REVIEW URRENT C OPINION
Quality improvement in pediatric sepsis Elliot Melendez a,b and Richard Bachur b
Purpose of review Although there is abundant literature detailing the impact of quality improvement in adult sepsis, the pediatric literature is lacking. Despite consensus definitions for sepsis, which patients along the sepsis spectrum should receive aggressive management and the exact onset of sepsis (‘time zero’) are not clearly established. In the adult emergency department (ED), sepsis onset is defined as the time of entry into the ED; however, this definition cannot be applied to hospitalized patients or patients who evolve during their ED course. Since the time of sepsis onset will dictate the timeliness of subsequent process measures, the variable definitions in the literature make it difficult to generalize findings among prior studies. Recent findings Despite the variation in defining time zero, aggressive fluid administration, timely antibiotics, and compliance with sepsis bundles have been shown to improve mortality and to reduce hospital and intensive care length of stay. In addition, early identification tools show promise in beginning to define sepsis onset and retrospective search tools may allow improved case finding of those children of concern for sepsis. Summary Quality improvement in pediatric sepsis is evolving. As we continue to define quality measures, we must standardize the definition of sepsis onset. This definition should be applicable to any treatment venue to ensure measures can be evaluated across all settings. In addition, we must delineate which patients along the sepsis spectrum should be candidates for timely interventions and standardize other outcome measures beyond mortality. Keywords pediatric, quality improvement, sepsis
INTRODUCTION Pediatric severe sepsis and septic shock is a leading cause of morbidity and mortality, and contributes to high healthcare costs [1 ,2 ]. However, understanding the true morbidity and mortality is challenging since there may be variable definitions for sepsis in both the research and quality improvement literature, and uniformity does not exist. Using the Pediatric Health Information Systems (PHIS) database, Balamuth et al. found that the incidence of pediatric severe sepsis increased over a 9-year period, but mortality decreased [1 ]. However, in selecting cases coded as sepsis by the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), the overall mortality for sepsis was 21.2%, which is more than twice the last estimate provided by Watson et al. [3]. Ruth et al. [2 ] looked at children admitted to a pediatric ICU in the same cohort of PHIS patients. Using a different definition, in which severe sepsis was assigned if an ICD9-CM code for infection was associated with one organ dysfunction, mortality was 14.4%. Though this methodology will identify more &
&
&
&
www.co-pediatrics.com
patients with risk of severe sepsis, the current consensus definition by Goldstein et al. [4] defines severe sepsis as two or more organ dysfunction in the setting of clinical sepsis. Both of these studies published in the same edition of ‘Pediatric Critical Care Medicine’ (November 2014) highlight the difficulty in accessing outcomes if considering quality improvement initiatives, because the definitions of whom is included can vary. Using uniform definitions, especially those that can be assigned prospectively, will allow for better measurement of outcomes, identify barriers to adherence with
a
Division of Medicine Critical Care and bDivision of Emergency Medicine, Boston Children’s Hospital, Boston, Massachusetts, USA Correspondence to Elliot Melendez, MD, Associate Director, Safety and Quality, Division of Medicine Critical Care, Sepsis Quality Leader, Division of Emergency Medicine, Boston Children’s Hospital, 300 Longwood Avenue, Main 11 South, Boston, MA 02115, USA. Tel: +1 617 355 4522; fax: +1 888 883 9238; e-mail: [email protected]. edu Curr Opin Pediatr 2015, 27:298–302 DOI:10.1097/MOP.0000000000000222 Volume 27 Number 3 June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Quality improvement in pediatric sepsis Melendez and Bachur
KEY POINTS Time zero, the time by which the onset of sepsis occurs, should be uniform in studies to allow appropriate comparisons of process and outcome measures. Process measures should be assessed and any quality improvement interventions should target local barriers. Outcome measures such as mortality should always be assessed, but intermediate outcomes, such as hospital and ICU length of stay, are meaningful outcomes. Sepsis quality improvement in pediatrics should be benchmarked across institutions while accounting for severity of illness or complex care conditions. A severity of illness score is needed to insure we understand risk as related to quality improvement interventions.
quality improvement processes, and clarify the controversies in the sepsis literature.
