Simple model for prediction of mortality in elderly patients with septic shock

Introduction

The number of elderly patients with septic shock has increased in past years, making it important to investigate the existence of prognostic factors in this population. This study aimed to develop a simple prognostic model of mortality in elderly patients with septic shock hospitalized in a medical ICU.

Methods

A prospective cohort of 67 elderly consecutive patients (age over 65 years), within a 32-month period, having their pulmonary artery monitored due to septic shock. The logistic regression model employed death as the dependent variable, and as independent variables those with P values highly significant in the univariate analysis: APACHE II score, the presence of previous cardiovascular diseases (systemic arterial hypertension, diabetes mellitus, stroke and chronic coronary insufficiency), positive troponin I, the need for noradrenalin in doses from 0.5 μg/kg/min and above and the number of organic failures according to Le Gall's criteria. Variables were dichotomized and added one by one to the model. Only APACHE II score and previous cardiovascular diseases remained in the end. The statistical package SPSS 10 was employed with a significance level of 5%.

Results

The patients presenting previous cardiovascular disease and APACHE II score from 20 and above had 85% probability of death, and those presenting previous cardiovascular disease and APACHE II score below 20 had 74% probability of death during hospitalization in the ICU. Those with APACHE II score below 20 and with no previous cardiovascular disease had only 5% probability of death.

Conclusion

The presence of previous cardiovascular disease should be included in the models to predict the probability of death in elderly patients with septic shock. This simple model shall be expanded and applied to a similar cohort for validation thereof.