Patients who arrive at the emergency department with an acute ischemic stroke and carry low values of a simple blood-derived index, the hemoglobin-albumin-lymphocyte-platelet (HALP) score, died more often within the following year than patients with higher values, according to a retrospective cohort study published in BMC Neurology. Of 735 adults treated at a single Turkish center during 2019, 185, or 25.2 percent, had died by one year, and the median HALP score among those who died was 27.16 compared with 42.73 among survivors, a difference the authors report as statistically significant.
The HALP score is calculated from four routine laboratory measurements: hemoglobin in grams per liter, multiplied by albumin in grams per liter, multiplied by the lymphocyte count in units of 10⁹ per liter, and divided by the platelet count in the same units. In this study the index was computed from the first blood samples drawn at emergency department presentation, meaning it reflects the patients’ condition at the moment of arrival rather than values measured later in the hospital stay. Because all four components are part of standard stroke workups, the score can be derived without additional testing or cost.
The cohort was assembled from 1,202 screened records of adults with a final discharge diagnosis of acute ischemic stroke who presented between 1 January and 31 December 2019. One-year vital status was ascertained using e-Nabız, Türkiye’s national personal health record system, which allowed the researchers to link each patient’s emergency presentation to survival information collected after discharge. The study design was retrospective, so the analysis relied on data already recorded during routine care.
To test whether the association between HALP and death was independent of other patient characteristics, the researchers used multivariable logistic regression with adjustment for age, sex, and selected comorbidities. After this adjustment, each doubling of the HALP score was associated with lower odds of death within one year, with an adjusted odds ratio of 0.479 and a 95 percent confidence interval of 0.385 to 0.596. In other words, patients whose HALP value was twice as high had roughly half the odds of one-year mortality compared with otherwise similar patients in the model, and the confidence interval did not include the null value of one.
The authors also examined how well the score distinguished patients who died from those who survived using receiver operating characteristic curve analysis. The area under the curve was 0.704, with a 95 percent confidence interval of 0.659 to 0.750. This corresponds to moderate discriminatory performance, clearly better than the 0.5 expected from chance assignment but well short of the near-perfect separation that would make a single laboratory index sufficient for individual risk assessment on its own.
An exploratory cut-off was selected by maximizing the Youden index, a standard method that balances sensitivity against specificity. Using the unrounded threshold of approximately 25.73, the score detected 49.2 percent of the patients who ultimately died, meaning just under half of the deaths occurred above this value, while it correctly classified 84.0 percent of survivors as being below it. The authors emphasize that this threshold is exploratory and was calculated using the unrounded value, so it should not be read as a validated clinical decision limit.
The study’s conclusion is deliberately cautious: lower HALP scores remained associated with higher one-year all-cause mortality after adjustment for age, sex, and selected comorbidities, but discriminatory performance was moderate and external validation is required before any clinical application. The authors frame the score as a candidate prognostic marker rather than a ready-to-use triage tool, and they do not claim that the HALP score itself modifies disease course or that correcting any of its components would change outcomes.
Several limitations follow directly from the design. The study was conducted at a single center, so the patient mix, laboratory methods, and follow-up infrastructure may differ elsewhere, which is precisely why the authors call for external validation. The analysis adjusted for age, sex, and selected comorbidities but not for stroke severity measures or acute treatments, so residual confounding cannot be excluded. As a retrospective cohort, the study can document associations between a marker measured at presentation and later mortality, but it cannot establish that the marker causes or predicts death independently of unmeasured factors. The reported odds ratios describe the association observed in this specific cohort of emergency department patients with acute ischemic stroke during 2019.
Within those boundaries, the findings add a low-cost, routinely available composite index to the set of variables that researchers studying post-stroke mortality can evaluate. The HALP score combines information about oxygen-carrying capacity, nutritional and hepatic status through albumin, immune cell levels through lymphocytes, and platelet count into a single number, and its association with one-year all-cause mortality in this cohort was consistent across both the unadjusted median comparison and the adjusted regression model. Whether that association holds in other health systems, in patients treated with newer reperfusion therapies, and against established severity scores remains a question for the validation studies the authors say are needed before the score could be applied clinically.
Subject of Research: Association between the HALP blood score and one-year mortality in emergency department patients with acute ischemic stroke
Article Title: Association of the HALP score with one-year mortality in emergency department patients with acute ischemic stroke: a retrospective cohort study
Article References: Avcı, M. A., Cander, B., Fettahoğlu, S., Fettahoğlu, S. T., Ak, E., & Kalafat, U. M. (2026). Association of the HALP score with one-year mortality in emergency department patients with acute ischemic stroke: a retrospective cohort study. BMC Neurology. https://doi.org/10.1186/s12883-026-05588-9
Image Credits: AI Generated
DOI: 10.1186/s12883-026-05588-9
Keywords: HALP score, acute ischemic stroke, one-year mortality, emergency department, retrospective cohort, prognostic marker, hemoglobin, albumin, lymphocyte count, platelet count, logistic regression, BMC Neurology
News Source: Cassandra Pierce. (November 9, 2026). Blood-based HALP score flags one-year mortality risk after acute ischemic stroke. Scienmag.