OUTCOMES RELATED TO SPECIFIC INTERVENTIONS IN PEDIATRIC SEVERE SEPSIS Ever since Carcillo et al. [5] showed that children who received less than 40 ml/kg of isotonic fluids during management for septic shock had a higher mortality rate, it has been recognized that inadequate fluid resuscitation is a predictor of worse outcomes. Paul et al. [6] reported the incidence of low compliance with septic shock processes as recommended by the Pediatric Advanced Life Support Sepsis Guidelines in 126 children who presented to a pediatric emergency department (ED) with severe sepsis and septic shock. Specifically, timely fluid administration of at least 60 ml/kg of isotonic fluid within the first hour of shock only occurred in 37% of the children. Nonadherence with timely fluid administration was statistically associated with both a longer ICU stay and a hospital length of stay [6]. Timely antibiotic administration in adult septic shock has also been recognized as a determinant of mortality as reported by Kumar et al. [7] in 2006. Kumar et al. [7] showed with every hour delay in antibiotics from the time of hypotension, there was a 7.6% increase in mortality. Recently, in children with severe sepsis, Weiss et al. [8] showed that delays in antibiotic administration above 3 h was associated with increased mortality and longer duration of organ dysfunction. In adults, if a central line has been placed, the surviving sepsis guidelines define fluid refractory shock (FRS) as having reached a central venous pressure (CVP) of 8–12 mmH2O with persistent
signs of shock [9]. Since central access in children is technically not as easy and may require sedation, it is understandable that CVP cannot be an appropriate endpoint for fluid resuscitation. Thus, FRS is defined in a child who has received greater than 60 ml/kg of isotonic fluid with persistent signs of shock. Despite these definitions, timely initiation of a vasopressor has not been clearly defined in pediatrics. Beck et al. [10] looked at timing of vasopressor initiation in adults and found a very weak association of time to increased mortality [adjusted odds ratio (OR) 1.02/h, 95% confidence interval (CI) 1.01–1.03, P < 0.001]. There is no pediatric literature on the impact of vasopressor initiation timeliness from the time of last fluid completion, though Paul et al. [6] a priori defined nonadherence if initiation was for more than 15 min from the first documented hypotension after administration of 60 ml/kg of crystalloid fluid. Unfortunately, this study did not specifically report if delays in vasopressor initiation were specifically associated with worse outcomes. Finally, though the pediatric sepsis guidelines make recommendations for vasopressor selection depending on clinically suspected hemodynamic derangement [11], quality outcomes have not yet been developed targeting appropriate vasopressor selection. In a quality improvement project to improve overall sepsis bundle compliance, Paul et al. [12 ] demonstrated that by performing rapid cycles of improvement with specific processes, bundle compliance reached 100% and a reduction in mortality, as detailed by an increase in the number of septic shock cases managed in the ED in between deaths due to sepsis. &&
BARRIERS TO IMPLEMENTATION OR ADHERENCE Understanding local barriers to implementation of sepsis quality improvement are imperative before adapting the latest intervention determined to be effective in improving sepsis outcomes. Otherwise, considerable effort can go into an initiative with little impact. Specifically, a multidisciplinary approach must be developed in order to implement quality improvement interventions. Burney et al. [13] assessed perceived causes of treatment delays of severe sepsis in adults from the perspective of both nurses (RN) and physicians (MD). She found that a delay in diagnosis by physicians (RN 28.1%, MD 6.8%), lack of recognition at triage (RN 19.3%, MD 20.5%), knowledge deficits (RN 14.0%, MD 2.3%), and nursing delays (RN 7.0%, MD 20.5%) were perceived as contributors [13]. This study shows the discordant perceptions between nurses and physicians, but also highlights opportunities for
1040-8703 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-pediatrics.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
299
Emergency and critical care medicine
multidisciplinary collaboration to improve sepsis management. Two recent adult sepsis studies from the Protocol-Based Care for Early Septic Shock (ProCESS) investigators and the Australasian Resuscitation in Sepsis Evaluation investigators showed that protocolled care was not superior to standard of care [14,15]. However, these studies should not be interpreted as a lack of need for protocolled care, since patients randomized to non-protocolled care also received timely interventions. This is likely due to study inclusion of academic centers where standardized sepsis management may already comply with surviving sepsis campaign guidelines [9]. In addition, both of these studies fail to account for a major determinant of sepsis outcomes – timely recognition. Prompt treatment can only begin once sepsis is recognized, and inherent to any randomized controlled trial, enrollment can only occur after recognition. In fact, those with clinically evident signs of organ dysfunction are more likely to have increased compliance with sepsis resuscitation bundles as compared to those with isolated biochemical signs, such as lactic academia [16]. As a result, quality improvement initiatives must ensure to implement strategies to recognize abnormal vital signs of sepsis to facilitate timely management. Through root cause analysis of critical events, Cruz et al. [17] identified recognition as a major barrier in the implementation of sepsis processes in a pediatric ED. As a result, they introduced a sepsis trigger tool to be used at triage to promptly recognize vital sign abnormalities and bring resources to the bedside. Cruz et al. [18] eventually implemented an automated age and temperatureadjusted tachycardia alert which identified 81% of children with shock. However, the tool triggered about 25 times a day for a total of 4553 triggers during the study period to identify only 170 patients (positive predictive value of only 3.7%). Twentynine sepsis patients were missed by the tool, though 11 were identified by the need for immediate resuscitation. Sepanski et al. [19] also designed and tested a multivariable screening tool which took into account not only systemic inflammatory response syndrome (SIRS) criteria but also laboratory findings of organ dysfunction. This screening tool was able to achieve a 48.7% positive predictive value. A potential barrier to the sustained effect of these interventions is that they may lead to alarm fatigue, resulting in degradation of effect over time, especially if they are not specific and lead to a resource burden. However, such trigger or screening tools maximize the chance that a child can be identified as soon as possible across multiple systems of care with variable pediatric expertise. 300
www.co-pediatrics.com
Paul et al. [6] also looked at barriers to adherence to sepsis guidelines and attempted to address whether delay in recognition or timely vascular access was the reason for delay in fluid and antibiotic administration. In the 126 children with septic shock, 79% were recognized as meeting criteria for sepsis within 5 min. In addition, 68% had no delay in vascular access. In isolating the 73 patients who had neither delay in recognition nor delay in vascular access, only 29% received 60 ml/kg within 60 min. Of note, if a delay occurred in both the recognition and vascular access, 0% achieved timely fluids. Although this was studied at a single pediatric ED, these findings suggest that delays in recognition and vascular access contribute to inadequate fluid resuscitation, but other factors must be investigated to understand nonadherence to sepsis protocols.
