The Effects of Implementing a Nurse-driven Sepsis Protocol in the Emergency Department

Introduction

In 2002, the Surviving Sepsis Campaign released the Barcelona Declaration to raise awareness of sepsis and affect a 25% decrease in the relative morality by 2008 (Slade, Tamber, & Vincent, 2003). In 2014 sepsis remained a leading cause of death in the United States where it came in 11^th (Centers for Disease Control and Prevention, 2015).  The key to surviving sepsis is early identification and treatment, and implementing early goal directed therapy (EGDT) in the emergency department has been shown to decrease sepsis mortality rates (Jones, Focht, Horton, & Kline, 2007). Triage nurses in the emergency department are the first point of medical contact on patient arrival, and nursing assessment frequently occurs before physician assessment. This presents an opportunity for early identification of the sepsis patient. There is evidence that education of nurses and introduction of a nurse-initiated sepsis protocol can improve time to sepsis bundle completion and antibiotic administration (Bruce, Maiden, Fedullo, & Son Chae, 2015; Tromp et al., 2010). A literature review of these two studies appears below.

Literature Review

Bruce, Maiden, Fedullo, & Son Chae (2015) conducted a retrospective chart review of 192 patients, over an 8-month period between 2011-2012, discharged with a diagnosis of severe sepsis or septic shock from two Southern California academic medical centers. The purpose of this study was twofold. First to assess the effects of a nurse-initiated sepsis protocol on time to antibiotic administration, compliance with sepsis bundle goals, and inpatient mortality. The second was to identify predictors of mortality in the same patients. All the patients included in the study were admitted through the emergency department. A nursing triage tool for sepsis identification was identified and, prior to implementation, all physicians and nurses received mandatory sepsis education. All cases were retrospectively placed in one of three groups. Patients that were admitted before education and implementation of the tool were placed in the pre-protocol group. The transition group included all cases that were admitted during education and implementation. Finally, the remainder were placed in the post protocol group. Data was collected on each patient that included demographic information and clinical characteristics to compare for statistical differences in samples. There were no statistical differences between the groups except for blood pressure. Among the data collected for analysis was ED admission time, fluid volume infused, serum lactate results, blood culture results, antibiotic administration time, and in-hospital mortality. No reliability or validity data assessment is available.

Data was analyzed using either the Mann-Whitney test or the Fisher exact test. Statistical significance was set at p = .05. Serum lactate measurement had a statistically significant improvement from 83.9% pre-protocol to 98.7% post protocol (p = .003). Median time to antibiotic administration also showed a statistically significant improvement, going from 135 minutes to 108 minutes (p = .021). There was no statistical significance in the other variables. The researchers noted that there were several limitations. The study examined only patients with a discharge diagnosis of severe sepsis or septic shock therefore it is unknown the total number of patients that triggered the protocol. There was no evaluation of online education for staff. And finally, there may be limited generalizability due to the study’s size and limited location. The researchers concluded that rapid recognition and implementation of a sepsis protocol could impact testing and treatment times. They feel that further research is necessary to determine ways to improve sepsis treatment goals in the emergency department.

Tromp et al. (2010) conducted a prospective before-and-after study on 825 adult patients presenting to an emergency department in the Netherlands with a known or suspected infection. The purpose was to evaluate the impact of education and a nurse-driven sepsis protocol on identification patients with sepsis and time to completion of sepsis bundle. The study included patients aged 16 years and older who had at least two criteria of the following criteria:

A sepsis protocol and screening list was developed and implemented during the study. Data was collected before and after implementation of these measures. After this, staff received education on sepsis and data was again collected. Baseline data was collected retrospectively from patients with a diagnosis of sepsis. This group was then narrowed to only those patients that met the criteria above. The remaining data was collected prospectively utilizing the screening and performance lists. Patient characteristic data and performance data were collected. Performance data included time to completion of the six sepsis bundle elements (lactate within 6 hours, two blood cultures prior to antibiotic, chest x-ray, urine specimen for urinalysis and culture, antibiotic administration within 3 hours, and admission or discharge of patient within 3 hours). There is no information available on the reliability and validity of the sepsis identification tool.

Data analysis was performed utilizing several methods. Frequencies, percentages, medians, and means were calculated on the performance data. Percentages calculated from these statistics were used to indicate levels of compliance with individual components and the complete sepsis bundle. The researchers then used a generalized linear model with a logarithmic link and Bernoulli distribution function to analyze the differences between the before and after results. There was a statistically significant improvement in overall bundle completion and in many of the individual components. Four components improved from baseline to completion of the study. Serum lactate collection (23% to 80%), chest x-ray completion (67% to 83%), urine collection (49% to 67%), and administration of antibiotics within 3 hours (38% to 56%) were the elements that improved. Overall the improvement in compliance with sepsis bundle improved from baseline to the second data collection point (incidence 3.0-3.9, 95% CI = 0.7-1.2) and again at the third collection (incidence 3.9-4.2, 95% CI = 0.03-0.5). Researchers noted that study limitations prevented generalization as the study was developed specifically for their situation and it was performed in only one facility. The screening tool was a possible limitation because the clinical signs used to screen for sepsis were not very specific. They concluded that the implementation of a nurse-driven sepsis protocol with education and screening tools can have a significantly positive impact on the identification and treatment of sepsis in the emergency department setting. They believe that future research using a controlled trial in multiple institutions will provide valuable information.

