Standardizing Sepsis Screening and Management via a Tele-ICU Program Improves Patient Care

Abstract

Objective: This article evaluates the feasibility of a tele-intensive care unit (ICU) nurse–driven early identification and treatment process for severe sepsis patients in improving compliance to evidence-based practice. Materials and Methods: Florence Nightingale identified that by using science, logic, and compassion to manipulate the patient care environment nurses could create the best possible conditions for healing to occur. Nurses in a tele-ICU used this premise to initiate a standardized screening and data collection program using a custom-built document sharing application that conformed to the Surviving Sepsis Campaign (SSC) criteria for identification and treatment of severe sepsis. Results: The tele-ICU nurses performed 89,921 screens on 36,353 ICU admissions to 161 ICU beds across a geographical range of 500 miles. Between January 1, 2006 and December 31, 2008, tele-ICU nurses identified 5,437 patients as meeting the criteria for severe sepsis. Statistically significant increases in compliance with SSC's bundled care recommendations were realized during this study period with four initial elements: antibiotic administration increased from 55% in 2006 to 74% in 2008 (p=0.001), serum lactate measurement increased from 50% to 66% (p=0.001), the initial fluid bolus of ≥20 mL/kg increased from 23% to 70% (p=0.001), and central line placement increased from 33% to 50% (p=0.001). Conclusions: A tele-ICU nurse–driven process can prompt earlier identification and improve compliance to evidence-based practice bundles for complex disease states such as severe sepsis.

Introduction

Health services research has highlighted the importance of medical errors. These reports define errors as a failure to execute a planned action or the use of a wrong plan to achieve a patient care goal.^1,2 Multiple researchers have emphasized the importance of systematically refining work processes and of care team communication and coordination to address errors of omission and commission in general inpatient medical and nursing practice.^3,4 Improvements in the care of the critically ill will require that two particular challenges be addressed.

Problem #1

Intensive care units (ICUs) provide specialized care to critically ill patients. ICUs face an increasing shortage of trained critical care nurses and physicians to safely manage highly complex patients.^5
–7 The gap between supply and demand of registered nurses is estimated to increase to over 400,000 by 2020.⁸ Although the strategic use of technology may not increase numbers in the nursing workforce, it can augment skill mix and optimize staffing.⁹ Evidence suggests that the presence of intensivists “at the bedside” in the ICU is linked to a decrease in complications and mortality rates as well as lower hospital resource utilization.¹⁰ Unfortunately, the shortage of intensivists is predicted to reach 35% in the United States by 2020.¹¹ These concerns highlight the need for innovation in critical care to discover more affordable and efficient ways to safely deliver care to diverse populations of patients.

Problem #2

Severe sepsis remains a significant cause of mortality and morbidity despite medical advances and evidence-based recommendations for the treatment of this disease state. Research suggests that over 500 people die each day of severe sepsis and that incidence and mortality rates are expected to increase as the population ages.^12,13 It is a complex disease state that is difficult to define, diagnosis, and treat.¹⁴

In 2003, an international team of experts joined to form the Surviving Sepsis Campaign (SSC). The SSC produced a bundled care approach that specifically targeted early and aggressive identification and treatment of severe sepsis.¹⁵ The treatment bundles were further refined and ratified in 2008.¹⁶ A “bundle” is defined as a group of therapies that when implemented together can result in better outcomes for specific disease states.¹⁷ The resuscitation bundle was derived from the Early Goal-Directed Therapy (EGDT) bundle and is based on seven evidence-based goals or elements that must be completed within 6 h of presentation of severe sepsis.¹⁸

A study by Brunkhorst et al. published in 2008 suggested that there were disparities between actual care and perceived care related to compliance with the SSC care bundles.¹⁹ Sutter Health noted inconsistency between the perceived rate of compliance and the documented rate of compliance to EGDT or the 6-h resuscitation bundle. In 2005, a regional ICU Quality Committee at Sutter Health reviewed outcomes using risk-adjusted mortality and length of stay data (Cerner APACHE III^®). This review revealed that sepsis mortality rates at the five hospitals within the region were variable and generally higher than predicted. Compliance to the 6-h bundle elements was observed to be less than optimal.

