Background: Sepsis is one of the most common admission diagnoses in the US. It is currently defined as organ dysfunction caused by a dysregulated host response to infection. The quickSOFA (QSOFA) score can be used to screen patients for the likelihood of poor outcomes typical of sepsis based on the presence of at least 2/3 clinical criteria including : respiratory rate ≥ 22/min or greater, altered mentation, or systolic blood pressure ≤ 100 mm Hg . Previous studies have established the association between initial lactate values ≥4, the shock index (systolic BP/heart rate) and subsequent deterioration. While initial clinical values are often investigated in studies, in practicality, bed placement is often guided by the response to initial resuscitation. The likelihood that patients are placed in lower levels of care than may be ideal may increase as the gaps between bed availability and demand widen across the country. It becomes increasingly important therefore that patient placement be guided by simple and easily accessible clinical criteria. We sought to examine clinical predictors of deterioration among patients admitted to hospital medicine with sepsis.
Methods: We reviewed all patients on the hospital medicine team who experienced a ‘Code Blue’ or rapid response event over a one-month period in 2025. Demographic and clinical information was extracted and collated using summary statistics.
Results: 61 events were reviewed of which 59 (97%) were rapid response events. 9/61 (14.7%) patients were admitted for sepsis and of these, 6/9 (67%) were admitted through the Emergency Room (ER) within the previous week. For these 6 patients, the mean age was 74.8 years and 5/6 were female (83%). The mean first heart rate was 104 beats per minute (range 67-140), mean first systolic was 157mm Hg (range 114-191). The mean shock index was 1.6 (range 0.9-2.8) and the mean first lactate was 2 mmol/L (range 1-5)
Conclusions: In this small sample, 15% of all code blue and rapid response events were in patients who were admitted with sepsis. Among patients who were recently admitted with sepsis through the ER and who subsequently experienced a clinical decline, first measured vital signs and laboratory parameters were not indicative of increased risk of adverse outcomes. However, the mean shock index was >1 and the normal mean shock index is 0.5-0.7. These preliminary findings need to be validated in larger samples and bed placement for patients with sepsis should be examined within the current constraints of bed availability and with a focus on trajectory and not solely on first parameters. Shock index may be a valuable clinical marker to predict deterioration in patients with sepsis.