 28-day mortality. A combined model incorporating these variables showed superior discriminative ability (AUC=0.788) with excellent internal validation (optimism-corrected AUC=0.775).

FI-lab independently predicts mortality in elderly sepsis patients and correlates with lymphocyte abnormalities. When comprehensive immune assessment is unavailable, lymphocyte trajectory patterns offer a practical approach for risk stratification.