Florence Nightingale identified that nursing practice must have the components of educated individuals using science, logic, and compassion for appropriate problem solving to occur.²⁰ Her model of nursing was built on the foundation that manipulations to the patient care environment can lead to the best possible conditions for healing to occur. Complex Adaptive System's model and the “butterfly effect” contend that small changes in the initial condition of a complex environment could produce large and lasting variation in system behavior.²¹ Using the aforementioned paradigms, the tele-ICU nurses were empowered to examine their practice, identify resources and tools needed to increase their efficiency, develop processes to change the environment of critically ill patients admitted to the ICUs that they monitored, and evaluate the effectiveness of changes.^22,23

The primary objective of this study was to evaluate the effectiveness of a nurse-driven early identification and treatment process for severe sepsis patients to improve compliance to evidence-based practice bundles.

The secondary objective was to demonstrate that the primary objective could be accomplished from a remote or tele-ICU environment with technology.

Materials and Methods

The Team and The Technology

Sutter Health is a large health system in North California, serving >5,000 hospital beds in over a 500-mile radius. Wide ranges in size, demographics, and location for the 26 hospitals within this health system exist. The Sutter experience mirrors what is being realized across the nation, that is, disparities in care are present because of shortages of specialists and varying levels of clinical expertise and resources within hospitals.

In January 2003, Sutter Health implemented remote monitoring of hospital ICUs via a Tele-ICU Program. This software application assesses and monitors critically ill patients via in-room video cameras with audio, Web-based connections to vital sign monitoring devices and radiological images, and interfaces to laboratory results and hospital registration information. The goal of the tele-ICU is to provide all patients the benefit of round-the-clock access to the expertise of experienced critical care nurses and intensivists. By having a tele-ICU, all patients in Sutter Health's ICUs have the same access to additional vital sign monitoring and timely assessments and care by a comprehensive critical care team. Because of Sutter Health's geographical range and volume of patients, two remote monitoring sites were established, one based in Sacramento (Sac) and the other in San Francisco. The Sac site is the source of the data for this study. The Sac tele-ICU serves the Central Valley, the Sierra Foothills, and the North Coast.

The tele-ICU nursing team allowed for the rapid development and deployment of care processes and tools because of the small number of nurses that needed to be trained, a high level of expertise, and a culture of nurse empowerment. The tele-ICU team is comprised of critical care physicians (>90% board certified) and registered nurses with an average of 15 years of critical care experience. This allowed for leveraging of resources and knowledge virtually across 10 hospitals.

The Setting

There were 161 ICU beds at 10 hospitals across a geographical range of 500 miles included in this study. ICUs ranged in size from 4 to 35 beds. Four hospitals were large and had tertiary care centers with specialist physicians available and the other hospitals were considered rural and/or small with limited or no specialist physicians. All sites had varying degrees of nursing skill mix, with lower expertise more prevalent on the night shift. Approximately one-third of the beds included in this study had started using a protocol or order set for the treatment of severe sepsis patients. The remaining two-thirds had some knowledge of severe sepsis but had not begun any formal screening and treatment processes, or they had limited or no knowledge of a standardized approach to define, diagnose, and/or treat severe sepsis.

