User login
Since its inception, the CMS Sep-1 Core Quality Measure has been unpopular in some circles. It is now under official attack by the American College of Emergency Physicians (ACEP) and the Infectious Diseases Society of America (IDSA), along with a handful of smaller professional societies. These societies appealed the National Quality Forum’s (NQF) 2021 recommendation that the measure be renewed. The NQF is the multidisciplinary and broadly representative group of evaluators who evaluate proposals for CMS-sponsored quality improvement on behalf of the American people and of the Centers for Medicare & Medicaid Services (CMS). Readers of CHEST Physician are likely familiar with core measures, in general, and with Sep-1, in particular. CMS requires hospitals to publicly report their compliance with several Core Quality Measures, and the failure to do so results in across the board reductions in Medicare payments. As of now, no penalties are levied for the degree of compliance but only for failure to report the degree of compliance.
The measure asks, in the main, for hospitals to perform what most physicians can agree should be standard care for patients with sepsis. Depending on whether shock is present, the measure requires:
1. Blood cultures before antibiotics
2. Antibiotics within 3 hours of recognition of sepsis
3. Serum lactate measurement in the first 3 hours and, if increased, a repeat measurement by 6 hours
4. If the patient is hypotensive, 30 mL/kg IV crystalloid within 3 hours, or documentation of why that is not appropriate for the patient
5. If hypotension persists, vasopressors within 6 hours
6. Repeat cardiovascular assessment within 6 hours for patients with shock
If I evaluate these criteria as a patient who has been hospitalized for a serious infection, which I am, they do not seem particularly stringent. In fact, as a patient, I would want my doctors and nurses to act substantially faster than this if I had sepsis or septic shock. If my doctor did not come back in less than 6 hours to check on my shock status, I would be disappointed, to say the least. Nevertheless, some physicians and professional societies see no reason why these should be standards and state that the data underlying them are of low quality. Meanwhile, according to CMS’ own careful evaluation, national compliance with the measures is less than 50%, while being compliant with the measures reduces absolute overall mortality by approximately 4%, from 26.3% to 22.2% (Townsend SR et al. Chest. 2022;161[2]:392-406). This would translate to between 14,000 and 15,000 fewer patients dying from sepsis per year, if all patients received bundled, measure-compliant care. These are patients I don’t care to ignore.
ACEP and IDSA point specifically to the new Surviving Sepsis Campaign Guidelines (SSC) recommendations as evidence that the antibiotic measure is based on low quality evidence (Evans L et al. Crit Care Med. 2021;49[11]:1974-82). In this regard, they are technically correct; the system of evidence review that the SSC panel uses, Grading of Assessment, Recommendations, and Evaluation (GRADE), considers that retrospective analyses, which nearly all of these studies are, can be graded no higher than low quality. Clearly, retrospective studies will never achieve the level of certainty that we achieve with randomized controlled trials, but the NQF, itself, typically views that when a number of well-performed retrospective studies point in the same direction, the level of evidence is at least moderate. After all, just as it would be inappropriate to randomize participants to decades of smoking vs nonsmoking in order prove that smoking causes lung cancer, it is not appropriate to randomize patients with sepsis to receive delayed antibiotics before we accept that such delays are harmful to them.
ACEP and IDSA also assert that the association of early antibiotics with survival is “stronger” for septic shock than for sepsis. In fact, the association is quite strong for both severities of illness. Until it progresses to septic shock, the expected mortality of sepsis is lower, and the percent reduction in mortality is less than for septic shock. However, the opportunity for lives preserved is quite large, because the number of patients with sepsis at presentation is approximately 10 times higher than the number with septic shock at presentation. Antibiotic delays are also associated with progression from infection or sepsis to septic shock (Whiles BB et al. Crit Care Med. 2017;45[4]:623-29; Bisarya R et al. Chest. 2022;161[1]:112-20). Importantly, SSC gave a strong recommendation for all patients with suspected sepsis to receive antibiotics within 3 hours of suspecting sepsis and within 1 hour of suspecting septic shock, a recommendation even stronger than that of Sep-1.
Critics opine that CMS should stop looking at the process measures and focus only on the outcomes of sepsis care. There is a certain attractiveness to this proposition. One could say that it does not matter so much how a hospital achieves lower mortality as long as they do achieve it. However, the question would then become – how low should the mortality rate be? I have a notion that whatever the number, the Sep-1 critics would find it unbearable.
There is a core principle embedded in the Sep-1 process measures, in SSC guidelines, and in the concept of early goal-directed therapy that preceded them: success is not dependent only on what we do but on when we do it. All of you have experienced this. Each of you has attended a professional school, whether medical, nursing, respiratory therapy, etc. None of you showed up unannounced on opening day of the semester and was admitted to that school. All of you garnered the grades, solicited the letters of recommendation, took the entrance exams, and submitted an application. Some of you went to an interview. All of these things were done in a timely fashion; professional schools do not accept incomplete applications or late applications. Doing the right things at the wrong time would have left us all pursuing different careers.
Very early in my career as an attending physician in the ICU, I found myself exasperated by the circumstances of many patients who we received in the ICU with sepsis. I would peruse their medical records and find that they had been septic, ie, had met criteria for severe sepsis, 1 to 2 days before their deterioration to septic shock, yet they had not been diagnosed with sepsis until shock developed. In the ICU, we began resuscitative fluids, ensured appropriate antibiotics, and started vasopressors, but it was often to no avail. The treatments we gave made no difference for many patients, because they were given too late. For me, this was career altering; much of my career since that time has focused on teaching medical personnel how to recognize sepsis, how to give timely and appropriate treatments, and how to keep the data to show when they have done that and when they have not.