CONTROVERSIES/DIFFICULTIES OF INTERPRETING LITERATURE In interpreting both adult and pediatric severe sepsis and quality improvement literature, attention is required to define sepsis onset. This is key in retrospective studies in which case the medical record may have incomplete data. Time zero, based on the established criteria of sepsis, should be the gold standard by which we measure all subsequent timeliness measures. For many adult ED-based sepsis studies, sepsis onset, or time zero, is defined as the time the patient presents to the ED. This definition can be extrapolated for pediatric sepsis, but there is no current consensus that supports this. A potential problem with using time zero at the time of entry to the ED in pediatric sepsis is that children can present with just fever without extreme tachycardia, and subsequently evolve while in ED. In addition, this definition neglects hospitalized patients who develop signs and symptoms of sepsis, and thus have no ‘time of entry’ to define sepsis onset. Thus, the variation in time zero in the literature makes interpretation of outcomes difficult to generalize. Ideally, a screening or trigger tool which immediately notifies a provider of the specific criteria to allow recognition is a good solution; however, the Cruz et al. trigger tool and the Sepanski et al. screening tool both can be limited by their generalizability to institutions that do not use the same electronic medical record system [17–19]. In adult studies outside the ED, sepsis onset is either not well defined [20], or time of hypotension or lactate level was used to determine onset [21–23]. For inpatient studies, presumably, the patient showed signs of sepsis before hypotension or before a healthcare provider was prompted to send a lactate. By waiting for hypotension to define time zero, Volume 27 Number 3 June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Quality improvement in pediatric sepsis Melendez and Bachur
there is a lost opportunity to understand if an earlier intervention could have improved outcomes. In pediatric studies, sepsis onset has also varied. Bissinger et al. [24] looked at the antibiotic administration timing in neonates with suspected hospital-acquired infection. Time zero was defined as the time when the decision to perform a sepsis evaluation occurred; however, this decision failed to account for delays from criteria to recognition, or from recognition to other important process measures. In the Weiss et al. [8] study assessing antibiotic timing in children with sepsis, time zero was defined in two ways: for pediatric ED patients, time zero was the time of ED entry, and for the general pediatric ward patients, time zero was the time that the first sepsis intervention was performed. The patients were combined for analysis and unfortunately this confounded the understanding of timeliness. Finally, a major confounder in the pediatric literature on sepsis outcomes is the absence of a severity score that can define overall risk of mortality that would allow a comparison of patient populations among studies. The adult literature may use the Predisposition Insult Response and Organ failure (PIRO) scoring system or the Mortality in Emergency Department Sepsis (MEDS) score [25]. Both have been shown to predict mortality; however, a similar severity scoring system does not exist for children with sepsis. In addition, since mortality is currently reported in studies carried out at large academic referral centers which are likely to have more children with chronic care conditions, reported mortality cannot be generalized.
SUMMARY AND FUTURE DIRECTIONS Despite the lack of uniform definitions of sepsis onset and no uniform reporting of timeliness of subsequent processes, the most current literature denotes that children with severe sepsis are at high risk of mortality (10.8–21.1%) [1 ,3,8], are more likely to have comorbidities (70.3–81.0%) [1 ,8], and are more likely to have longer hospital and ICU lengths of stay [6]. These outcomes are worsened when sepsis management guidelines are not followed, but comparing time-efficiency interventions among studies is difficult due to variation in time zero definitions. The literature suggests that the same process measures in adult sepsis are appropriate targets for timeliness in pediatrics. In time, the pediatric sepsis literature will evolve and should help determine the optimal amount of fluid resuscitation that should be given, and this can help determine additional process measures, such as timeliness to pressors in FRS. In addition, &
&
measurement of appropriate pressor selection needs to be studied and related to outcomes. Finally, a sepsis severity score is needed to ensure that the severity of illness among pediatric sepsis studies is comparable.
CONCLUSION Pediatric sepsis is a prevalent disease that has been defined by consensus guidelines [4]. These definitions serve as a reference, but are difficult to apply prospectively. One possible definition that can be applied across treatment settings is to define sepsis onset as the time the patient first meets criteria for SIRS before progressing to severe sepsis/shock, which is the earliest time a clinical team could have recognized a patient at risk for severe sepsis. We expect mortality will continue to improve in pediatric sepsis; as a result, meaningful outcomes should also include hospital and ICU length of stay. Finally, to benchmark any of these outcomes, a severity score will need to be developed to include comorbidities and severity at the time of presentation. Acknowledgements None. Financial support and sponsorship None. Conflicts of interest There are no conflicts of interest.
