Amit Uppal, MD

 

In 2002, the European Society of Intensive Care Medicine, the Society of Critical Care Medicine, and the International Sepsis Forum formed the Surviving Sepsis Campaign (SSC) aiming to reduce sepsis-related mortality by 25% within 5 years, mimicking the progress made in the management of STEMI (http://www.survivingsepsis.org/About-SSC/Pages/History.aspx).

SSC bundles: a historic perspective

The first guidelines were published in 2004. Recognizing that guidelines may not influence bedside practice for many years, the SSC partnered with the Institute for Healthcare Improvement to apply performance improvement methodology to sepsis management, developing the “sepsis change bundles.” In addition to hospital resources for education, screening, and data collection, the 6-hour resuscitation and 24-hour management bundles were created. Subsequent data, collected as part of the initiative, demonstrated an association between bundle compliance and survival.

In 2008, the SSC guidelines were revised, and the National Quality Forum (NQF) adopted sepsis bundle compliance as a quality measure. NQF endorsement is often the first step toward the creation of mandates by the Centers for Medicare and Medicaid Services (CMS), but that did not occur at the time.

In 2012, the SSC guidelines were updated and published with new 3- and 6-hour bundles. That year, Rory Staunton, an otherwise healthy 12-year-old boy, died of septic shock in New York. The public discussion of this case, among other factors, prompted New York state to develop a sepsis care mandate that became state law in 2014. An annual public report details each hospital’s compliance with process measures and risk-adjusted mortality. The correlation between measure compliance and survival also holds true in this data set.

In 2015, CMS developed the SEP-1 measure. While the symbolic importance of a sepsis federal mandate and its potential to improve patient outcomes is recognized, concerns remain about the measure itself. The detailed and specific way data must be collected may disconnect clinical care provided from measured compliance. The time pressure and the “all-or-nothing” approach might incentivize interventions potentially harmful in some patients. No patient-centered outcomes are reported. This measure might be tied to reimbursement in the future.

The original version of SEP-1 was based on the 2012 SSC bundles, which reflected the best evidence available at the time (the 2001 Early Goal-Directed Therapy trial). By 2015, elements of that strategy had been challenged, and the PROCESS, PROMISE, and ARISE trials contested the notion that protcolized resuscitation decreased mortality. Moreover, new definitions of sepsis syndromes (Sepsis-3) were published in 2016 (Singer M, et al. JAMA. 2016;315[8]:801).

The 2016 SSC guidelines adopted the new definitions and recommended that patients with sepsis-induced hypoperfusion immediately receive a 30 mL/kg crystalloid bolus, followed by frequent reassessment. CMS did not adopt the Sepsis-3 definitions, but updates were made to allow the clinicians flexibility to demonstrate reassessment of the patient.

 

Comparing the 1-hour bundle to STEMI care

This year, the SSC published a 1-hour bundle to replace the 3- and 6-hour bundles (Levy MM et al. Crit Care Med. 2018;46[6]:997). Whereas previous bundles set time frames for completion of the elements, the 1-hour bundle focuses on the initiation of these components. The authors revisited the parallel between early management of sepsis and STEMI. The 1-hour bundle includes serum lactate, blood cultures prior to antibiotics, broad-spectrum antibiotics, a 30 mL/kg crystalloid bolus for patients with hypotension or lactate greater than or equal to 4 mmol/L, and vasopressors for persistent hypotension.

Elements of controversy after the publication of this bundle include:


1. One hour seems insufficient for complex clinical decision making and interventions for a syndrome with no specific diagnostic test: sepsis often mimics, or is mimicked by, other conditions.

2. Some bundle elements are not supported by high-quality evidence. No controlled studies exist regarding the appropriate volume of initial fluids or the impact of timing of antibiotics on outcomes.

3. The 1-hour time frame will encourage empiric delivery of fluids and antibiotics to patients who are not septic, potentially leading to harm.

4. While the 1-hour bundle is a quality improvement tool and not for public reporting, former bundles have been adopted as federally regulated measures.


Has the SSC gone too far? Are these concerns enough to abandon the 1-hour bundle? Or are the concerns regarding the 1-hour bundle an example of “perfect is the enemy of better”? To understand the potential for imperfect guidelines to drive tremendous patient-level improvements, one must consider the evolution of STEMI management.

