Problem Addressed

• A common, fatal condition: Severe sepsis occurs in approximately 750,000 intensive care unit (ICU) patients each year in the United States, with almost 40 percent dying as a result.² At Christiana Care, annual severe sepsis/septic shock mortality rates reached as high as 61.7 percent before the 2005 implementation of the Sepsis Alert Program.
• Effective treatment often not provided: Studies of trauma patients have found that outcomes are improved when patients are adequately resuscitated within 60 minutes of the event.³ In addition, studies have shown that timely antibiotic administration is key to improving sepsis outcomes.⁴ However, suboptimal treatment of sepsis is common. For example, even though the Surviving Sepsis Campaign (a collaborative initiative of the European Society of Intensive Care Medicine, the International Sepsis Forum, and the Society of Critical Care Medicine) has disseminated expert guidelines that suggest the administration of antibiotics within 1 hour of suspected bacterial infection, the average time to first infusion of antibiotics exceeds 3 hours. Other treatment problems that contribute to poor sepsis patient outcomes include use of narrow-spectrum (affecting a limited number of organism types) rather than broad-spectrum (affecting a wide variety of organism types) antibiotics; inconsistent monitoring of physiological parameters such as blood pressure, central venous pressure, and mixed venous oxygen saturation; and lack of treatment protocols and processes to support early goal-directed therapy.⁵

Description of the Innovative Activity

• Prompt screening and evaluation: The program specifies four screening criteria for systemic inflammatory response syndrome (SIRS): elevated heart rate, elevated respiratory rate or low partial pressure of carbon dioxide, elevated or low temperature, and high or low white blood cell count. The four SIRS screening criteria are used in evaluating patients' risk for sepsis; if at least two SIRS criteria are met, a physician's evaluation for sepsis is warranted. At Christiana Care, approximately 85 percent of sepsis patients present to the emergency department (ED); the ED triage nurse performs the initial screen and, for patients in whom sepsis is suspected, requests a complete blood count, blood cultures, serum lactate concentration, and a rapid physician evaluation. Once identified as septic, the patient is immediately placed in an ED room, and the physician is asked to evaluate the patient as soon as possible. The remaining 15 percent of patients with sepsis are identified in the general inpatient area by the floor nurses and physicians during sepsis evaluations or by the rapid response team.
• Sepsis Alert Packet: A Sepsis Alert Packet was designed to expedite the evaluation and the treatment of sepsis patients identified in ED and hospital units. The packet includes the treatment protocol, treatment order set, antibiotic selection chart, and other reference materials. A first-dose drug kit (including single-dose vials of antibiotics and other medications that may be needed, with administration guidelines) is also available. For inpatient units, a rapid response team nurse brings the Sepsis Alert Packet and antibiotic kit to the patient's bedside and initiates appropriate therapy, under a doctor's supervision, within 1 hour of suspected infection.
• Early administration of antibiotics: Physicians order and nurses initiate antibiotic administration within 1 hour of suspected infection, using the following tools to assist:
  
