Sepsis is the body's response to an infection, which can progress to organ dysfunction. Mortality is higher for patients with multiple organ failure due to severe sepsis and septic shock. Sepsis continues to increase in incidence and is the most common cause of death in non-coronary critical care units.1,2 Evidence suggests that heightened recognition and timely initiation of appropriate therapy in the early hours after diagnosis of severe sepsis improve clinical outcomes.3-8 As reported by Rivers et al3 in 2001, early goal-directed therapy with timely intervention to achieve hemodynamic resuscitation goals led to reduced in-hospital mortality for patients with severe sepsis and septic shock.3 This study involving 263 emergency department (ED) patients was the foundation for the Surviving Sepsis Campaign (SSC) guidelines that promote 6- and 24-hour bundled care with specific hemodynamic targets.3-6 Based on the initial work of Rivers et al, optimal hemodynamic goals for successful resuscitation in sepsis include central venous pressure (CVP) greater than 8 mm Hg, which indicates adequate fluid resuscitation, and both mean arterial pressure (MAP) greater than 65 mm Hg and mixed venous oxygen saturation (ScvO₂) greater than 70%, indicating adequate tissue perfusion. A pilot study conducted by Girardis and colleagues7 evaluated sepsis-related outcomes associated with 6- and 24-hour bundled care in a cohort of 67 intensive care unit (ICU) patients. Although compliance with bundled care was only 35% in this cohort, results demonstrated a significant decrease in inpatient mortality from 68% to 27% (P < 0.01). More recently, the SSC steering committee collaborated with the Institute for Healthcare Improvement to evaluate integration of SSC guidelines into bedside practice.8 Data were collected on 15 022 subjects at 165 sites to evaluate compliance with bundled care. Compliance with 6- and 24-hour bundles increased from 10.9% to 31.3% (P < .0001) and 18.4% to 36.1% (P < .008) by the end of 2 years, respectively. Hospital mortality decreased from 37% to 30.8% over the same 2-year period (P < .001).

In response to global initiatives and local trends, an interdisciplinary team was created at Ochsner Medical Center to implement strategies consistent with current evidence-based guidelines for sepsis management. The team's goal was to decrease preventable mortality in patients with severe sepsis and septic shock by improving early recognition and process of care.

PROJECT DESCRIPTION AND DESIGN

Initial undertakings included acquiring administrative support and formulating the team for the Sepsis Project. Target patient areas included the 37-bed ED and 48 beds in 2 adult critical care units at this 563-bed tertiary care teaching facility. Key stakeholders who could impact patients across the continuum of care were identified. The medical director for Clinical Practice Improvement collaborated with physician leaders from the ED and critical care service (CCS) to construct a workable plan to manage sepsis (Figure 1). Individual physicians from emergency medicine and critical care medicine assumed leadership roles as cochairs of the Sepsis Committee. Additional committee members included nurses and physicians from the ED and CCS, respiratory therapy, pharmacy, medical informatics, performance improvement, endocrine, dietary, palliative care, and hospital administration.

FIGURE 1. The Sepsis Project plan.

The critical care clinical nurse specialist (CNS) was one of the nurse representatives on the Sepsis Committee. The CNS position was new to the ICU setting in this facility, and therefore the role was not well defined prior to this project. The Sepsis Project offered an opportunity to validate the versatility of the CNS role and to build key relationships with clinicians and nonclinicians (ie, performance improvement, purchasing, and medical informatics) within the organization. The CNS served as a clinical expert, consultant, leader, educator, and researcher to direct changes within the three spheres of influence. The CNS functioned as a critical link in the team by facilitating collaborative relationships, complementing the strengths of all disciplines to create a supportive environment for effective communication, and ensuring diffusion of scientific knowledge and relevant research in support of evidence-based practice. As a content expert, the CNS acted as a change agent to provide an interface between disciplines and geographic areas.