Tools and Process

The evaluation model chosen was a realistic evaluation approach or context plus mechanism equals outcome (CMO) using quantitative data as endpoints. This approach has been discussed in previous literature as an alternative to the more traditional randomized control trial method for several reasons: (1) it allows the care providers to define an intervention that they think might be effective, (2) it formalizes the hypothetical context, mechanisms, and outcome patterns (CMO), and (3) a realistic evaluation does not attempt to control the outcomes but rather seeks to understand patterns between CMO.^24
–26

The tele-ICU nursing team defined that their intervention would be identifying and advocating for the care of patients with severe sepsis. They identified that a key barrier to early identification of this patient population was related to screening and sorting the high volume of patients that they monitored. The absence of an easy-to-use, standardized form that could be readily shared among the remote team was needed. Information services (IS) staff collaborated with the tele-ICU nursing team to define and develop a document sharing application to record, track, and retrieve the screening results and associated treatments. Input from nurse and physician experts from within and outside the organization was solicited during the development phase. The IS application collected the screening and treatment information into discrete data fields, provided for easy sharing of the document among the tele-ICU staff, and in a subsequent phase, semiautomated the repeat screening of ICU patients. This solution also allowed for near real-time reports to be generated on compliance to bundle metrics and the incidence of severe sepsis in the ICUs served by the tele-ICU.

Although this solution allowed for the capture of discrete data elements, the form doubled as a real-time “checklist to care.” Using checklists in the care of critically ill patients has demonstrated improved adherence to evidence-based protocols and bundled care approaches.²⁷ This solution was a significant improvement over traditional paper forms and even electronic forms that capture information as a graphic. The goal set for the pilot was 85% of patients screened at least once during their ICU stay. Approval was sought and granted from the Sutter Health Institute for Research and Education Investigational Research Board. This study used de-identified data collected during the routine course of medical care. A waiver of written consent and authorization was obtained.

The design, development, and approval of the electronic form took place in September 2005. The nursing staff at the tele-ICU coordinated with the bedside staff to collect the information to complete the form. The new process was defined as screening every patient every shift (there are two nursing shifts per day at the tele-ICU) and screening new patients within 1 h of admission to the ICU. The pilot was conducted from October 1 to December 31, 2005. A goal of 85% screening of all ICU patients was reached by the end of the 3-month pilot. Leadership decided that the demonstrated progress in screening compliance and the reaching of the 85% goal was sufficient to launch a system-wide implementation.

The sepsis audit and screening tool moved into production for the Sac Tele-ICU in January 2006 and for the San Francisco Tele-ICU in June 2006. As hospitals integrated into the tele-ICU, they were immediately included in the sepsis screening process. By April 2006, a data repository with standard and flexible reporting was implemented with monthly reporting to the hospitals served by the Sac Tele-ICU. The supporting details were provided for each positive screen for in-depth analyses and validation of data. This provided timely reports of compliance to the severe sepsis bundle elements.

The Sac Tele-ICU nurses working with the tele-ICU intensivists defined a standard process for the treatment of patients meeting criteria for severe sepsis. The tele-ICU team worked collaboratively with bedside teams to execute care processes with the goal of improving compliance to the SSC bundle and reduction in mortality. The IS department designed and built enhancements to the electronic form that supported these process improvements. Additionally, the rate of screening and the compliance with the initial treatment bundle as well as outcome results were studied and reported in the literature.^28

–33

Subjects

Between January 1, 2006 and December 31, 2008 nursing staff at both tele-ICUs combined performed a total of 211,833 sepsis screens on 68,933 unique patient admissions. Of these, the Sac hub performed 89,921 screens on 36,535 patient admissions. The standard process was to screen every patient within 1 h of admission to the ICU and once per shift for ICU patients if the patient had a suspicion of infection. This process resulted in multiple screens per patient. Our analysis focuses only on screens that were positive for severe sepsis and uses a distinct criterion that counted only one positive screen per patient per admission. There were a total of 5,437 distinct patients positive for severe sepsis during this time period.

Our initial analysis focused on process metrics (screening rate, percent positive screens, and location of patient when criteria for severe sepsis were met). The subsequent analysis for treatment compliance was in accordance with the SSC 6-h resuscitation or EGDT bundle and included the following process metrics:

• Blood cultures prior to antibiotics

• Antibiotics in the prescribed time

• Baseline labs

• Lactate measurement

• Fluid bolus of 20 mL/kg

• Central line access

The electronic form enabled the tele-ICU team to identify patients as “not septic,” “comfort care only,” or “care is appropriate.” These phenomena were insignificant at 0.8% and, therefore, were not excluded from the analysis. Statistical analysis using asymptotic confidence interval test for a difference between two binomial proportions was performed to analyze the percentage of patients receiving each of the initial treatments.

