Recognition, response and outcomes of sepsis: A dual site retrospective observational study.

OBJECTIVES

To describe clinical recognition, response and outcomes of patients with sepsis.

METHODS

A retrospective, observational study was undertaken at two hospitals. Inclusion criteria were: adult patients admitted via the Emergency Department (ED) between 1 January and 30 April 2014 allocated a primary ICD-10-AM discharge from hospital code related to sepsis. Recognition of sepsis was considered based on the presence of clinical documentation that reflects the Sepsis Kills criteria being met. Response to sepsis was considered based on the presence of clinical documentation where the patient received a response consistent with the 'Sepsis Six' strategies. Outcomes pertained to response to sepsis (e.g. time to antibiotics) and ED measures (e.g. time to be seen, ED length of stay). Sub-group analysis considered location where sepsis was recognised (ED/ward).

RESULT

In total, 96 patients met the inclusion criteria; most were admitted under general medicine (37%) followed by intensive care (18%). Sepsis was recognised in the ED for most patients (n = 64), with a history of fevers/rigors the most common (60%) indication of infection. Regarding response and ED outcomes for this group, the median time from triage nurse assessment i) to being seen by the treating clinician was 19 min; ii) to sepsis recognition was 27 min; and iii) to antibiotics was 181 min; 35% received antibiotics within 60 min from recognition. Those recognised in the ED had a longer ED stay than those where sepsis was recognised on the ward (336 min vs. 225 min, p = 0.013).

CONCLUSIONS

Sepsis can develop at various stages throughout the patient's journey. In this small sample, ED recognition was associated with longer ED stay, likely due to more interventions. Whilst guidelines recommend antibiotics be administered within 60 min of triage, this was not achieved for most patients. Given the dynamic nature of sepsis, future indicators may focus on time from recognition rather than time from triage.