The second step in changing care delivery was equipment acquisition. Prior to project implementation in the various clinical areas, preliminary work involved obtaining equipment for central venous oxygen saturation (ScvO₂) monitoring and point-of-care testing. Procurement of monitoring equipment and intravascular catheters necessitated presenting the business case to hospital administration. The medical informatics department assisted clinician leaders in performing a cost-benefit analysis to compare the use of basic central line catheters versus those with ScvO₂ monitoring capability. The Sepsis Committee felt that the use of special ScvO₂ catheters (Pre-Sep catheter; Edwards Lifesciences Corporation, Irvine, CA) and hemodynamic monitors would decrease complexity of nursing work, improve accuracy of numerical values, and reduce blood wastage to limit transfusion requirements. The CNS partnered with the purchasing department and industry associates to acquire catheter kits and extra monitors as needed for the Sepsis Project. Although point-of-care testing was already in place for monitoring ScvO₂, the committee determined that the additional burden of running lactate levels at the bedside would be avoided by drawing lactate as a "stat" bedside laboratory test. Therefore, lactate is not a point-of-care test at this facility, but rather sent to the laboratory for analysis.

The third step in changing care delivery was the creation of standardized order sets. Order sets were developed by the Sepsis Team to allow for implementation of bundled care for sepsis into daily practice. Two initial order sets were created: one for the ED and one for critical care. Although many disciplines had input, the sepsis order sets were required to include all the recommended items for the 6-hour resuscitation and 24-hour management bundles as outlined by the SSC.5,6 Patients admitted through the ED with severe sepsis/septic shock were expected to be assigned an ED order set with subsequent admission to a critical care unit. The critical care order set was initially designed to transition ED care to the critical care units. However, the critical care order set was also used to treat septic patients with severe sepsis who were transferred from "other inpatient units." The Sepsis Committee expected that patients admitted via the ED would be treated in accordance with both order sets, whereas septic patients transferred directly to a critical care area would receive a critical care order set only. The ED order set focused on the 6-hour resuscitation bundled items including orders for volume, vasopressors, and inotropes, in addition to standardized broad-spectrum antibiotics and laboratory testing. The critical care order set contained items in the 6-hour resuscitation and 24-hour management bundles including orders for continued volume resuscitation, vasopressors, and inotropes as needed to meet hemodynamic goals of CVP greater than 8 mm Hg, MAP greater than 65, and ScvO₂ greater than 70%. Other items contained in the critical care order set included continued antibiotic coverage, ventilator management, insulin therapy, steroids, and use of activated protein C as indicated. The order sets underwent the stringent multidisciplinary review process required. After institutional approval, the orders were made available on the hospital intranet for easy access by all staff. The Sepsis Project mandated that order sets would be initiated on all adult patients with severe sepsis or septic shock.8-10 Patients were identified as suffering from severe sepsis/septic shock as defined by demonstrating 2 or more systemic inflammatory response syndrome conditions, suspected infection, elevated lactic acid levels greater than 4 mmol/L, and hypotension refractory to volume resuscitation (Table).9,10 Order set initiation was a trigger for staff to expedite septic patients through the continuum of care in a standardized fashion. Whereas ED patients meeting criteria for sepsis were typically admitted to a general inpatient medical-surgical units, all patients meeting criteria for severe sepsis or septic shock (sepsis cohort) were to be admitted to a critical care unit.

Table. Definitions and Criteria for Sepsis

The final step in changing care delivery was staff education. The CNS was responsible for educating the large group of interdisciplinary health care providers who treat septic patients. Education sessions highlighted epidemiology, pathophysiology, sepsis classifications, clinical presentation, and treatment of sepsis. Early recognition and awareness of sepsis were emphasized with the aim of improving diagnosis and increasing appropriate treatment. Training was conducted in diverse venues including lectures, video recordings, bedside teaching, and media streaming using the hospital intranet. Nurse and physician leaders participated in promoting the Sepsis Project and educating staff. Unit directors and clinical coordinators facilitated staff availability for education time, which visibly showed support for the sepsis initiative. Physicians were educated during the monthly critical care rotation didactic schedule and on interdisciplinary team rounds.