  • Antibiotic selection chart: Physicians can quickly determine the optimal antibiotic to prescribe by consulting a one-page chart outlining which antibiotics are most likely to be effective for specific infections. The chart includes therapeutic options for primary therapy and alternatives for those with penicillin/cephalosporin allergies. The chart covers infections associated with community-acquired pneumonia, health care–associated pneumonia, febrile neutropenia, intra-abdominal infection, and meningitis. The chart helps to prioritize the order of antibiotic administration to maximize treatment efficacy. The antibiotic recommendations are reviewed yearly by an independent team of infectious disease, critical care, and ED physicians in conjunction with the pharmacy department. When changes are made, they are incorporated into the antibiotic packets, and the recommendation charts are edited.
  • Antibiotic posters: Posters that are placed in visible locations in the ED, adult ICUs, and elsewhere in the hospital reinforce the appropriate order of antibiotic administration. This poster helps clarify and ensure that patients get the broadest spectrum antibiotic first and most rapidly. The posters detail which antibiotics are infused over the shortest time.
  • Easy antibiotic availability: All antibiotic options are available in each ED's and ICU's medication-dispensing machines. The rapid response team nurse brings an antibiotic kit to all inpatient calls thought to be related to infection.
• Early goal-directed therapy: A one-page treatment protocol helps to ensure initiation of early goal-directed therapy for patients with confirmed sepsis. The protocol is an algorithm that depicts the recommended order of treatment steps (IV administration of fluid, central line administration, vasopressor agents, and other therapies) and associated therapeutic goals for achieving target rates for systolic blood pressure, mean arterial pressure, central venous pressure, and venous oxygen saturation. The protocol specifies that each step's goals should be achieved before moving on to a subsequent step.
• Treatment order set: With the implementation of computer physician order entry, the physicians are presented with the various treatment recommendations electronically. When the computer system is down, clinicians use a two-page standardized treatment order set with checkboxes to quickly order all laboratory tests and treatments that are needed to monitor and care for the patient.
• Backup support for busy ED physicians: Information provided in October 2012 indicates that, to facilitate early resuscitation and aggressive care for the critically ill, an ICU Alert Team was created. This team is composed of a dedicated critical care nurse and a critical care midlevel provider. Once a patient who has passed through the sepsis alert pathway in the ED is felt to require an ICU admission, the ICU Alert Team is notified and assumes immediate care in the ED.
• Emergency medical services (EMS) Sepsis Alert Program: Information provided in October 2012 indicates that in 2011, a pilot project was initiated with local paramedics to ensure earlier identification and triage of patients with potentially severe sepsis. The paramedics were extensively educated on the SIRS criteria and were equipped with a portable lactate machine for each ambulance. When the patients are found to have an elevated lactate level, they are immediately started on fluid resuscitation (via ED physician orders). They are stratified into two categories: EMS Sepsis Alert (EMS lactate level > 4 mmol/L with two SIRS criteria) or EMS Sepsis Advisory (EMS lactate level of 2.5 to 3.9 mmol/L with two SIRS criteria). On arrival, patients are immediately placed in ED beds as opposed to presenting to triage. Aggressive sepsis resuscitation is continued with the goal to significantly reduce time to resuscitation and antibiotic administration.
• Ongoing data collection: A tool is used to collect data on patient demographics; infection site; antibiotic(s) used; and the time to identification, sepsis confirmation, antibiotic administration, and achievement of each goal.

References/Related Articles

Contact the Innovator

Innovator Disclosures

Results

• Decrease in severe sepsis/septic shock–related mortality, morbidity, and LOS: Updated information provided in November 2011 indicates that severe sepsis/septic shock mortality was reduced from 61.7 to 16.7 percent (this reduced the overall sepsis-related mortality by 53 percent); the incidence of sepsis-related organ dysfunction declined by 63 percent for acute respiratory distress syndrome and by 65.7 percent for acute renal insufficiency/failure. Average LOS for sepsis patients (excluding 70-day outliers) decreased from 18.2 to 9.3 days as of November 2011. Information provided in October 2012 indicates that, between program inception and April 2012, mortality has been further reduced to 15.68 percent, and ICU LOS has been reduced from 11.9 to 4.1 days.
• Discharge of more patients to home after stays in the ICU: Information provided in November 2011 indicates that the percentage of patients discharged to home from the ICU increased by 199 percent, which suggests that fewer patients experienced a decline in health status from a prolonged ICU stay.
• Quicker fluid resuscitation and initial improvement in administration of antibiotics: Information provided in November 2011 indicates that time to fluid resuscitation decreased from 1.21 to 1.17 hours after sepsis diagnosis. The percentage of sepsis patients receiving antibiotics less than 1 hour after diagnosis increased from 92.4 to 96.7 percent in 2007, but declined to 95 percent in 2011.
• Increase in scores for Institute for Healthcare Improvement (IHI) sepsis resuscitation bundle: Information provided in November 2011 shows that Christiana Care's compliance with IHI's acute sepsis resuscitation bundle increased from 28.6 to 45 percent of patients. (For more information about the bundle components, see link to the IHI under Tools and Other Resources.)
• Successful care for EMS Sepsis program patients: Information provided in October 2012 indicates that 68 patients were placed in the EMS Sepsis Alert category; 75 percent had a final diagnosis of sepsis/severe infection, and 41 percent (of total) were admitted to the ICU, with median antibiotic administration time of 60 minutes. Of 73 patients placed in the EMS Sepsis Advisory category, 68 percent had a final diagnosis of severe infection/sepsis, and 26 percent (of total) were admitted to the ICU.
• Presumed financial benefit: Although actual financial savings have not been calculated, the results described above suggest a positive financial benefit owing to fewer nursing home admissions, which generates savings for the system; and lower LOS, which reduces costs and enhances revenues by freeing up beds to handle new admissions.