Research Question

Will the initiation of a nurse-driven sepsis protocol affect time to identification, treatment, and admission of the sepsis patient in an emergency department setting? The purpose of this study is to determine if nursing intervention can decrease the time to antibiotic and length of emergency department stay in the sepsis patient.

Design

This investigation will be a quantitative and quasi-experimental study. It will be conducted using a one-group (pretest-posttest) design. This will allow a comparison of patients admitted with a diagnosis of sepsis from the emergency department both retrospective and prospective to the implementation of a nurse-driven sepsis protocol. The independent variable in this research will be the nursing staff in the emergency department of a 650-bed level one trauma center and teaching hospital. The introduction of a nurse driven sepsis protocol will be the intervention used. This protocol will allow nurses in the emergency department to initiate diagnostic testing for suspected sepsis patients. This testing will include a complete blood count (CBC), serum lactate level, and two sets of blood cultures. Initiation of this protocol will also trigger an alert on the patient tracking board. An icon will appear next to the patient’s name to alert the medical staff that the patient is a suspected sepsis patient. Currently sepsis bundles must be initiated by a physician. Retrospective data will be collected prior to the implementation of the nursing sepsis protocol. Prospective data will be collected following the initiation of the nursing sepsis protocol. The dependent variables included will be the time to specimen collection (CBC, serum lactate, and first set of blood cultures), time to administration of first antibiotic, and time to admission order placement.

Sample

The sample for this study will consist of all adult patients with a diagnosis of sepsis. The study will include only patients admitted through the emergency department at Maine Medical Center. Because the goal of this study is to evaluate the overall effectiveness of the nursing sepsis protocol, all patients meeting inclusion criteria will be included and convenience sampling will be used. Data collected for the retrospective sample will include all patients that meet the inclusion criteria for a minimum of three months prior to the implementation of the nursing sepsis protocol. After implementation of the nursing sepsis protocol a three-month transition period will be observed to control confounding variables arising from unfamiliarity with the protocol. Following this transition period, prospective data will again be collected on all patients that meet inclusion criteria for a minimum of three months. The minimum sample size will be 100 for each phase of the study. If a sufficient sample size cannot be obtained over the three-month period, the time frame will be increased until it is obtained. All retrospective data will be collected prior to instituting the nursing sepsis protocol. Prospective data will be collected at completion. The data will be collected, from electronic medical records of identified patients, utilizing a standardized data collection tool. Data will be analyzed using both the independent t-test and ANOVA with a significance of p = .05.

Instruments

The instrument used in this study will be a chart review of each subject’s medical record. During the review, the following data will be collected:

• ED arrival time
• ED triage time
• Time of nursing sepsis protocol activation
• Time of specimen collection for CBC, serum lactate, and 1^st set of blood cultures
• Time of 1^st physician contact
• Time antibiotic order obtained
• Time of initial antibiotic administration
• Time admission order entered in system.

All data collected and variables will be measures of time and thus have content validity. ED arrival time, ED triage time, and admission order time are automatically documented by Epic at the time they are completed. The remaining items may or may not be documented at the time of their occurrence. Given the urgency of treatment initiation in the sepsis patient, nurses and physicians often delay documentation until after the intervention is completed. Reliability is dependent on the accuracy of times documented in the medical record. Assessment of reliability in this case would require direct observation of staff and comparison of the observed intervention times to the documented observation times. Accurate documentation of interventions in sepsis patients is currently an expectation in this emergency department and staff are aware of this. Given these factors, I believe it would be acceptable to conduct the study without determining statistical reliability of the documentation times.

References

Bruce, H.R., Maiden, J., Fedullo, P.F., & Son Chae, K. (2015). Impact of nurse-initiated ED sepsis protocol on compliance with sepsis bundles, time to initial antibiotic administration, and in-hospital mortality. Journal of Emergency Nursing, 41(2), 130-137. doi:10.1016/jen.2014.12.007

Centers for Disease Control and Prevention. (2015). Deaths, percent of total deaths, and death rates for the 15 leading causes of death: United States and each State, 2014 [Data file]. Retrieved from https://www.cdc.gov/nchs/nvss/mortality/lcwk9.htm

Jones, A.E., Focht, A., Horton, J.M., & Kline, J.A. (2007). Prospective external validation of the clinical effectiveness of an emergency department-based early goal-directed therapy protocol for severe sepsis and septic shock. Chest, 132(2), 425-32. doi: 10.1378/chest.07-0234

Seymour, C., Liu, V.X., Iwashyna, J., Brunkhorst, F.M., Rea, T.D., Sherag, A., . . . Angus, D.C. (2016). Assessment of clinical criteria for sepsis: For the third international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA, 315(8), 762-774. doi:10.1001/jama.2016.0288

Slade, E., Tamber, P.S., & Vincent, J. (2003). The Surviving Sepsis Campaign: raising awareness to reduce mortality. Critical Care, 7(1), 1. doi:10.1186/cc1876

Tromp, M., Hulscher, M., Bleeker-Rovers, C.P., Peters, L., Van den Berg, D., Borm, G., . . . Pickkers, P. (2010). The role of nurses in the recognition and treatment of patients with sepsis in the emergency department: A prospective before-and-after intervention study. International Journal of Nursing Studies, 47(12), 1464-1473. doi:10.1016/j.ijnurstu.2010.04,007.