REFERENCES AND RECOMMENDED READING Papers of particular interest, published within the annual period of review, have been highlighted as: & of special interest && of outstanding interest 1. Balamuth F, Weiss SL, Neuman MI, et al. Pediatric severe sepsis in U.S. & children’s hospitals. Pediatr Crit Care Med 2014; 15:798–805. Highlights an alternative way to possibly be more inclusive diagnostically when retrospectively trying to identify children with severe sepsis. 2. Ruth A, McCracken CE, Fortenberry JD, et al. Pediatric severe sepsis: current & trends and outcomes from the Pediatric Health Information Systems database. Pediatr Crit Care Med 2014; 15:828–838. Highlights an alternative way to possibly be more inclusive diagnostically when retrospectively trying to identify children with severe sepsis. 3. Watson RS, Carcillo JA, Linde-Zwirble WT, et al. The epidemiology of severe sepsis in children in the United States. Am J Respir Crit Care Med 2003; 167:695–701. 4. Goldstein B, Giroir B, Randolph A, et al. International pediatric sepsis consensus conference: definitions for sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med 2005; 6:2–8. 5. Carcillo JA, Davis AL, Zaritsky A. Role of early fluid resuscitation in pediatric septic shock. J Am Med Assoc 1991; 266:1242–1245. 6. Paul R, Neuman MI, Monuteaux MC, et al. Adherence to PALS sepsis guidelines and hospital length of stay. Pediatrics 2012; 130:e273–e280. 7. Kumar A, Roberts D, Wood KE, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med 2006; 34:1589–1596.
1040-8703 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-pediatrics.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
301
Emergency and critical care medicine 8. Weiss SL, Fitzgerald JC, Balamuth F, et al. Delayed antimicrobial therapy increases mortality and organ dysfunction duration in pediatric sepsis. Crit Care Med 2014; 42:2409–2417. 9. Dellinger RP, Levy MM, Rhodes A, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012. Intensive Care Med 2013; 39:165–228. 10. Beck V, Chateau D, Bryson GL, et al. Timing of vasopressor initiation and mortality in septic shock: a cohort study. Crit Care 2014; 18:R97. 11. Brierley J, Carcillo JA, Choong K, et al. Clinical practice parameters for hemodynamic support of pediatric and neonatal septic shock: 2007 update from the American College of Critical Care Medicine. Crit Care Med 2009; 37:666–688. 12. Paul R, Melendez E, Stack A, et al. Improving adherence to PALS septic shock && guidelines. Pediatrics 2014; 133:e1358–e1366. Study shows how a multidisciplinary intervention to improve sepsis management bundle compliance can reach high reliability and reduce mortality due to sepsis. 13. Burney M, Underwood J, McEvoy S, et al. Early detection and treatment of severe sepsis in the emergency department: identifying barriers to implementation of a protocol-based approach. J Emerg Nurs 2012; 38:512–517. 14. ARISE Investigators; ANZICS Clinical Trials Group. Peake SL, Delaney A, Bailey M, et al. Goal-directed resuscitation for patients with early septic shock. N Engl J Med 2014; 371:1496–1506. 15. ProCESS Investigators. Yealy DM, Kellum JA, Huang DT, et al. A randomized trial of protocol-based care for early septic shock. N Engl J Med 2014; 370:1683–1693. 16. Kakebeeke D, Vis A, de Deckere ER, et al. Lack of clinically evident signs of organ failure affects ED treatment of patients with severe sepsis. Int J Emerg Med 2013; 6:4.
302
www.co-pediatrics.com
17. Cruz AT, Perry AM, Williams EA, et al. Implementation of goal-directed therapy for children with suspected sepsis in the emergency department. Pediatrics 2011; 127:e758–e766. 18. Cruz AT, Williams EA, Graf JM, et al. Test characteristics of an automated ageand temperature-adjusted tachycardia alert in pediatric septic shock. Pediatr Emerg Care 2012; 28:889–894. 19. Sepanski RJ, Godambe SA, Mangum CD, et al. Designing a pediatric severe sepsis screening tool. Front Pediatr 2014; 2:56. 20. Seoane L, Winterbottom F, Nash T, et al. Using quality improvement principles to improve the care of patients with severe sepsis and septic shock. Ochsner J 2013; 13:359–366. 21. Yokota PK, Marra AR, Martino MDV, et al. Impact of appropriate antimicrobial therapy for patients with severe sepsis and septic shock - a quality improvement study. PLoS One 2014; 9:e104475. 22. Mok K, Christian MD, Nelson S, et al. Time to administration of antibiotics among inpatients with severe sepsis or septic shock. Can J Hosp Pharm 2014; 67:213–219. 23. Miller RR, Dong L, Nelson NC, et al. Multicenter implementation of a severe sepsis and septic shock treatment bundle. Am J Respir Crit Care Med 2013; 188:77–82. 24. Bissinger RL, Mueller M, Cox TH, et al. Antibiotic timing in neonates with suspected hospital-acquired infections. Adv Neonatal Care 2013; 13:22– 28. (quiz 29–30). 25. Macdonald SP, Arendts G, Fatovich DM, et al. Comparison of PIRO, SOFA, and MEDS scores for predicting mortality in emergency department patients with severe sepsis and septic shock. Acad Emerg Med 2014; 21:1257– 1263.
