Systemic symptoms predict presence or development of severe sepsis and septic shock

Abstract

Background: Severe sepsis is a major cause of mortality and morbidity globally. As the time to adequate treatment is directly linked to outcome, early recognition is of critical importance. Early, accessible markers for severe sepsis are desirable. The systemic inflammatory response in sepsis leads to changes in vital signs and biomarkers and to symptoms unrelated to the focus of infection. This study investigated whether the occurrence of any of six systemic symptoms could predict severe sepsis in a cohort of patients admitted to hospital for suspected bacterial infections. Methods: A retrospective, consecutive study was conducted. All adult patients admitted during 1 month to a 550-bed secondary care hospital in western Sweden and given intravenous antibiotics for suspected community-acquired infection were included (n = 289). Symptoms (fever/chills, muscle weakness, localised pain, dyspnea, altered mental status and gastrointestinal symptoms) were registered along with age, sex, vital signs and laboratory values. Patients who fulfilled criteria of severe sepsis within 48 h were compared with patients who did not. Odds ratios for severe sepsis were calculated, adjusted for age, sex and comorbidities. Results: Criteria for severe sepsis were fulfilled by 90/289 patients (31.1%). Altered mental status (OR = 4.29, 95% CI = 2.03–9.08), dyspnea (OR = 2.92, 95% CI = 1.69–5.02), gastrointestinal symptoms (OR = 2.31, 95% CI = 1.14–4.69) and muscle weakness (OR = 2.24, 95% CI = 1.06–4.75) were more common in patients who had or later developed severe sepsis. Conclusions: Systemic symptoms in combination with other signs of infection should be considered warning signs of severe sepsis.

Keywords:

• Sepsis
• severe sepsis
• septic shock
• symptoms
• altered mental status
• dyspnea
• gastrointestinal symptoms
• muscle weakness

Acknowledgements

The authors would like to thank Mattias Molin, biostatistician at Statistiska kontrollgruppen, Göteborg, Sweden, for valuable assistance with statistical methods.

Declaration of interest

The study was funded by the Health & Medical Care Committee of the Region Västra Götaland. Expenses for statistical consulting were funded by the Research and development council of the county Södra Älvsborg. There are no commercial relationships or conflicts of interest to declare.