Analysis of the effectiveness of the SIRS, qSOFA and MEWS scale for the sepsis identification in oncohematological patients. Gematologiya i transfuziologiya

Abstract

Introduction. Sepsis is one of the main causes of high mortality in oncohematological patients. Sepsis is diagnosed using different scoring scales, but the accuracy of the diagnosis varies.Aim — to determine the effectiveness of SIRS, qSOFA and MEWS scales for the diagnosis of sepsis in oncohematological patients.Materials and methods. Study participants included 202 patients hospitalized in the intensive care unit with infectious complications — 112 (55 %) male, 90 (45 %) female. The median age was 57 years. The estimated predictors were the criteria of SIRS, qSOFA and MEWS scales and other significant indicators. The Boruta method was employed to select the predictors. A multiple binomial logistic regression model was used to determine the prognostic value of the predictors. The probability of sepsis was calculated based on the regression equation. The correlation between sepsis probability and the prognostic scales was assessed using correlation analysis (Kendall rank correlation coefficient).Results. Diagnoses were confirmed by scales in 95 of 202 (47 %) patients: SIRS — in 77 (81 %), qSOFA — in 31 (33 %), MEWS — in 65 (68 %). Of the 19 predictors initially included in the study, the final logistic regression model included 6: respiratory rate, heart rate, impaired urine output, systolic blood pressure, body temperature, and Glasgow coma scale level, which matched MEWS criteria. The median probability of sepsis was 0.38 (0.079–0.921). The results of the calculated probability of sepsis, according to the logistic regression model, correlated most closely with the score on the MEWS scale, to a lesser extent — with that on SIRS and qSOFA.Conclusion. The MEWS scale is a more suitable tool for the diagnosis of sepsis than SIRS and sofa in oncohematological patients.