Volume 27 Number 3 June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Quality improvement in pediatric sepsis Elliot Melendez a,b and Richard Bachur b
Purpose of review Although there is abundant literature detailing the impact of quality improvement in adult sepsis, the pediatric literature is lacking. Despite consensus definitions for sepsis, which patients along the sepsis spectrum should receive aggressive management and the exact onset of sepsis (‘time zero’) are not clearly established. In the adult emergency department (ED), sepsis onset is defined as the time of entry into the ED; however, this definition cannot be applied to hospitalized patients or patients who evolve during their ED course. Since the time of sepsis onset will dictate the timeliness of subsequent process measures, the variable definitions in the literature make it difficult to generalize findings among prior studies. Recent findings Despite the variation in defining time zero, aggressive fluid administration, timely antibiotics, and compliance with sepsis bundles have been shown to improve mortality and to reduce hospital and intensive care length of stay. In addition, early identification tools show promise in beginning to define sepsis onset and retrospective search tools may allow improved case finding of those children of concern for sepsis. Summary Quality improvement in pediatric sepsis is evolving. As we continue to define quality measures, we must standardize the definition of sepsis onset. This definition should be applicable to any treatment venue to ensure measures can be evaluated across all settings. In addition, we must delineate which patients along the sepsis spectrum should be candidates for timely interventions and standardize other outcome measures beyond mortality. Keywords pediatric, quality improvement, sepsis
INTRODUCTION Pediatric severe sepsis and septic shock is a leading cause of morbidity and mortality, and contributes to high healthcare costs [1 ,2 ]. However, understanding the true morbidity and mortality is challenging since there may be variable definitions for sepsis in both the research and quality improvement literature, and uniformity does not exist. Using the Pediatric Health Information Systems (PHIS) database, Balamuth et al. found that the incidence of pediatric severe sepsis increased over a 9-year period, but mortality decreased [1 ]. However, in selecting cases coded as sepsis by the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), the overall mortality for sepsis was 21.2%, which is more than twice the last estimate provided by Watson et al. [3]. Ruth et al. [2 ] looked at children admitted to a pediatric ICU in the same cohort of PHIS patients. Using a different definition, in which severe sepsis was assigned if an ICD9-CM code for infection was associated with one organ dysfunction, mortality was 14.4%. Though this methodology will identify more &
&
&
&
www.co-pediatrics.com
patients with risk of severe sepsis, the current consensus definition by Goldstein et al. [4] defines severe sepsis as two or more organ dysfunction in the setting of clinical sepsis. Both of these studies published in the same edition of ‘Pediatric Critical Care Medicine’ (November 2014) highlight the difficulty in accessing outcomes if considering quality improvement initiatives, because the definitions of whom is included can vary. Using uniform definitions, especially those that can be assigned prospectively, will allow for better measurement of outcomes, identify barriers to adherence with
a
Division of Medicine Critical Care and bDivision of Emergency Medicine, Boston Children’s Hospital, Boston, Massachusetts, USA Correspondence to Elliot Melendez, MD, Associate Director, Safety and Quality, Division of Medicine Critical Care, Sepsis Quality Leader, Division of Emergency Medicine, Boston Children’s Hospital, 300 Longwood Avenue, Main 11 South, Boston, MA 02115, USA. Tel: +1 617 355 4522; fax: +1 888 883 9238; e-mail: [email protected]. edu Curr Opin Pediatr 2015, 27:298–302 DOI:10.1097/MOP.0000000000000222 Volume 27 Number 3 June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Quality improvement in pediatric sepsis Melendez and Bachur
KEY POINTS Time zero, the time by which the onset of sepsis occurs, should be uniform in studies to allow appropriate comparisons of process and outcome measures. Process measures should be assessed and any quality improvement interventions should target local barriers. Outcome measures such as mortality should always be assessed, but intermediate outcomes, such as hospital and ICU length of stay, are meaningful outcomes. Sepsis quality improvement in pediatrics should be benchmarked across institutions while accounting for severity of illness or complex care conditions. A severity of illness score is needed to insure we understand risk as related to quality improvement interventions.
quality improvement processes, and clarify the controversies in the sepsis literature.
