Prognostic indicators related to risk of death in shock patients: A new simplified score

Abstract

Introduction: We analysed the parameters associated with mortality outcome in shock patients. Methods: The databases of intensive care unit patients were retrieved, and shock patients were selected for further analysis. Logistic regression was used to identify the predictors of mortality outcome. The area under curve (AUC) of receiver operating characteristic (ROC) curve was calculated for the power of prediction model. Results: A total of 467 patients were recruited, of which 183 patients were diagnosed with shock. The variables predicting mortality outcomes were heart rate above 130 beats/minute (p-value is 0.015, odds ratio [OR] 4.38, 95 percent confidence interval [CI] 1.338-14.321), pH less than or equal to 7.24 (p-value is 0.001, OR 6.11, 95 percent CI 2.17-17.18), creatinine more than 1.5 mg/dl (p-value is 0.048, OR 3.05, 95 percent CI 1.01-9.19), and Glasgow Coma Score less than 7 (p-value is 0.038, OR 3.476, 95 percent CI 1.07-11.27). The sensitivity, specificity and AUC of ROC of this model was 82.5, 60.5 and 0.826 percent, respectively. The positive and negative predictive values and AUC of ROC at a score below 2 was 82.8, 67.3 and 0.81 percent, respectively. The results revealed a significant improvement in survival in shock patients with a score below 2 (p-value less than 0.001). We prospectively validated the score in 107 shock patients and found very high AUC of ROC. Conclusion: Acidosis, tachycardia, renal impairment and impaired consciousness within the first 24 hours are the main predictors of shock state, and should be used for assessment of survival outcome in shock patients.