Simple Blood Ratio Predicts Death Risk in Cirrhosis Patients With Pneumonia

A routine blood test that costs almost nothing to perform may help doctors identify which patients with cirrhosis and pneumonia are most likely to die during their hospital stay, according to a new multicenter study from China. The research, published in Frontiers in Medicine, found that a simple ratio derived from standard blood counts—the neutrophil-to-lymphocyte ratio, or NLR—was independently associated with in-hospital mortality in this high-risk population, and that adding it to established severity scores significantly improved their ability to distinguish survivors from non-survivors.

The study analyzed data from 851 patients treated at five Chinese hospitals between 2016 and 2025. All had been diagnosed with both cirrhosis—the end-stage scarring of liver tissue that affects roughly 112 million people worldwide in its compensated form—and pneumonia, one of the most dangerous infections this group can face. Among the cohort, 14.9 percent died before discharge. The researchers focused on the NLR calculated at admission: the number of neutrophils, the frontline cells of the innate immune system, divided by the number of lymphocytes, the adaptive immune cells that orchestrate targeted responses. Because raw NLR values are heavily right-skewed, the team used the base-10 logarithm of the ratio in their statistical models.

The biological rationale is straightforward. When pneumonia strikes, the body floods the bloodstream with neutrophils to fight the infection, while inflammatory signaling molecules such as tumor necrosis factor-alpha and interleukin-6 accelerate the death of lymphocytes. Patients with cirrhosis start from a compromised position: their immune systems are already dysfunctional, marked by chronic inflammation, impaired splenic filtering of blood cells, and a pre-existing imbalance between neutrophil and lymphocyte populations. A pneumonia-triggered inflammatory surge therefore pushes the ratio even higher, making it a potential barometer of both infection severity and underlying immune collapse.

The numbers were striking. In the fully adjusted statistical model—which accounted for age, sex, comorbidities such as coronary artery disease, hypertension and diabetes, hepatitis type, and laboratory markers including blood urea nitrogen, white cell count, albumin and international normalized ratio—each tenfold increase in NLR was associated with 3.82-fold higher odds of in-hospital death. When patients were divided into three groups by NLR level, mortality climbed in lockstep: 5.3 percent in the lowest tertile, 15.2 percent in the middle, and 24.3 percent in the highest. Restricted cubic spline analysis confirmed the relationship was linear, with no statistically significant departure from a straight-line dose–response pattern.

What makes the finding clinically interesting is how the ratio performed against the established yardsticks. The Child-Pugh score, which grades liver failure severity using five measures including bilirubin, albumin, and the presence of ascites and encephalopathy, achieved an area under the receiver operating characteristic curve (AUC) of 0.647. The CURB-65 score, a five-point pneumonia severity tool, reached 0.658. The log-transformed NLR alone scored 0.707—numerically higher than either, though the differences did not reach statistical significance under the DeLong test used for pairwise comparison.

The real payoff came from combining the markers. When NLR was added to Child-Pugh, the AUC jumped from 0.647 to 0.737, an increase of 0.090 that was highly significant. Added to CURB-65, it rose from 0.658 to 0.734, a gain of 0.076. Calibration curves showed acceptable agreement between predicted and observed risks, and decision-curve analysis indicated that models incorporating the ratio delivered greater net benefit across clinically relevant threshold probabilities, roughly between 10 and 40 percent. In practical terms, this means clinicians using the combined scores would make better-informed decisions about escalation of care across a wide range of risk tolerances.

The researchers also identified an exploratory cutoff. Using the Youden index to maximize the sum of sensitivity and specificity, an NLR of approximately 5 emerged as the threshold, correctly flagging 74 percent of the patients who died while classifying 57.6 percent of survivors correctly. The negative predictive value was 92.7 percent, meaning a low ratio was a fairly reliable signal that a patient would survive the admission. The team stressed, however, that this cutoff was derived from the same cohort it was tested on and requires independent validation before it can guide clinical decisions.

Robustness checks supported the core result. A complete-case analysis restricted to the 392 patients with no missing data produced an even stronger association. A mixed-effects model with a random intercept for each hospital—necessary because baseline mortality risk varied noticeably between centers—still yielded an adjusted odds ratio of 3.82. Additional adjustments for calendar year, given the decade-long study window, left the findings essentially unchanged. Subgroup analyses across age, sex, comorbidity status, severity scores, ICU admission, and study period found no significant interactions, suggesting the ratio’s prognostic value is broadly consistent across patient types.

The study has limitations worth noting. It was retrospective, so residual confounding cannot be excluded despite the extensive adjustments. Only admission values were measured; tracking how the ratio changes over the hospital stay might add further prognostic information, and levels can be influenced by stress and glucocorticoid treatment. The specific pathogen causing each pneumonia was not identified, and the analysis did not stratify by cause of cirrhosis. The endpoint was in-hospital death rather than 30-day or 90-day survival, which future work should address.

Even with those caveats, the appeal of the marker is its accessibility. Unlike composite scores that require multiple organ-system assessments or laboratory panels that may take hours to complete, the NLR falls out of any standard complete blood count—the single most common blood test performed in hospitals worldwide. For emergency departments and wards where sophisticated severity calculators are not routinely applied, an elevated neutrophil-to-lymphocyte ratio in a cirrhotic patient with pneumonia could serve as an early, zero-cost warning sign that prompts closer monitoring, earlier escalation to intensive care, or more aggressive empiric treatment. The authors call for prospective, multicenter validation, but the message of the data is clear: sometimes the immune system’s most basic arithmetic tells you who is in trouble.

Subject of Research: Association between the neutrophil-to-lymphocyte ratio and in-hospital mortality in patients with cirrhosis and pneumonia

Article Title: Neutrophil-to-lymphocyte ratio is associated with in-hospital mortality in patients with cirrhosis and pneumonia: a multicenter retrospective study

Article References: Zhang, H., Huang, L., Sun, P., Dong, Y., Gu, Y., Chen, X., Suo, S., Zhang, M., & Zhang, Y. (2026). Neutrophil-to-lymphocyte ratio is associated with in-hospital mortality in patients with cirrhosis and pneumonia: a multicenter retrospective study. Frontiers in Medicine, 13, Article 1963013. https://doi.org/10.3389/fmed.2026.1963013

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.1963013

Keywords: cirrhosis, pneumonia, neutrophil-to-lymphocyte ratio, in-hospital mortality, prognosis, Child-Pugh score, CURB-65, biomarker, immune dysfunction, risk stratification, multicenter study, liver disease