Since the 1970s, the in-hospital mortality for STEMI has decreased from 25% to around 5%. The most significant factor in this achievement was the recognition that early reperfusion improves outcomes and that doing it consistently requires complex coordination. In 2004, a Door-to-Balloon (D2B) time of less than 90 minutes was included as a guideline recommendation (Antman EM, et al. Circulation. 2004;110[5]:588). CMS started collecting performance data on this metric, made that data public, and later tied the performance to hospital reimbursement.

Initially, the 90-minute goal was achieved in only 44% of cases. In 2006, the D2B initiative was launched, providing recommendations for public education, coordination of care, and emergent management of STEMI. Compliance with these recommendations required significant education and changes to STEMI care at multiple levels. Data were collected and submitted to inform the process. Six years later, compliance with the D2B goal had increased from 44% to 91%. The median D2B dropped from 96 to 64 minutes. Based on high compliance, CMS discontinued the use of this metric for reimbursement as the variation between high and low performers was minimal. Put simply, the entire country had gotten better at treating STEMI. The “time-zero” for STEMI was pushed back further, and D2B has been replaced with first-medical-contact (FMC) to device time. The recommendation is to achieve this as quickly as possible, and in less than 90 minutes (O’Gara P, et al. JACC. 2013;61[4]:485).

Consider the complexity of getting a patient from their home to a catheterization lab within 90 minutes, even in ideal circumstances. This short time frame encourages, by design, a low threshold to activate the system. We accept that some patients will receive an unnecessary catheterization or systemic fibrinolysis although the recommendation is based on level B evidence.

Compliance with the STEMI guidelines is more labor-intensive and complex than compliance with the 1-hour sepsis bundle. So, is STEMI a fair comparison to sepsis? Both syndromes are common, potentially deadly, and time-sensitive. Both require early recognition, but neither has a definitive diagnostic test. Instead, diagnosis requires an integration of multiple complex clinical factors. Both are backed by imperfect science that continues to evolve. Over-diagnosis of either will expose the patient to potentially harmful therapies.

The early management of STEMI is a valid comparison to the early management of sepsis. We must consider this comparison as we ponder the 1-hour sepsis bundle.

Is triage time the appropriate time-zero? In either condition, triage time is too early in some cases and too late in others. Unfortunately, there is no better alternative, and STEMI guidelines have evolved to start the clock before triage. Using a point such as “recognition of sepsis” would fail to capture delayed recognition.

Is it possible to diagnose and initiate treatment for sepsis in such a short time frame? Consider the treatment received by the usual care group of the PROCESS trial (The ProCESS Investigators. N Engl J Med. 2014;370:1683). Prior to meeting entry criteria, which occurred in less than 1 hour, patients in this group received an initial fluid bolus and had a lactate assessment. Prior to randomization, which occurred at around 90 minutes, this group completed 28 mL/kg of crystalloid fluid, and 76% received antibiotics. Thus, the usual-care group in this study nearly achieved the 1-hour bundle currently being contested.

Is it appropriate for a guideline to strongly recommend interventions not backed by level A evidence? The recommendation for FMC to catheterization within 90 minutes has not been studied in a controlled way. The precise dosing and timing of fibrinolysis is also not based on controlled data. Reperfusion devices and antiplatelet agents continue to be rigorously studied, sometimes with conflicting results.

Finally, should the 1-hour bundle be abandoned out of concern that it will be used as a national performance metric? First, there is currently no indication that the 1-hour bundle will be adopted as a performance metric. For the sake of argument, let’s assume the 1-hour bundle will be regulated and used to compare hospitals. Is there reason to think this bundle favors some hospitals over others and will lead to an unfair comparison? Is there significant inequity in the ability to draw blood cultures, send a lactate, start IV fluids, and initiate antibiotics?

Certainly, national compliance with such a metric would be very low at first. Therein lies the actual problem: a person who suffers a STEMI anywhere in the country is very likely to receive high-quality care. Currently, the same cannot be said about a patient with sepsis. Perhaps that should be the focus of our concern.


Dr. Uppal is Assistant Professor, NYU School of Medicine, Bellevue Hospital Center, New York, New York.
 

