Evaluating the Utility of Initial Prehospital Shock Index and Modified Shock Index to Predict Hospital Sepsis and Septic Shock Diagnosis

Introduction: Early recognition and initiation of treatment for suspected sepsis patients is important to reduce morbidity and mortality. Shock index (SI) and modified shock index (MSI) can be readily obtained in the prehospital setting and may be useful predictors of sepsis; however, their predictive value for sepsis and septic shock has not been tested for use by EMS. 

Objective: To evaluate SI and MSI in the prehospital setting as predictors of hospital sepsis diagnosis. 

Methods: This retrospective analysis used linked prehospital and hospital patient care records for January 1, 2018–December 31, 2018 from the large national health data exchange research database maintained by ESO. All medical emergency EMS responses with linked hospital data were included. Patients under 18 and those with traumatic injuries were excluded. Any sepsis diagnosis was defined by hospital ICD-10 codes A40, A41, R65.20, and R65.21, and septic shock was restricted to R65.21 only. SI was calculated by dividing initial heart rate by systolic blood pressure. MSI was calculated by dividing initial heart rate by mean arterial pressure. Based on previous research, elevated SI was defined as greater than 1.0, and elevated MSI was greater than 1.3. Four multivariable logistic regression models were created to separately evaluate SI and MSI as predictors of sepsis and septic shock, controlling for patient age, gender, race/ethnicity, community size, and census region. 

Results: This analysis included 325,558 patients. The median age was 60 (IQR 43–75), 53% were female, 69% were white (non-Hispanic), and 94% were in urban communities. Five percent (16,881) were diagnosed with sepsis. Of these, 19% had septic shock (3,144). Over a fivefold increase in odds of any sepsis diagnosis was noted for patients with an elevated initial SI (AOR 5.30, 95% CI, 5.08–5.52) or MSI (AOR 5.42, 95% CI, 5.21–5.64). Approximately a ninefold increase in odds of septic shock was observed for patients with an elevated SI (AOR 9.05, 95% CI, 8.32–9.85) or MSI (AOR 8.61, 95% CI, 7.92–9.36). 

Conclusion: Our findings suggest elevated initial prehospital SI and MSI are both strong predictors of hospital sepsis diagnosis and stronger predictors of septic shock.