OPERATIONAL IMPLEMENTATION

The expected course for patients arriving in the ED involved nurse triage of the patient as potentially septic and prompt examination by a physician. Once severe sepsis was diagnosed, sepsis order sets were implemented, a ScvO₂ catheter was placed, and early goal-directed therapy was initiated. To foster successful team participation, ED physicians agreed to place central lines, whereas CCS physicians agreed to see the patient within an hour of diagnosis and expeditiously admit the patient to a critical care unit. Subsequent care was delivered by continuing the sepsis order set begun in the ED in a critical care unit. Prior to this initiative, ED physicians did not routinely place central lines, and ED nursing staff was not accustomed to monitoring patients for CVP and ScvO₂. Therefore, both physicians and nurses required education on hemodynamic monitoring. The CNS facilitated this education as an ongoing consultant for both physician and nurse groups by being available for questions and troubleshooting problems. The CNS also participated in daily rounds with the critical care team to give "real time" feedback and educate staff. Timely communication among the clinical pharmacist, CNS, ED clinical coordinator, ICU nursing staff, and physician leaders played a critical role in the project's outcomes. The Six-Source Influencer Model was used to initiate and sustain change by encouraging personal, social, and structural motivation and ability through rapid cycle feedback.11

DATA COLLECTION PROCEDURE

Organizational approval by the institutional review board was obtained for this project. Patients were reviewed daily on interdisciplinary rounds. Data measurement strategies included concurrent data collection from the bedside chart by the clinical pharmacist. Health Insurance Portability and Accountability Act protection was applied by entering information into a password-protected database. Prospective data collection included unit where order set was initiated, measurement of time to goal attainment with criteria for "6-hour goals met." Specific goals to be met within 6 hours of sepsis diagnosis included serum lactate level drawn, blood cultures prior to antibiotic administration, antibiotic administration within 2 hours, CVP greater than 8 mm Hg, MAP greater than 65 mm Hg, and ScvO₂ > 70%.

DATA ANALYSIS

Rapid cycle review played a critical role in analyzing the processes of care. This review methodology was implemented by a sepsis subcommittee, called the "Sepsis Metrics Team," to improve patient management and provide gap analysis. The Sepsis Metrics Team included the ED and CCS physician leaders, CNS, clinical pharmacist, ED clinical coordinator, a medical informatics analyst, and a performance improvement representative. This small team met monthly to review each individual case, order set usage, time to goals, missed opportunities, and patient outcomes. As a result, individualized timely feedback was provided to staff on opportunities for clinical practice improvement. Feedback to physicians was provided by the physician leader for that area, and nursing feedback was given by the CNS to the unit director, clinical coordinators, and bedside nurses. A second meeting occurred each month with the larger Sepsis Committee, allowing for action plans to be shared based on data abstracted from the metrics meeting.

Statistical analyses included description of frequency distributions for usage of ED and critical care order sets and 6-hour goals met. [chi]² Analyses and simple linear regression using a stepwise method were used to test for relationships between "6-hour goals met" outcomes and location of sepsis order set initiation (ED or critical care unit). SPSS statistical software (PASW, version 18; SPSS Inc., Chicago) was used to complete 2-tailed statistical analyses with significance when P < .05.

RESULTS

Patients within the sepsis cohort (those with severe sepsis and septic shock) who used an order set totaled 674. Twelve patients who were admitted as transfers from other facilities were excluded from statistical analyses (n = 662) as time of sepsis onset could not be established. The sepsis cohort included 449 (68%) of 662 eligible patients who were admitted via the ED, and 213 (32%) of 662 patients transferred to a critical care unit from within the facility. Figure 2 shows the location of admission, and the number of order sets initiated. Patients admitted through the ED were associated with a significantly increased rate of meeting 6-hour goals than those transferred from other inpatient units to critical care ([chi]²₁ [n = 662] = 13.50, P < .001) (Figure 3). The ED order set was initiated in 300 (67%) of 449 patients. Both order sets were used in 230 (51%) of 449. Seventy-five percent (224/300) of ED patients received an ED order set and met 6-hour goals. Seventy-six percent (175/230) of ED patients received both order sets and met 6-hour goals. Lower mortality rates were associated with ED order set use (14%) versus no ED order set use (22%). Two hundred thirteen patients were transferred to a critical care unit from other inpatient units. The critical care order set was used 144 times (144/213, 68%) with a mortality rate of 22% (32/144). The order set was not used 69 times of 213 with a mortality rate of 30.4% (21/69), indicating that order set use affected mortality. Overall, significantly more patients met 6-hour goals when sepsis order sets were used ([chi]²₁ [n = 662] = 36.16, P < .001) ((Figure 4). Findings from simple linear regression indicate that the use of order sets explains 24% of the variation in meeting goals, R² = 0.24, F_1,661 = 38.51, P < .0001.

FIGURE 2. Location of order set initiation.

FIGURE 3. Six-hour goals met and location.