 

In 2002, the European Society of Intensive Care Medicine, the Society of Critical Care Medicine, and the International Sepsis Forum formed the Surviving Sepsis Campaign (SSC) aiming to reduce sepsis-related mortality by 25% within 5 years, mimicking the progress made in the management of STEMI (http://www.survivingsepsis.org/About-SSC/Pages/History.aspx).

SSC bundles: a historic perspective

The first guidelines were published in 2004. Recognizing that guidelines may not influence bedside practice for many years, the SSC partnered with the Institute for Healthcare Improvement to apply performance improvement methodology to sepsis management, developing the “sepsis change bundles.” In addition to hospital resources for education, screening, and data collection, the 6-hour resuscitation and 24-hour management bundles were created. Subsequent data, collected as part of the initiative, demonstrated an association between bundle compliance and survival.

In 2008, the SSC guidelines were revised, and the National Quality Forum (NQF) adopted sepsis bundle compliance as a quality measure. NQF endorsement is often the first step toward the creation of mandates by the Centers for Medicare and Medicaid Services (CMS), but that did not occur at the time.

In 2012, the SSC guidelines were updated and published with new 3- and 6-hour bundles. That year, Rory Staunton, an otherwise healthy 12-year-old boy, died of septic shock in New York. The public discussion of this case, among other factors, prompted New York state to develop a sepsis care mandate that became state law in 2014. An annual public report details each hospital’s compliance with process measures and risk-adjusted mortality. The correlation between measure compliance and survival also holds true in this data set.

In 2015, CMS developed the SEP-1 measure. While the symbolic importance of a sepsis federal mandate and its potential to improve patient outcomes is recognized, concerns remain about the measure itself. The detailed and specific way data must be collected may disconnect clinical care provided from measured compliance. The time pressure and the “all-or-nothing” approach might incentivize interventions potentially harmful in some patients. No patient-centered outcomes are reported. This measure might be tied to reimbursement in the future.

The original version of SEP-1 was based on the 2012 SSC bundles, which reflected the best evidence available at the time (the 2001 Early Goal-Directed Therapy trial). By 2015, elements of that strategy had been challenged, and the PROCESS, PROMISE, and ARISE trials contested the notion that protcolized resuscitation decreased mortality. Moreover, new definitions of sepsis syndromes (Sepsis-3) were published in 2016 (Singer M, et al. JAMA. 2016;315[8]:801).

The 2016 SSC guidelines adopted the new definitions and recommended that patients with sepsis-induced hypoperfusion immediately receive a 30 mL/kg crystalloid bolus, followed by frequent reassessment. CMS did not adopt the Sepsis-3 definitions, but updates were made to allow the clinicians flexibility to demonstrate reassessment of the patient.

 

Comparing the 1-hour bundle to STEMI care

This year, the SSC published a 1-hour bundle to replace the 3- and 6-hour bundles (Levy MM et al. Crit Care Med. 2018;46[6]:997). Whereas previous bundles set time frames for completion of the elements, the 1-hour bundle focuses on the initiation of these components. The authors revisited the parallel between early management of sepsis and STEMI. The 1-hour bundle includes serum lactate, blood cultures prior to antibiotics, broad-spectrum antibiotics, a 30 mL/kg crystalloid bolus for patients with hypotension or lactate greater than or equal to 4 mmol/L, and vasopressors for persistent hypotension.

Elements of controversy after the publication of this bundle include:


1. One hour seems insufficient for complex clinical decision making and interventions for a syndrome with no specific diagnostic test: sepsis often mimics, or is mimicked by, other conditions.

2. Some bundle elements are not supported by high-quality evidence. No controlled studies exist regarding the appropriate volume of initial fluids or the impact of timing of antibiotics on outcomes.

3. The 1-hour time frame will encourage empiric delivery of fluids and antibiotics to patients who are not septic, potentially leading to harm.

4. While the 1-hour bundle is a quality improvement tool and not for public reporting, former bundles have been adopted as federally regulated measures.


Has the SSC gone too far? Are these concerns enough to abandon the 1-hour bundle? Or are the concerns regarding the 1-hour bundle an example of “perfect is the enemy of better”? To understand the potential for imperfect guidelines to drive tremendous patient-level improvements, one must consider the evolution of STEMI management.

