Effect of sepsis protocol in inpatient departments triggered by Ramathibodi Early Warning Score (REWS) on treatment processes

Abstract

Background: Sepsis needs to be more focused on the effect of patient management at the ward level. We aimed to evaluate the effect of implementing the sepsis protocol triggered by the Ramathibodi Early Warning Score (REWS) on treatment processes in inpatients with new-onset sepsis. Methods: We conducted a prospective observational study among adult medical patients admitted to the general wards. A 25-month pre-protocol period was assigned as a control, and a 14-month protocol period was allocated to a protocol group. An inpatient sepsis protocol comprised a nurse-initiated sepsis protocol with REWS ≥2 plus suspected infection, prompt antibiotic, lactate measurement, and fluid resuscitation. Primary outcomes were the achievement of sepsis treatment processes, including the resuscitation and management bundle, namely: 1) the percentage of patients who were taken for the initial laboratory workup for sepsis, especially lactate and blood culture taking before antibiotics; 2) the percentage of patients who received appropriate antibiotics; 3) the percentage of patients who received optimal fluid resuscitation and management; 4) the percentage of patients who performed inferior vena cava ultrasound; 5) the percentage of patients who received steroid and vasopressor drugs; 6) "time-to-antibiotic," the duration from diagnosis of sepsis to receiving antibiotic treatment; 7) "time-to-optimal intravenous fluid management;" 8) "time-to-transfer to ICU. Results: 282 patients were evaluated (141 pre-implementation, 141 post-implementation); 94.7% of patients with sepsis had REWS ≥2. More patients in the protocol period had a lactate measurement and fluid management (89 [63.1%] vs. 44 patients [31.2%], p<0.001 and (50 [35.4%] vs. 22 patients [15.6%], p<0.001, respectively). More patients in the protocol period received antibiotics within 1 hour than in the pre-protocol period (80 [56.7%] vs. 53 patients [37.6%], p=0.001). The time to antibiotic treatment (mean, SD) in the protocol period was shorter than that in the pre-protocol period (81.7 [77.86] vs. 138.22 [145.17], p=0.007). The length of the intensive care unit (ICU) stay was shorter in the protocol period (8 d [3, 16.5] vs. 10 d [5, 20.5], p=0.011). The two groups did not differ in in-hospital mortality, length of hospital stay, or time-to-transfer to the ICU. Conclusions: Implementing an in-hospital sepsis protocol was associated with significant improvement in sepsis treatment processes, namely lactate measurement, starting antibiotic treatment within 1 hour, fluid management, and a shorter length of ICU stay.