Evidence Rating (What is this?)

Context of the Innovation

Planning and Development Process

• Senior management approval: In February 2004, Christiana Care's director of critical care medicine obtained approval from Christiana Care's vice president of performance improvement to develop a task force to improve sepsis care.
• Creation of interdisciplinary team: The director of critical care medicine formed a multidisciplinary team that included physicians from internal medicine, emergency medicine, and surgery, as well as representatives from nursing, pharmacy, and performance improvement. Approximately 20 individuals met monthly over a 6-month period, conducting a review of evidence-based clinical guidelines and expert recommendations. The team also monitored the clinical recommendations being developed by the Surviving Sepsis Campaign, with the goal of making the campaign's recommendations actionable in a clinical setting.
• Program design and development: The team designed the program to reflect a process for rapidly identifying at-risk patients, evaluating clinical status, and providing appropriate therapy. The team brainstormed ideas for clinical tools that would help support the process, outlined necessary tasks, and made assignments based on team members' areas of expertise. For example, pharmacy representatives on the team were charged with determining how to ensure that the right antibiotics were available quickly, while ED nursing representatives took responsibility for educating ED triage nurses about how to identify patients with sepsis and how to correctly measure central venous pressure. Individual team members also designed the program components, including the antibiotic poster, treatment protocol, ED treatment order sets, and data collection tool.
• Team review and refinements: Processes and tools developed by individual team members were presented at team meetings; they were critiqued and changed accordingly.
• Education and training: Team members designed and implemented ideas for staff education, including Web-based seminars and posters to be hung throughout the hospital. In addition, the director of critical care medicine gave grand-round lectures to multiple departments. Clinicians receive additional education every 3 months through lectures, newsletters, and other efforts designed to keep sepsis care "top of mind."
• Continuous quality improvement: Monthly performance improvement meetings were held to assess compliance and review outcomes data. The team utilized the Christiana Care Plan-Do-Check-Act method for continuous improvement, using rapid cycle improvements to make adjustments and then evaluate their impact. Information provided in November 2009 indicates that Christiana Care now maintains the process through quarterly meetings with multidisciplinary representation.

Resources Used and Skills Needed

• Staffing: No additional staffing was required to develop or implement the program.
• Costs: Costs associated with the program were minimal, consisting primarily of costs for the development of educational materials, the first-dose antibiotic kit, and other materials.

Funding Sources

Tools and Other Resources

Getting Started with This Innovation

• Adopt a "can-do" attitude: The Christiana Care team was highly successful because all members were willing to address and change processes to improve sepsis care.
• Emphasize standardization: It was made clear to clinicians that care algorithms are developed based on clinical evidence or expert opinion.
• Ensure cross-disciplinary interactions: A "silo" mentality with carefully designated responsibilities can be detrimental to patient care. Ongoing contact between ED and ICU/e-ICU™ physicians is critical to ensuring prompt care for every sepsis patient.
• Expand the focus to the hospital floors as well: Floor nurses are not necessarily trained or equipped to manage an acutely ill patient with sepsis. Designating the rapid response team to care for these patients and having all necessary materials readily accessible on the units can improve sepsis outcomes throughout the hospital.

Sustaining This Innovation

• Monitor compliance: Ongoing monitoring of compliance and performance is key to achieving and maintaining good outcomes. Data should be tracked and new processes tested to assess their impact. Information provided in October 2012 indicates that teams should ensure adequate resources for enhanced monitoring of compliance with components of the sepsis resuscitation bundle.
• Reinforce education on a regular basis: To keep sepsis care "top of mind" among clinicians, educational messages should be periodically reinforced through new posters, newsletter articles, and other forums. In addition, the positive results achieved from the program should be widely disseminated and promoted. Information provided in October 2012 indicates that grand rounds can be used to provide an update on anticipated new guidelines, and lectures can be delivered to medical house staff, ED physicians, and hospitalists.
• Incorporate new recommendations: Review the literature for evidence-based findings or new expert consensus statements on sepsis diagnosis and treatment, and revise processes and recommendations accordingly.
• Encourage interdisciplinary collaboration among specialties: Information provided in October 2012 indicates that program developers can help sustain the program by making a specific effort to improve recognition of inhospital sepsis by collaborating with infectious disease consultants, intensivists, and hospitalists.