OUTCOMES RELATED TO SPECIFIC INTERVENTIONS IN PEDIATRIC SEVERE SEPSIS Ever since Carcillo et al. [5] showed that children who received less than 40 ml/kg of isotonic fluids during management for septic shock had a higher mortality rate, it has been recognized that inadequate fluid resuscitation is a predictor of worse outcomes. Paul et al. [6] reported the incidence of low compliance with septic shock processes as recommended by the Pediatric Advanced Life Support Sepsis Guidelines in 126 children who presented to a pediatric emergency department (ED) with severe sepsis and septic shock. Specifically, timely fluid administration of at least 60 ml/kg of isotonic fluid within the first hour of shock only occurred in 37% of the children. Nonadherence with timely fluid administration was statistically associated with both a longer ICU stay and a hospital length of stay [6]. Timely antibiotic administration in adult septic shock has also been recognized as a determinant of mortality as reported by Kumar et al. [7] in 2006. Kumar et al. [7] showed with every hour delay in antibiotics from the time of hypotension, there was a 7.6% increase in mortality. Recently, in children with severe sepsis, Weiss et al. [8] showed that delays in antibiotic administration above 3 h was associated with increased mortality and longer duration of organ dysfunction. In adults, if a central line has been placed, the surviving sepsis guidelines define fluid refractory shock (FRS) as having reached a central venous pressure (CVP) of 8–12 mmH2O with persistent
signs of shock [9]. Since central access in children is technically not as easy and may require sedation, it is understandable that CVP cannot be an appropriate endpoint for fluid resuscitation. Thus, FRS is defined in a child who has received greater than 60 ml/kg of isotonic fluid with persistent signs of shock. Despite these definitions, timely initiation of a vasopressor has not been clearly defined in pediatrics. Beck et al. [10] looked at timing of vasopressor initiation in adults and found a very weak association of time to increased mortality [adjusted odds ratio (OR) 1.02/h, 95% confidence interval (CI) 1.01–1.03, P < 0.001]. There is no pediatric literature on the impact of vasopressor initiation timeliness from the time of last fluid completion, though Paul et al. [6] a priori defined nonadherence if initiation was for more than 15 min from the first documented hypotension after administration of 60 ml/kg of crystalloid fluid. Unfortunately, this study did not specifically report if delays in vasopressor initiation were specifically associated with worse outcomes. Finally, though the pediatric sepsis guidelines make recommendations for vasopressor selection depending on clinically suspected hemodynamic derangement [11], quality outcomes have not yet been developed targeting appropriate vasopressor selection. In a quality improvement project to improve overall sepsis bundle compliance, Paul et al. [12 ] demonstrated that by performing rapid cycles of improvement with specific processes, bundle compliance reached 100% and a reduction in mortality, as detailed by an increase in the number of septic shock cases managed in the ED in between deaths due to sepsis. &&
BARRIERS TO IMPLEMENTATION OR ADHERENCE Understanding local barriers to implementation of sepsis quality improvement are imperative before adapting the latest intervention determined to be effective in improving sepsis outcomes. Otherwise, considerable effort can go into an initiative with little impact. Specifically, a multidisciplinary approach must be developed in order to implement quality improvement interventions. Burney et al. [13] assessed perceived causes of treatment delays of severe sepsis in adults from the perspective of both nurses (RN) and physicians (MD). She found that a delay in diagnosis by physicians (RN 28.1%, MD 6.8%), lack of recognition at triage (RN 19.3%, MD 20.5%), knowledge deficits (RN 14.0%, MD 2.3%), and nursing delays (RN 7.0%, MD 20.5%) were perceived as contributors [13]. This study shows the discordant perceptions between nurses and physicians, but also highlights opportunities for
1040-8703 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-pediatrics.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
299
Emergency and critical care medicine
multidisciplinary collaboration to improve sepsis management. Two recent adult sepsis studies from the Protocol-Based Care for Early Septic Shock (ProCESS) investigators and the Australasian Resuscitation in Sepsis Evaluation investigators showed that protocolled care was not superior to standard of care [14,15]. However, these studies should not be interpreted as a lack of need for protocolled care, since patients randomized to non-protocolled care also received timely interventions. This is likely due to study inclusion of academic centers where standardized sepsis management may already comply with surviving sepsis campaign guidelines [9]. In addition, both of these studies fail to account for a major determinant of sepsis outcomes – timely recognition. Prompt treatment can only begin once sepsis is recognized, and inherent to any randomized controlled trial, enrollment can only occur after recognition. In fact, those with clinically evident signs of organ dysfunction are more likely to have increased compliance with sepsis resuscitation bundles as compared to those with isolated biochemical signs, such as lactic academia [16]. As a result, quality improvement initiatives must ensure to implement strategies to recognize abnormal vital signs of sepsis to facilitate timely management. Through root cause analysis of critical events, Cruz et al. [17] identified recognition as a major barrier in the implementation of sepsis processes in a pediatric ED. As a result, they introduced a sepsis trigger tool to be used at triage to promptly recognize vital sign abnormalities and bring resources to the bedside. Cruz et al. [18] eventually implemented an automated age and temperatureadjusted tachycardia alert which identified 81% of children with shock. However, the tool triggered about 25 times a day for a total of 4553 triggers during the study period to identify only 170 patients (positive predictive value of only 3.7%). Twentynine sepsis patients were missed by the tool, though 11 were identified by the need for immediate resuscitation. Sepanski et al. [19] also designed and tested a multivariable screening tool which took into account not only systemic inflammatory response syndrome (SIRS) criteria but also laboratory findings of organ dysfunction. This screening tool was able to achieve a 48.7% positive predictive value. A potential barrier to the sustained effect of these interventions is that they may lead to alarm fatigue, resulting in degradation of effect over time, especially if they are not specific and lead to a resource burden. However, such trigger or screening tools maximize the chance that a child can be identified as soon as possible across multiple systems of care with variable pediatric expertise. 300
www.co-pediatrics.com
Paul et al. [6] also looked at barriers to adherence to sepsis guidelines and attempted to address whether delay in recognition or timely vascular access was the reason for delay in fluid and antibiotic administration. In the 126 children with septic shock, 79% were recognized as meeting criteria for sepsis within 5 min. In addition, 68% had no delay in vascular access. In isolating the 73 patients who had neither delay in recognition nor delay in vascular access, only 29% received 60 ml/kg within 60 min. Of note, if a delay occurred in both the recognition and vascular access, 0% achieved timely fluids. Although this was studied at a single pediatric ED, these findings suggest that delays in recognition and vascular access contribute to inadequate fluid resuscitation, but other factors must be investigated to understand nonadherence to sepsis protocols.
