Sepsis accounts for nearly 37,000 deaths per year, 100,000 hospital admissions and costs the National Health Service (NHS) approximately £2 billion per year. Mortality rates are close to 41.1% in Europe and 28.3% in the United States but adherence to sepsis performance bundles reduces the relative risk in mortality rate by 25%. As majority of these patients are initially treated with sepsis performance bundles in the emergency department, our aim is to follow up their immediate and long-term outcomes.
The primary objective was to identify the mortality rate of patients admitted via the ED with severe sepsis and septic shock during initial admission and eight months later. The secondary objectives were to establish the most common source of sepsis amongst this patient cohort and the use of sepsis performance bundles.
A retrospective observational study was conducted. Patients with sepsis were identified between June 2016 and August 2016 using their case notes. Inclusion criteria were patients aged 18 years or over with sepsis with or without evidence of tissue hypoperfusion (high lactate), organ dysfunction or hypotension despite adequate fluid resuscitation. The cause of death, length of stay and time to death were analyzed. Intention to treat analysis was applied to the use of sepsis six bundles - full set of recorded observations, intravenous fluid and oxygen administration, prescription of antibiotics within one hour of admission and serum lactate estimation, obtaining blood cultures and fluid balance documentation within 4 hours of arrival.
During the study period 7930 cases were identified by hospital coding. 103 consecutive patients were studied. The mean age was 72 years (21-100 years). Among the 103 patients studied, 18% died during their initial admission, but subsequent follow up showed a mortality of 46% by the end of the first year.
The median age for mortality group was 81.5 years (50-100 years).
There was no significant gender difference with 24 females and 22 males. All but 2 patients had 3 or more co-morbidities. Average length of stay was 11.2 days (95% CI 9.2-13.3). Mean time from initial admission to death was 35.4 days (95% CI 13.2-53.5). Main sources of sepsis in the group who died included chest (41%), urine (26%), abdominal (15%) and unknown (15%). 48% had 3 or more co-morbidities. The most common co-morbidities included malignancy (43%), hypertension (30%), diabetes (28%) and lung disease (22%).
The overall adherence to the sepsis six bundle as a whole varied for different parameters. Amongst the parameters measured, 93% of patients received antibiotics, 83% had fluids initiated, 82% had blood cultures drawn, and 85% had a serum lactate measured. 39% of patients had their observations documented, 37% a senior doctor review, 40% had oxygen administered and 49% had fluid balance documented.
Severe sepsis is a frequent reason for presentation to the ED. This requires a clear, timely and co-ordinated response from a multi-disciplinary team. Our study is possibly only the second study which looked at long term mortality related to sepsis. One previous study reported a mortality of 37% at the end of 1-year period after initiation of sepsis performance bundles. However, the mean age was 58 years in the previous study compared to our study where the mean age was 72 years. Also, the median age in the mortality group in our study was 81.5 years and patients had significant co-morbidities. Although ours is a smaller study, the short-term outcomes of early initiation of sepsis bundles are associated with good prognosis (18% mortality). This is not however reflected in long term mortality which remains significantly high at 46% especially in the elderly and is often associated multiple co-morbidities.