FIGURE 4. Goals met and order set initiation.

DISCUSSION

Overall, sepsis order set use at this facility was associated with improvements in process of care. Significantly more patients admitted through the ED met 6-hour goals than those transferred from other inpatient units to critical care. This outcome is probably due to timely primary diagnosis of sepsis and maintenance of direct patient control through the continuum of care. When therapy was initiated outside the ED setting, other variables may have affected treatment such as time to recognition of sepsis, time to antibiotic administration, central line placement, volume resuscitation, vasopressor initiation, and ScvO₂ monitoring. Sufficient scientific evidence supports early goal-directed therapy as an effective treatment modality when initiated within the first 6 hours of diagnosis.3-8 Rivers et al3,4 studied 6-hour bundled care in ED patients, although ICU process was not specifically studied. Girardis et al7 observed a small group of patients in the ICU setting; however, patients were admitted through a variety of pathways; therefore, initial therapies are unclear. The SSC steering committee reviewed a large number of patients over a 2-year period in 30 countries for clinical compliance with evidence-based bundled care and concluded that adherence to recommended bundles improved patient outcomes.8

In the current project, some in-hospital patients may have had a lower rate of goal attainment due to late recognition of sepsis and progression of sepsis beyond a point at which order sets were helpful. Findings at this facility are similar to those by Rivers et al,3,4 Girardis et al,7 and the SSC steering committee,8 where beneficial outcomes appear to be related to raised awareness of sepsis, staff education, and targeted process change in accordance with evidence-based guidelines.3,4,7,8

Although the Sepsis Project initially targeted patients admitted though the ED with a diagnosis of severe sepsis, subsequent rapid cycle review revealed a large proportion of septic patients transferred to critical care units via the rapid response team (RRT) from other inpatient units. The RRT was identified for specific training in sepsis recognition and management. Intensive care unit nurses and respiratory therapy respond for RRT with backup CCS physician response. Further training and publicity campaigns have been put in place by the CNS to promote use of RRT and to educate members of the RRT itself on the importance of early treatment of sepsis, appropriate laboratory work, antibiotics, and volume resuscitation.

One of the most vital team accomplishments during the project was candid feedback affecting personal, social, and structural motivation and ability. Using the Six-Source Influencer Model, an atmosphere of teamwork was created that overcame reluctance to change and altered vital behavior.11 The Sepsis Committee and the Sepsis Metrics Team both met monthly independent of one another to discuss patient care and change management. Staff communicated with each other between those times whenever missed opportunities were recognized. Review of data on an ongoing basis helped the Sepsis Committee to focus educational efforts toward specific behaviors and care practices. The care of each patient was individually reviewed, and each measure was evaluated: time to diagnosis, blood cultures, antibiotic administration, lactate, ScvO₂ catheter placement, and time to targets for CVP, MAP, and ScvO₂. Major reasons that patients failed to meet goals included no ScvO₂ catheter placed, no CVP or ScvO₂ measurement documented, and failure to administer antibiotics within 2 hours. Great emphasis has now been placed on documentation of CVP and ScvO₂ and immediate administration of antibiotics at the time of sepsis diagnosis or documentation of rationale for inability to do so.

CONCLUSION

The findings from this performance improvement project support use of "bundled" order sets to improve process of care delivery in patients with severe sepsis and septic shock.

IMPLICATIONS FOR PRACTICE

Historically, it has been difficult for the CNS to define individual value to an organization. This project allowed recognition of the facets and adaptability unique to the CNS skill set and permitted the CNS to work in all spheres of influence to integrate various components essential to the project plan. Often, projects are successful when initiated, but maintaining momentum and sustaining results can be challenging. The Sepsis Project has been successful at this facility because of excellent planning prior to project initiation, proper selection of team membership, and steadfast executive support. The Sepsis Team worked collectively to establish a smooth transition for patient care from the ED to critical care. The interdisciplinary panel made communication and accountability a priority that has led to continued performance improvement. The principal success and outcomes of the Sepsis Team are evident in culture change by frontline staff throughout the facility. Positive change has occurred as the Sepsis Protocol has become the practice "norm," and deviation is considered an abnormal occurrence within the organization. As an added benefit of the success of this project, the CNS is now recognized as an integral part of the team and is acknowledged within the institution as an expert in clinical and organizational strategies.

References