CONTROVERSIES/DIFFICULTIES OF INTERPRETING LITERATURE In interpreting both adult and pediatric severe sepsis and quality improvement literature, attention is required to define sepsis onset. This is key in retrospective studies in which case the medical record may have incomplete data. Time zero, based on the established criteria of sepsis, should be the gold standard by which we measure all subsequent timeliness measures. For many adult ED-based sepsis studies, sepsis onset, or time zero, is defined as the time the patient presents to the ED. This definition can be extrapolated for pediatric sepsis, but there is no current consensus that supports this. A potential problem with using time zero at the time of entry to the ED in pediatric sepsis is that children can present with just fever without extreme tachycardia, and subsequently evolve while in ED. In addition, this definition neglects hospitalized patients who develop signs and symptoms of sepsis, and thus have no ‘time of entry’ to define sepsis onset. Thus, the variation in time zero in the literature makes interpretation of outcomes difficult to generalize. Ideally, a screening or trigger tool which immediately notifies a provider of the specific criteria to allow recognition is a good solution; however, the Cruz et al. trigger tool and the Sepanski et al. screening tool both can be limited by their generalizability to institutions that do not use the same electronic medical record system [17–19]. In adult studies outside the ED, sepsis onset is either not well defined [20], or time of hypotension or lactate level was used to determine onset [21–23]. For inpatient studies, presumably, the patient showed signs of sepsis before hypotension or before a healthcare provider was prompted to send a lactate. By waiting for hypotension to define time zero, Volume 27 Number 3 June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Quality improvement in pediatric sepsis Melendez and Bachur
there is a lost opportunity to understand if an earlier intervention could have improved outcomes. In pediatric studies, sepsis onset has also varied. Bissinger et al. [24] looked at the antibiotic administration timing in neonates with suspected hospital-acquired infection. Time zero was defined as the time when the decision to perform a sepsis evaluation occurred; however, this decision failed to account for delays from criteria to recognition, or from recognition to other important process measures. In the Weiss et al. [8] study assessing antibiotic timing in children with sepsis, time zero was defined in two ways: for pediatric ED patients, time zero was the time of ED entry, and for the general pediatric ward patients, time zero was the time that the first sepsis intervention was performed. The patients were combined for analysis and unfortunately this confounded the understanding of timeliness. Finally, a major confounder in the pediatric literature on sepsis outcomes is the absence of a severity score that can define overall risk of mortality that would allow a comparison of patient populations among studies. The adult literature may use the Predisposition Insult Response and Organ failure (PIRO) scoring system or the Mortality in Emergency Department Sepsis (MEDS) score [25]. Both have been shown to predict mortality; however, a similar severity scoring system does not exist for children with sepsis. In addition, since mortality is currently reported in studies carried out at large academic referral centers which are likely to have more children with chronic care conditions, reported mortality cannot be generalized.
SUMMARY AND FUTURE DIRECTIONS Despite the lack of uniform definitions of sepsis onset and no uniform reporting of timeliness of subsequent processes, the most current literature denotes that children with severe sepsis are at high risk of mortality (10.8–21.1%) [1 ,3,8], are more likely to have comorbidities (70.3–81.0%) [1 ,8], and are more likely to have longer hospital and ICU lengths of stay [6]. These outcomes are worsened when sepsis management guidelines are not followed, but comparing time-efficiency interventions among studies is difficult due to variation in time zero definitions. The literature suggests that the same process measures in adult sepsis are appropriate targets for timeliness in pediatrics. In time, the pediatric sepsis literature will evolve and should help determine the optimal amount of fluid resuscitation that should be given, and this can help determine additional process measures, such as timeliness to pressors in FRS. In addition, &
&
measurement of appropriate pressor selection needs to be studied and related to outcomes. Finally, a sepsis severity score is needed to ensure that the severity of illness among pediatric sepsis studies is comparable.
CONCLUSION Pediatric sepsis is a prevalent disease that has been defined by consensus guidelines [4]. These definitions serve as a reference, but are difficult to apply prospectively. One possible definition that can be applied across treatment settings is to define sepsis onset as the time the patient first meets criteria for SIRS before progressing to severe sepsis/shock, which is the earliest time a clinical team could have recognized a patient at risk for severe sepsis. We expect mortality will continue to improve in pediatric sepsis; as a result, meaningful outcomes should also include hospital and ICU length of stay. Finally, to benchmark any of these outcomes, a severity score will need to be developed to include comorbidities and severity at the time of presentation. Acknowledgements None. Financial support and sponsorship None. Conflicts of interest There are no conflicts of interest.
