Predictors of mortality in patients with severe sepsis or septic shock in the ICU of a public teaching hospital

Background

Sepsis is a complex and multifactorial syndrome, whose incidence, morbidity and mortality have been increasing worldwide. The knowledge of clinical, epidemiological and hemodynamic parameters responsible for its evolution, diagnosis and treatment are still the subject of many studies. Therefore, this study aims to evaluate clinical, laboratory and hemodynamic parameters of morbidity and mortality in patients with severe sepsis and septic shock.

Materials and methods

As the diagnostic criteria of the systemic inflammatory response syndrome (SIRS) are very sensitive and very little specific, we selected patients with severe sepsis and septic shock in the first 24 hours of ICU admission, 18 years old or more, with two general and one or more inflammatory criteria of SIRS (ACCP/SCCM/2003). Patients with pathologies that could confound clinical and laboratory evaluations and advanced comorbidities or on immunosuppressive drug therapy were excluded. The ICU had 35 beds, five of them are resuscitation beds located in the emergency room (ER). The same intensivist team assists all patients in the ER and during ICU permanence. The principal investigator did not perform any orientation or intervention in the treatment of selected patients. Clinical (age, sex, infection focus, fluid balance, hemodialysis, use of corticosteroids, antibiotic therapy, APACHE II, SOFA), laboratory (blood cell counting, lactate, creatinine, bilirubin, glucose, cortisol, NT-proBNP, C-reactive protein (CRP), procalcitonin (PCT), Troponin I), hemodynamic (blood pressure, heart rate, left ventricular systolic function (echocardiography)) and respiratory parameters (respiratory rate, PaO₂/FiO₂), PEEP and peak inspiratory pressure (PIP)) were analyzed from ICU admission until discharge or death. Echocardiography was performed at 48 hours and on the 10th day after ICU admission.

Results

Seventy-two patients (64% male), mean age 52 ± 19 years, were consecutively included, 21% (15/72) with severe sepsis and 79% (57/72) with septic shock. Mortality was 18% (13/72), of these 21% (3/13) for severe sepsis and 79% (10/13) for septic shock. Median APACHE II score was 28 (16 to 37) and SOFA score 6 (5 to 10) (Table 1). There was positive correlation between mortality with: male gender, APACHE II, SOFA, positive 24-hour fluid balance, hemodialysis indication, corticosteroid use, leukopenia, lactate, NT-proBNP and PCT levels (Table 2). From univariate analysis, practically the same significant association with mortality was observed (Table 3). In addition, the final multivariate Cox model showed that male gender, hypotension (first 24 hours), leukopenia and positive fluid balance (first 24 hours) had an impact on mortality (Table 4). Glycemic control and early antibiotic use were not relevant.

Table 1 Demographic and clinical characteristics observed in patients with severe sepsis and septic shock

Table 2 Clinical and laboratory variables related to mortality in patients with severe sepsis and septic shock

Table 3 Univariate analysis of variables associated with mortality in patients with severe sepsis and septic shock

Table 4 Multivariate analysis of variables associated with mortality in patients with severe sepsis and septic shock

Conclusions

Precocious treatment, judicious fluid management and individualized care showed benefit in the treatment of patients with severe sepsis, septic shock.