CHICAGO – A high neutrophil to lymphocyte ratio on days 2 and 5 of surgical ICU hospitalization independently predicts increased mortality in critically ill trauma patients, an award-winning study showed.
The risk of death was two times higher for patients with a neutrophil to lymphocyte ratio (NLR) of at least 10.45 on day 2 (adjusted hazard ratio, 2.07; 95% confidence interval, 1.38-3.13; P = .001) and 5.7 times higher for those with an NLR of at least 7.91 on day 5 (adjusted HR, 5.79; 95% CI, 2.93-11.44; P less than .001) in a multivariate analysis, after adjustment for age 65 years or older, male sex, a systolic blood pressure of 90 mm Hg or less, a Glasgow Coma Scale (GCS) score of 8 or less, an Injury Severity Score (ISS) of at least 25, and operation on admission.
“The neutrophil to lymphocyte ratio is easily accessible, the calculation is simple, it adds no additional costs, and virtually all critically ill patients will have these labs,” study author Dr. Evren Dilektasli said at the American College of Surgeons annual clinical congress.
The simple calculation has been shown to be useful in the diagnosis of appendicitis and to be associated with overall survivalin metastatic colorectal cancer, but its association with mortality in trauma patients is not known, he said.
The retrospective cohort comprised 1,356 trauma patients, at least 16 years old, admitted to the Los Angeles County–University of Southern California Medical Center surgical ICU between January 2013 and January 2014. The median NLR was calculated for each day of the surgical ICU stay. At baseline, 16% of patients had an ISS of at least 25, 16.5% had a GCS of 8 or less, 4.5% had a systolic BP of 90 mm Hg or less, 74.3% were male, 23.7% were aged 65 or older, and 86% had a blunt injury. The most common operations on admission were laparotomy (39.6%) and craniectomy/craniotomy (20.7%).
In receiver operating characteristic (ROC) analysis for the first 10 days of hospitalization, the area under the curve (AUC) values for predicting mortality were between 0.55 on day 1 and a high of 0.79 on day 5, Dr. Dilektasli reported.
Starting from day 2 to day 10, the AUCs were statistically significant for predicting mortality.
The NLRs on day 2 (AUC, 0.73; P less than .001) and day 5 (AUC, 0.79; P less than .001) were selected in order to adjust for the clinical probability of early and late complications, he said.
Subsequent ROC curve analysis revealed an NLR cutoff of 10.45 on day 2 (AUC, 0.73; sensitivity, 73.2%; specificity, 61.8%) and a cutoff of 7.91 on day 5 (AUC, 0.79; sensitivity, 82.8%; specificity, 65.2%).
A high NLR on day 2 (at least 10.45) versus a low NLR (less than 10.45) was associated with significantly more ventilator days (5 days vs. 3 days), a longer surgical ICU length of stay (5 days vs. 3 days), a longer hospital stay (11 days vs. 8 days), and greater mortality (18.3% vs. 4.8%; all P values less than .001), reported Dr. Dilektasli, who was a research fellow at USC at the time of the study and has returned to Turkey to continue his training.
On day 5, a high NLR (at least 7.91) versus a low NLR (less than 7.91) was associated with significantly more ventilator days (7 days vs. 4 days; P less than .001), a longer surgical ICU stay (9 days vs. 5 days; P less than .001), and increased mortality (20.4% vs. 2.8%; P less than .001), but not a longer hospital stay (17 days vs. 14 days; P = .119).
In Kaplan-Meier analysis, a significant difference was observed between the high and low NLR groups on day 2 (log rank P less than .001) and day 5 (log rank P less than .001), he said.
“NLR may be a promising tool for assessing the risk of in-hospital mortality,” Dr. Dilektasli concluded. “Prospective external validation is warranted in a larger heterogeneous trauma population.”
During a discussion of the study, it was noted that the ROC curves were impressive, but that other biologic markers known to be associated with poor survival such as C-reactive protein level and class II major hepatitis C expression should have been included in the analysis.
When asked whether any patients with a low NLR on day 2 went on to have a high NLR on day 5, Dr. Dilektasli said there were such patients and that they also had an increased risk of death.