Can the leukocyte`s parameters in peripheral blood smear predict risk of in-hospital death of patients undergoing high-risk gastrointestinal surgery?

Pol Przegl Chir. 2018 Dec 10;91(1):22-28. doi: 10.5604/01.3001.0012.7792.

Abstract

Introduction: White blood cell (WBC) count constitutes a part of routine peripheral blood examination (FBC, full blood count). Precise analysis of leukocytes' parameters in blood smear is usually performed only when leukopenia or leukocytosis is found. We aimed to assess the usefulness of leukocytes' smear test in predicting in-hospital death of patients undergoing high-risk gastrointestinal (GI) surgery. Materials & Methods. We prospectively enrolled 101 subjects undergoing high-risk GI surgery from 01.01.2017 till 31.12.2017. Blood tests were performed preoperatively, with analysis focused on the assessment of total WBC count, as well as the count of neutrophils (NEUT), lymphocytes (LYM), monocytes (MONO), eosinophils (EOS) and basophils (BASO). Indices of NEUT/LYM (NLR) and MONO/LYM (MLR) were calculated. In-hospital mortality was considered the outcome.

Results: Mortality reached 5%. There was no significant difference in WBC count between survivors and the deceased (p=0.2) and WBC failed to predict the outcome (AUC=0.69; p=0.3). MONO (p=0.009) and BASO (p=0.02) counts, as well as MLR (p=0.007) were significantly higher in patients who died. MONO count and MLR index predicted in-hospital death with good accuracy, respectively: AUC[MONO]=0.85 (p<0.001) and AUC[MLR]=0.86 (p<0.001). Other investigated parameters played no significant role in outcome prediction.

Conclusion: Routine peripheral blood smear evaluation should be considered in all patients undergoing high-risk GI surgery, because the number of monocytes can be a valuable predictor of in-hospital death.

Keywords: Peripheral blood morphology; monocytes; perioperative medicine; risk.

MeSH terms

  • Aged
  • Digestive System Surgical Procedures / adverse effects*
  • Female
  • Hospital Mortality*
  • Humans
  • Leukocyte Count*
  • Male
  • Middle Aged
  • Predictive Value of Tests*
  • Risk Assessment / methods*