Since the 1970s, the in-hospital mortality for STEMI has decreased from 25% to around 5%. The most significant factor in this achievement was the recognition that early reperfusion improves outcomes and that doing it consistently requires complex coordination. In 2004, a Door-to-Balloon (D2B) time of less than 90 minutes was included as a guideline recommendation (Antman EM, et al. Circulation. 2004;110[5]:588). CMS started collecting performance data on this metric, made that data public, and later tied the performance to hospital reimbursement.

Initially, the 90-minute goal was achieved in only 44% of cases. In 2006, the D2B initiative was launched, providing recommendations for public education, coordination of care, and emergent management of STEMI. Compliance with these recommendations required significant education and changes to STEMI care at multiple levels. Data were collected and submitted to inform the process. Six years later, compliance with the D2B goal had increased from 44% to 91%. The median D2B dropped from 96 to 64 minutes. Based on high compliance, CMS discontinued the use of this metric for reimbursement as the variation between high and low performers was minimal. Put simply, the entire country had gotten better at treating STEMI. The “time-zero” for STEMI was pushed back further, and D2B has been replaced with first-medical-contact (FMC) to device time. The recommendation is to achieve this as quickly as possible, and in less than 90 minutes (O’Gara P, et al. JACC. 2013;61[4]:485).

Consider the complexity of getting a patient from their home to a catheterization lab within 90 minutes, even in ideal circumstances. This short time frame encourages, by design, a low threshold to activate the system. We accept that some patients will receive an unnecessary catheterization or systemic fibrinolysis although the recommendation is based on level B evidence.

Compliance with the STEMI guidelines is more labor-intensive and complex than compliance with the 1-hour sepsis bundle. So, is STEMI a fair comparison to sepsis? Both syndromes are common, potentially deadly, and time-sensitive. Both require early recognition, but neither has a definitive diagnostic test. Instead, diagnosis requires an integration of multiple complex clinical factors. Both are backed by imperfect science that continues to evolve. Over-diagnosis of either will expose the patient to potentially harmful therapies.

The early management of STEMI is a valid comparison to the early management of sepsis. We must consider this comparison as we ponder the 1-hour sepsis bundle.

Is triage time the appropriate time-zero? In either condition, triage time is too early in some cases and too late in others. Unfortunately, there is no better alternative, and STEMI guidelines have evolved to start the clock before triage. Using a point such as “recognition of sepsis” would fail to capture delayed recognition.

Is it possible to diagnose and initiate treatment for sepsis in such a short time frame? Consider the treatment received by the usual care group of the PROCESS trial (The ProCESS Investigators. N Engl J Med. 2014;370:1683). Prior to meeting entry criteria, which occurred in less than 1 hour, patients in this group received an initial fluid bolus and had a lactate assessment. Prior to randomization, which occurred at around 90 minutes, this group completed 28 mL/kg of crystalloid fluid, and 76% received antibiotics. Thus, the usual-care group in this study nearly achieved the 1-hour bundle currently being contested.

Is it appropriate for a guideline to strongly recommend interventions not backed by level A evidence? The recommendation for FMC to catheterization within 90 minutes has not been studied in a controlled way. The precise dosing and timing of fibrinolysis is also not based on controlled data. Reperfusion devices and antiplatelet agents continue to be rigorously studied, sometimes with conflicting results.

Finally, should the 1-hour bundle be abandoned out of concern that it will be used as a national performance metric? First, there is currently no indication that the 1-hour bundle will be adopted as a performance metric. For the sake of argument, let’s assume the 1-hour bundle will be regulated and used to compare hospitals. Is there reason to think this bundle favors some hospitals over others and will lead to an unfair comparison? Is there significant inequity in the ability to draw blood cultures, send a lactate, start IV fluids, and initiate antibiotics?

Certainly, national compliance with such a metric would be very low at first. Therein lies the actual problem: a person who suffers a STEMI anywhere in the country is very likely to receive high-quality care. Currently, the same cannot be said about a patient with sepsis. Perhaps that should be the focus of our concern.


Dr. Uppal is Assistant Professor, NYU School of Medicine, Bellevue Hospital Center, New York, New York.
 