Results

Using the asymptotic confidence interval test, statistically significant increases in compliance with SSC bundled care recommendations were realized during this study period with the following initial elements. Timeliness of antibiotics increased from 55% to 74% (p<0.001). Measuring lactates increased from 50% to 66% (p<0.001). Obtaining blood cultures before antibiotics showed a small increase from 65% to 70%, which was not statistically significant (p=0.12). Measuring baseline labs showed a small improvement and then a steady decrease (81%–75%, p=0.05).

The percentage of patients receiving the 20 mL/kg fluid bolus and receiving a central line were compared with the count of patients meeting the criteria for needing EGDT. The criteria used for needing EGDT was blood pressure <90 mm/Hg, mean arterial pressure (MAP) <65 mm/Hg, or lactate >4 mM. The percentage of patients receiving the fluid bolus shows marked improvement, rising from an initial value of 23% to a sustained value of 70% (p<0.001). The central line access also increased from 33% to a sustained value of 50% (p<0.001).

In summary, in 2008, patients were more likely to receive timely antibiotic administration, have a lactate level measured, receive the initial fluid bolus of 20 mL/kg, and have a central line placed. Unfortunately, because of lack of local buy-in, the measurement and treatment of central venous oxygen saturation (ScVO2) were not performed in a large enough quantity to analyze. Although not directly measured in this study, there is substantial research that shows that increased compliance with the SSC bundle improves patient outcomes.

Discussion

Patients deserve to receive care that is based on the best available science and appropriate care should not be hindered because of the lack of knowledge or resources. Although implementing evidence-based practice at the bedside has historically proven to be difficult to achieve, we believe that these data suggest that a tele-ICU intervention can improve the patient care environment. Improvements in care delivery can be realized and disparities of care mitigated when clinical expertise is leveraged through innovative models of care. These data suggest that the following points:

(1) A nurse-driven early identification and treatment process for severe sepsis patients enhanced by technology tools can improve compliance to evidence-based practice bundles.

(2) Tools developed with input from end users can achieve a higher rate of adoption.

(3) Tele-ICU nursing practice can prompt earlier identification and treatment for a complex disease state.

(4) Leveraging nurse and physician expertise through a virtual environment is a viable solution to mitigate disparities of care across a large healthcare system.

At the time of the writing of this article, data on compliance to the SSC bundle was collected and published on a monthly basis. Compliance to the SSC bundle has continued to improve. System-wide education and focus on quality improvement for this population of patients continues.

Limitations

Knowledge deficits and resistance to a standardized care approach for the treatment of this population of patients was present at all sites in varying degrees. The fact that this was a multicentered study strengthens the validity of the results, despite the fact that multiple variables were present, such as size, location, and other hospital demographics. The degree of acceptance of the tele-ICU team interventions also varied and is not discussed in this study. This could be viewed as a weakness in this study design. A well-designed, randomized study with thorough economic analysis needs to be performed to validate these conclusions and evaluate the cost effectiveness of this model of care.

Conclusions

Implementing evidence-based practice at the bedside has historically proven to be difficult to achieve. Statistically significant increases in compliance with the SSC's bundled care recommendations were realized during the study period: (1) increased antibiotic administration, (2) increased measurement of serum lactate, (3) increased administration of initial fluid bolus, and (4) increased placement of a central line.

We believe that our experience and observations are typical of most large healthcare systems. The healthcare industry is facing increased pressures to provide high-level quality of care in a fiscally responsible manner. An aging population, increased costs, increased regulatory demands, and decreased reimbursements coupled with a diminishing critical care labor force are confounding variables to achieve these goals. To continue to provide effective, safe, and equitable care, we will be required to continually define and implement innovative models of care. More studies using innovative care models and technology solutions are needed.

