Evaluation of the neutrophil-to-lymphocyte and platelet-to-lymphocyte ratios in critically ill dogs

Lisa H Dourmashkin,Bridget Lyons,Rebecka S Hess,Koranda Walsh,Deborah C Silverstein
DOI: https://doi.org/10.1111/vec.13269
Abstract:Objective: To evaluate whether the neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) are accurate prognostic indicators and correlate with illness severity scores in critically ill dogs. Design: Prospective observational study from December 2016 to May 2017. Setting: ICU at a veterinary teaching hospital. Animals: Seventy-two client-owned dogs admitted to the ICU with CBCs and abbreviated and complete acute patient physiologic and laboratory evaluation (APPLEfast and APPLEfull ) scores were enrolled. Interventions: None. Measurements and main results: The NLR, PLR, APPLEfast , and APPLEfull scores were calculated for each patient on the day of admission. Patients were followed from admission to discharge, and diagnosis, survival, and length of hospitalization were recorded. The patient population was assessed as a whole and as subcategories of patients with neoplastic disease, infectious disease, sepsis, and severe hemorrhage. Dogs with nonseptic disease processes (n = 52) that died had a significantly higher median PLR (P = 0.04) of 441 (range: 106-986) compared to those that survived with a median PLR of 217 (range: 28.4-3225). The PLR was strongly predictive of ICU length of stay in dogs with severe hemorrhage (P = 0.03, Spearman's rho = 0.84). The NLR had a poor positive correlation with APPLEfull score (P = 0.04, Spearman's rho = 0.24), and PLR had a poor negative correlation with APPLEfast score (P = 0.02, Spearman's rho = -0.27). Conclusions: The PLR correlated with ICU length of stay for patients with severe hemorrhage and with survival for patients with nonseptic disease processes. The PLR and NLR correlated with illness severity as measured by APPLE scores. Future studies with larger sample sizes are warranted to further determine the merit of NLR and PLR as indicators of morbidity, mortality, and illness severity.
What problem does this paper attempt to address?