 

In 2002, the European Society of Intensive Care Medicine, the Society of Critical Care Medicine, and the International Sepsis Forum formed the Surviving Sepsis Campaign (SSC) aiming to reduce sepsis-related mortality by 25% within 5 years, mimicking the progress made in the management of STEMI (http://www.survivingsepsis.org/About-SSC/Pages/History.aspx).

SSC bundles: a historic perspective

The first guidelines were published in 2004. Recognizing that guidelines may not influence bedside practice for many years, the SSC partnered with the Institute for Healthcare Improvement to apply performance improvement methodology to sepsis management, developing the “sepsis change bundles.” In addition to hospital resources for education, screening, and data collection, the 6-hour resuscitation and 24-hour management bundles were created. Subsequent data, collected as part of the initiative, demonstrated an association between bundle compliance and survival.

In 2008, the SSC guidelines were revised, and the National Quality Forum (NQF) adopted sepsis bundle compliance as a quality measure. NQF endorsement is often the first step toward the creation of mandates by the Centers for Medicare and Medicaid Services (CMS), but that did not occur at the time.

In 2012, the SSC guidelines were updated and published with new 3- and 6-hour bundles. That year, Rory Staunton, an otherwise healthy 12-year-old boy, died of septic shock in New York. The public discussion of this case, among other factors, prompted New York state to develop a sepsis care mandate that became state law in 2014. An annual public report details each hospital’s compliance with process measures and risk-adjusted mortality. The correlation between measure compliance and survival also holds true in this data set.

In 2015, CMS developed the SEP-1 measure. While the symbolic importance of a sepsis federal mandate and its potential to improve patient outcomes is recognized, concerns remain about the measure itself. The detailed and specific way data must be collected may disconnect clinical care provided from measured compliance. The time pressure and the “all-or-nothing” approach might incentivize interventions potentially harmful in some patients. No patient-centered outcomes are reported. This measure might be tied to reimbursement in the future.

The original version of SEP-1 was based on the 2012 SSC bundles, which reflected the best evidence available at the time (the 2001 Early Goal-Directed Therapy trial). By 2015, elements of that strategy had been challenged, and the PROCESS, PROMISE, and ARISE trials contested the notion that protcolized resuscitation decreased mortality. Moreover, new definitions of sepsis syndromes (Sepsis-3) were published in 2016 (Singer M, et al. JAMA. 2016;315[8]:801).

The 2016 SSC guidelines adopted the new definitions and recommended that patients with sepsis-induced hypoperfusion immediately receive a 30 mL/kg crystalloid bolus, followed by frequent reassessment. CMS did not adopt the Sepsis-3 definitions, but updates were made to allow the clinicians flexibility to demonstrate reassessment of the patient.

 

Comparing the 1-hour bundle to STEMI care

This year, the SSC published a 1-hour bundle to replace the 3- and 6-hour bundles (Levy MM et al. Crit Care Med. 2018;46[6]:997). Whereas previous bundles set time frames for completion of the elements, the 1-hour bundle focuses on the initiation of these components. The authors revisited the parallel between early management of sepsis and STEMI. The 1-hour bundle includes serum lactate, blood cultures prior to antibiotics, broad-spectrum antibiotics, a 30 mL/kg crystalloid bolus for patients with hypotension or lactate greater than or equal to 4 mmol/L, and vasopressors for persistent hypotension.

Elements of controversy after the publication of this bundle include:


1. One hour seems insufficient for complex clinical decision making and interventions for a syndrome with no specific diagnostic test: sepsis often mimics, or is mimicked by, other conditions.

2. Some bundle elements are not supported by high-quality evidence. No controlled studies exist regarding the appropriate volume of initial fluids or the impact of timing of antibiotics on outcomes.

3. The 1-hour time frame will encourage empiric delivery of fluids and antibiotics to patients who are not septic, potentially leading to harm.

4. While the 1-hour bundle is a quality improvement tool and not for public reporting, former bundles have been adopted as federally regulated measures.


Has the SSC gone too far? Are these concerns enough to abandon the 1-hour bundle? Or are the concerns regarding the 1-hour bundle an example of “perfect is the enemy of better”? To understand the potential for imperfect guidelines to drive tremendous patient-level improvements, one must consider the evolution of STEMI management.