Footnotes

Acknowledgments

The authors thank the Sutter Health eICU^® and IS Teams, John Mesic, M.D., Daniel Ikeda, M.D., Tom Shaughnessy, M.D., Gordon Hunt, M.D., and Bill Farrell, Ph.D., for help with this study.

Disclosure Statement

No competing financial interests exist.

References

Institute of Medicine. To err is human: Building a safer health system. Washington, DC: National Academy of Sciences, 2000.

National Academy of Sciences. Crossing the quality chasm: A new health system for the 21st century. 2000. www.iom.edu. 2010 June 20.

McCauley

, Irwin

. Changing the work environment in intensive care units to achieve patient-focused care: The time has come. Am J Crit Care, 2006; 15:541–548.

Kim

, Barnato

, Angus

, Fleisher

, Kahn

. The effect of multidisciplinary care teams on intensive care unit mortality. Arch Intern Med, 2010; 170:369–376.

Auerbach

, Buerhaus

, Straiger

. Better late than never: Workforce supply implications of later entry into nursing. Health Aff, 2007; 26:178–185.

Kozar

, Shackford

, Cocanour

. Challenges to the care of the critically ill: Novel staffing paradigms. J Trauma, 2008; 64:366–373.

The Leap Frog Group. ICU physician staffing (IPS) factsheet. 2008. www.leapfroggroup.org/media/file/FactSheet_IPS.pdf. 2010 July 4.

Buerhaus

, Needleman

, Mattke

, Stewart

. Strengthening hospital nursing. Health Aff, 2002; 21:123–132.

NACNEP Sixth Report to the Secretary of Human Services the Congress. U.S. Department of Health and Human Services. Washington, DC: U.S. Government Printing Office, 2008. ftp://ftp.hrsa.gov/bhpr/nursing/sixth.pdf. 2010 January 25.

10.

Provonost

, Needham

, Waters

, Birkmeyer

, Calinawan

, BMath

, Birkmeyer

, Dorman

. Intensive care physician staffing: Financial modeling of the leapfrog standard. Crit Care Med, 2006; 32:1247–1253.

11.

Angus

, Shorr

, White

, Dremsizov

, Schmitz

, Kelley

. Committee on Manpower for Pulmonary Critical Care Societies. Current and projected workforce requirements for care of the critically ill and patients with pulmonary disease: Can we meet the requirements of an aging population? JAMA, 2000; 284:2762–2770.

12.

Angus

, Linde-Zwirble

, Lidicker

, Clermont

, Carcillo

, Pinsky

. Epidemiology of severe sepsis in the United States: Analysis of incidence, outcome, and associated costs of care. Crit Care Med, 2001; 29:1303–1310.

13.

Dombrovskiy

, Martin

, Sunderram

, Paz

. Rapid increase in hospitalization and mortality rates for severe sepsis in the United States: A trend analysis from 1993 to 2003. Crit Care Med, 2007; 35:1244–1250.

14.

Surviving Sepsis Campaign. Surviving sepsis: About the campaign. 2008. www.survivingsepsis.org/About_the_Campaign/Pages/default.aspx. 2010 December 11.

15.

Dellinger

, Carlet

, Masur

, Gerlach

, Calandra

, Cohen

, Gea-Banacloche

, Keh

, Marshall

, Parker

, Ramsay

, Zimmerman

, Vincent

, Levy

. Surviving sepsis campaign guidelines for management of severe sepsis and septic shock. Crit Care Med, 2004; 32:858–873.

16.

Dellinger

, Levy

, Carlet

, Bion

, Parker

, Jaeschke

, Reinhart

, Angus

, Brun-Buisson

, Beale

, Calandra

, Dhainaut

, Gerlach

, Harvey

, Marini

, Marshall

, Ranieri

, Ramsay

, Sevransky

, Thompson

, Townsend

, Vender

, Zimmerman

, Vincent

. Surviving sepsis campaign: International guidelines for management of severe sepsis and septic shock. Crit Care Med, 2008; 36:296–327.