REFERENCES AND RECOMMENDED READING Papers of particular interest, published within the annual period of review, have been highlighted as: & of special interest && of outstanding interest 1. Balamuth F, Weiss SL, Neuman MI, et al. Pediatric severe sepsis in U.S. & children’s hospitals. Pediatr Crit Care Med 2014; 15:798–805. Highlights an alternative way to possibly be more inclusive diagnostically when retrospectively trying to identify children with severe sepsis. 2. Ruth A, McCracken CE, Fortenberry JD, et al. Pediatric severe sepsis: current & trends and outcomes from the Pediatric Health Information Systems database. Pediatr Crit Care Med 2014; 15:828–838. Highlights an alternative way to possibly be more inclusive diagnostically when retrospectively trying to identify children with severe sepsis. 3. Watson RS, Carcillo JA, Linde-Zwirble WT, et al. The epidemiology of severe sepsis in children in the United States. Am J Respir Crit Care Med 2003; 167:695–701. 4. Goldstein B, Giroir B, Randolph A, et al. International pediatric sepsis consensus conference: definitions for sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med 2005; 6:2–8. 5. Carcillo JA, Davis AL, Zaritsky A. Role of early fluid resuscitation in pediatric septic shock. J Am Med Assoc 1991; 266:1242–1245. 6. Paul R, Neuman MI, Monuteaux MC, et al. Adherence to PALS sepsis guidelines and hospital length of stay. Pediatrics 2012; 130:e273–e280. 7. Kumar A, Roberts D, Wood KE, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med 2006; 34:1589–1596.
1040-8703 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-pediatrics.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
301
Emergency and critical care medicine 8. Weiss SL, Fitzgerald JC, Balamuth F, et al. Delayed antimicrobial therapy increases mortality and organ dysfunction duration in pediatric sepsis. Crit Care Med 2014; 42:2409–2417. 9. Dellinger RP, Levy MM, Rhodes A, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012. Intensive Care Med 2013; 39:165–228. 10. Beck V, Chateau D, Bryson GL, et al. Timing of vasopressor initiation and mortality in septic shock: a cohort study. Crit Care 2014; 18:R97. 11. Brierley J, Carcillo JA, Choong K, et al. Clinical practice parameters for hemodynamic support of pediatric and neonatal septic shock: 2007 update from the American College of Critical Care Medicine. Crit Care Med 2009; 37:666–688. 12. Paul R, Melendez E, Stack A, et al. Improving adherence to PALS septic shock && guidelines. Pediatrics 2014; 133:e1358–e1366. Study shows how a multidisciplinary intervention to improve sepsis management bundle compliance can reach high reliability and reduce mortality due to sepsis. 13. Burney M, Underwood J, McEvoy S, et al. Early detection and treatment of severe sepsis in the emergency department: identifying barriers to implementation of a protocol-based approach. J Emerg Nurs 2012; 38:512–517. 14. ARISE Investigators; ANZICS Clinical Trials Group. Peake SL, Delaney A, Bailey M, et al. Goal-directed resuscitation for patients with early septic shock. N Engl J Med 2014; 371:1496–1506. 15. ProCESS Investigators. Yealy DM, Kellum JA, Huang DT, et al. A randomized trial of protocol-based care for early septic shock. N Engl J Med 2014; 370:1683–1693. 16. Kakebeeke D, Vis A, de Deckere ER, et al. Lack of clinically evident signs of organ failure affects ED treatment of patients with severe sepsis. Int J Emerg Med 2013; 6:4.
302
www.co-pediatrics.com
17. Cruz AT, Perry AM, Williams EA, et al. Implementation of goal-directed therapy for children with suspected sepsis in the emergency department. Pediatrics 2011; 127:e758–e766. 18. Cruz AT, Williams EA, Graf JM, et al. Test characteristics of an automated ageand temperature-adjusted tachycardia alert in pediatric septic shock. Pediatr Emerg Care 2012; 28:889–894. 19. Sepanski RJ, Godambe SA, Mangum CD, et al. Designing a pediatric severe sepsis screening tool. Front Pediatr 2014; 2:56. 20. Seoane L, Winterbottom F, Nash T, et al. Using quality improvement principles to improve the care of patients with severe sepsis and septic shock. Ochsner J 2013; 13:359–366. 21. Yokota PK, Marra AR, Martino MDV, et al. Impact of appropriate antimicrobial therapy for patients with severe sepsis and septic shock - a quality improvement study. PLoS One 2014; 9:e104475. 22. Mok K, Christian MD, Nelson S, et al. Time to administration of antibiotics among inpatients with severe sepsis or septic shock. Can J Hosp Pharm 2014; 67:213–219. 23. Miller RR, Dong L, Nelson NC, et al. Multicenter implementation of a severe sepsis and septic shock treatment bundle. Am J Respir Crit Care Med 2013; 188:77–82. 24. Bissinger RL, Mueller M, Cox TH, et al. Antibiotic timing in neonates with suspected hospital-acquired infections. Adv Neonatal Care 2013; 13:22– 28. (quiz 29–30). 25. Macdonald SP, Arendts G, Fatovich DM, et al. Comparison of PIRO, SOFA, and MEDS scores for predicting mortality in emergency department patients with severe sepsis and septic shock. Acad Emerg Med 2014; 21:1257– 1263.
Volume 27 Number 3 June 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