Since the 1970s, the in-hospital mortality for STEMI has decreased from 25% to around 5%. The most significant factor in this achievement was the recognition that early reperfusion improves outcomes and that doing it consistently requires complex coordination. In 2004, a Door-to-Balloon (D2B) time of less than 90 minutes was included as a guideline recommendation (Antman EM, et al. Circulation. 2004;110[5]:588). CMS started collecting performance data on this metric, made that data public, and later tied the performance to hospital reimbursement.

Initially, the 90-minute goal was achieved in only 44% of cases. In 2006, the D2B initiative was launched, providing recommendations for public education, coordination of care, and emergent management of STEMI. Compliance with these recommendations required significant education and changes to STEMI care at multiple levels. Data were collected and submitted to inform the process. Six years later, compliance with the D2B goal had increased from 44% to 91%. The median D2B dropped from 96 to 64 minutes. Based on high compliance, CMS discontinued the use of this metric for reimbursement as the variation between high and low performers was minimal. Put simply, the entire country had gotten better at treating STEMI. The “time-zero” for STEMI was pushed back further, and D2B has been replaced with first-medical-contact (FMC) to device time. The recommendation is to achieve this as quickly as possible, and in less than 90 minutes (O’Gara P, et al. JACC. 2013;61[4]:485).

Consider the complexity of getting a patient from their home to a catheterization lab within 90 minutes, even in ideal circumstances. This short time frame encourages, by design, a low threshold to activate the system. We accept that some patients will receive an unnecessary catheterization or systemic fibrinolysis although the recommendation is based on level B evidence.

Compliance with the STEMI guidelines is more labor-intensive and complex than compliance with the 1-hour sepsis bundle. So, is STEMI a fair comparison to sepsis? Both syndromes are common, potentially deadly, and time-sensitive. Both require early recognition, but neither has a definitive diagnostic test. Instead, diagnosis requires an integration of multiple complex clinical factors. Both are backed by imperfect science that continues to evolve. Over-diagnosis of either will expose the patient to potentially harmful therapies.

The early management of STEMI is a valid comparison to the early management of sepsis. We must consider this comparison as we ponder the 1-hour sepsis bundle.

Is triage time the appropriate time-zero? In either condition, triage time is too early in some cases and too late in others. Unfortunately, there is no better alternative, and STEMI guidelines have evolved to start the clock before triage. Using a point such as “recognition of sepsis” would fail to capture delayed recognition.

Is it possible to diagnose and initiate treatment for sepsis in such a short time frame? Consider the treatment received by the usual care group of the PROCESS trial (The ProCESS Investigators. N Engl J Med. 2014;370:1683). Prior to meeting entry criteria, which occurred in less than 1 hour, patients in this group received an initial fluid bolus and had a lactate assessment. Prior to randomization, which occurred at around 90 minutes, this group completed 28 mL/kg of crystalloid fluid, and 76% received antibiotics. Thus, the usual-care group in this study nearly achieved the 1-hour bundle currently being contested.

Is it appropriate for a guideline to strongly recommend interventions not backed by level A evidence? The recommendation for FMC to catheterization within 90 minutes has not been studied in a controlled way. The precise dosing and timing of fibrinolysis is also not based on controlled data. Reperfusion devices and antiplatelet agents continue to be rigorously studied, sometimes with conflicting results.

Finally, should the 1-hour bundle be abandoned out of concern that it will be used as a national performance metric? First, there is currently no indication that the 1-hour bundle will be adopted as a performance metric. For the sake of argument, let’s assume the 1-hour bundle will be regulated and used to compare hospitals. Is there reason to think this bundle favors some hospitals over others and will lead to an unfair comparison? Is there significant inequity in the ability to draw blood cultures, send a lactate, start IV fluids, and initiate antibiotics?

Certainly, national compliance with such a metric would be very low at first. Therein lies the actual problem: a person who suffers a STEMI anywhere in the country is very likely to receive high-quality care. Currently, the same cannot be said about a patient with sepsis. Perhaps that should be the focus of our concern.


Dr. Uppal is Assistant Professor, NYU School of Medicine, Bellevue Hospital Center, New York, New York.