17.

Surviving Sepsis Campaign. Implement the resuscitation bundle—within the first 6 hours of care. 2008. www.survivingsepsis.org/Bundles/Pages/SepsisResuscitationBundle.aspx. 2010 December 11.

18.

Rivers

, Nguyen

, Havstad

, Ressler

, Muzzin

, Knoblich

, Peterson

, Tomlanovich

. Early Goal-Directed Therapy Collaborative Group. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med, 2001; 345:1368–1377.

19.

Brunkhorst

, Engel

, Ragaller

, Welte

, Rossaint

, Gerlach

, Mayer

, John

, Stuber

, Weiler

, Oppert

, Moerer

, Bogatsch

, Reinhart

, Loeffler

, Hartog

. German Sepsis Competence Network (SepNet). Practice and perception—a nationwide survey of therapy habits in sepsis. Crit Care Med, 2008; 36:2719–2725.

20.

Selanders

. The power of environmental adaptation. J Holistic Nurs, 2010; 28:81–88.

21.

Holden

. Complex adaptive systems: Concept analysis. J Adv Nurs, 2005; 52:651–657.

22.

Jacobson

. Research utilization in nursing: The power of one. Orthop Nurs, 2000; 19:61–65.

23.

Gawlinski

. The power of clinical nursing research: Engage clinicians, improve patients' lives, and forge a professional legacy. Am J Crit Care, 2008; 17:315–326.

24.

Clark

, Whelan

, Barbour

, MacIntyre

. A realist study of the mechanisms of cardiac rehabilitation. J Adv Nurs, 2005; 52:362–371.

25.

Ogrinc

, Batalden

. Realist evaluation as a framework for the assessment of teaching about the improvement of care. J Nurs Educ, 2009; 48:661–667.

26.

Berwick

. The science of improvement. JAMA, 2008; 299:1182–1184.

27.

Byrnes

, Schuerer

, Schallom

, Sona

, Mazuski

, Taylor

, McKenzie

, Thomas

, Emerson

, Nemeth

, Bailey

, Boyle

, Buchman

, Coopersmith

. Implementation of a mandatory checklist of protocols and objectives improves compliance with a wide range of evidence-based intensive care unit practices. Crit Care Med, 2009; 37:2775–2781.

28.

Rincon

, Bourke

, Ikeda

. Centralized, remote care improves sepsis identification, bundle compliance, and outcomes. Chest, 2007; 132:557b–558b.

29.

Rincon

, Bourke

, Ikeda

, Seiver

. Screening for severe sepsis: An incidence analysis. Crit Care Med, 2007; 35:A257.

30.

Hayatdavoudi

, Yee

, Rojas

, Rincon

, Winchell

, Ikeda

. Impact of a surviving sepsis protocol employing a vasopressor restrictive strategy on the survival of the sepsis subgroups when compared to predicted APACHE III risk adjusted outcomes. Crit Care Med, 2007; 34:A259.

31.

Ikeda

, Hayatdavoudi

, Winchell

, Rojas

, Rincon

, Yee

. The impact of using a standard protocol for the surviving sepsis 6 and 24 hour bundles in septic patients on total ICU risk adjusted mortality. Crit Care Med, 2006; 34:A108.

32.

Hayatdavoudi

, Ikeda

, Seiver

, Winchell

, Rojas

, Rincon

, Yee

. Impact of a protocol treating severe sepsis on renal function and survival of septic shock patients in an open adult ICU. Crit Care Med, 2006; 34:A18.

33.

Ikeda

, Hayatdavoudi

, Winchell

, Rojas

, Rincon

, Yee

. Implementation of a standard protocol for the surviving sepsis 6 and 24 hr bundles in patients with an APACHE III admission diagnosis of sepsis decreases mortality in an adult open ICU. Crit Care Med, 2006; 34:A2.