Before Sep-1 many, if not most, of the hospitals in the United States had no particular strategy in place to recognize and treat patients with sepsis, even though it was and is the most common cause of death and the costliest condition in American hospitals. Now, most hospitals do have such strategies. Assertions by professional societies that it is difficult to collect the data for Sep-1 reporting are likely true. However, keeping patients safe and alive is a hospital’s primary reason for existing. As long as hospitals are tracking each antibiotic and every liter of fluid so that they can bill for them, my own ears are deaf to hearing that it is too difficult to make sure that we are doing our job. Modifying or eliminating Sep-1 for any reason except data that show we can clearly further improve the outcome for all patients with sepsis is the wrong move to make. So far, other professional societies want to remove elements of Sep-1 without evidence that it would improve our care for patients with sepsis or their outcomes. Thankfully, from the time we proposed the first criteria for diagnosing sepsis, CHEST has promoted what is best for patients, whether it is difficult or not.
Dr. Simpson is a pulmonologist and intensivist with an extensive background in sepsis and in critical care quality improvement, including by serving as a senior adviser to the Solving Sepsis initiative of the Biomedical Advanced Research and Development Authority (BARDA) of the U.S. Department of Health and Human Services and an author of the 2016 and 2020 updates of the Surviving Sepsis Campaign Guidelines. Dr. Simpson is the senior medical adviser for Sepsis Alliance, a nationwide patient information and advocacy organization. He is the immediate past president of CHEST.
Since its inception, the CMS Sep-1 Core Quality Measure has been unpopular in some circles. It is now under official attack by the American College of Emergency Physicians (ACEP) and the Infectious Diseases Society of America (IDSA), along with a handful of smaller professional societies. These societies appealed the National Quality Forum’s (NQF) 2021 recommendation that the measure be renewed. The NQF is the multidisciplinary and broadly representative group of evaluators who evaluate proposals for CMS-sponsored quality improvement on behalf of the American people and of the Centers for Medicare & Medicaid Services (CMS). Readers of CHEST Physician are likely familiar with core measures, in general, and with Sep-1, in particular. CMS requires hospitals to publicly report their compliance with several Core Quality Measures, and the failure to do so results in across the board reductions in Medicare payments. As of now, no penalties are levied for the degree of compliance but only for failure to report the degree of compliance.
The measure asks, in the main, for hospitals to perform what most physicians can agree should be standard care for patients with sepsis. Depending on whether shock is present, the measure requires:
1. Blood cultures before antibiotics
2. Antibiotics within 3 hours of recognition of sepsis
3. Serum lactate measurement in the first 3 hours and, if increased, a repeat measurement by 6 hours
4. If the patient is hypotensive, 30 mL/kg IV crystalloid within 3 hours, or documentation of why that is not appropriate for the patient
5. If hypotension persists, vasopressors within 6 hours
6. Repeat cardiovascular assessment within 6 hours for patients with shock
If I evaluate these criteria as a patient who has been hospitalized for a serious infection, which I am, they do not seem particularly stringent. In fact, as a patient, I would want my doctors and nurses to act substantially faster than this if I had sepsis or septic shock. If my doctor did not come back in less than 6 hours to check on my shock status, I would be disappointed, to say the least. Nevertheless, some physicians and professional societies see no reason why these should be standards and state that the data underlying them are of low quality. Meanwhile, according to CMS’ own careful evaluation, national compliance with the measures is less than 50%, while being compliant with the measures reduces absolute overall mortality by approximately 4%, from 26.3% to 22.2% (Townsend SR et al. Chest. 2022;161[2]:392-406). This would translate to between 14,000 and 15,000 fewer patients dying from sepsis per year, if all patients received bundled, measure-compliant care. These are patients I don’t care to ignore.
ACEP and IDSA point specifically to the new Surviving Sepsis Campaign Guidelines (SSC) recommendations as evidence that the antibiotic measure is based on low quality evidence (Evans L et al. Crit Care Med. 2021;49[11]:1974-82). In this regard, they are technically correct; the system of evidence review that the SSC panel uses, Grading of Assessment, Recommendations, and Evaluation (GRADE), considers that retrospective analyses, which nearly all of these studies are, can be graded no higher than low quality. Clearly, retrospective studies will never achieve the level of certainty that we achieve with randomized controlled trials, but the NQF, itself, typically views that when a number of well-performed retrospective studies point in the same direction, the level of evidence is at least moderate. After all, just as it would be inappropriate to randomize participants to decades of smoking vs nonsmoking in order prove that smoking causes lung cancer, it is not appropriate to randomize patients with sepsis to receive delayed antibiotics before we accept that such delays are harmful to them.
ACEP and IDSA also assert that the association of early antibiotics with survival is “stronger” for septic shock than for sepsis. In fact, the association is quite strong for both severities of illness. Until it progresses to septic shock, the expected mortality of sepsis is lower, and the percent reduction in mortality is less than for septic shock. However, the opportunity for lives preserved is quite large, because the number of patients with sepsis at presentation is approximately 10 times higher than the number with septic shock at presentation. Antibiotic delays are also associated with progression from infection or sepsis to septic shock (Whiles BB et al. Crit Care Med. 2017;45[4]:623-29; Bisarya R et al. Chest. 2022;161[1]:112-20). Importantly, SSC gave a strong recommendation for all patients with suspected sepsis to receive antibiotics within 3 hours of suspecting sepsis and within 1 hour of suspecting septic shock, a recommendation even stronger than that of Sep-1.
Critics opine that CMS should stop looking at the process measures and focus only on the outcomes of sepsis care. There is a certain attractiveness to this proposition. One could say that it does not matter so much how a hospital achieves lower mortality as long as they do achieve it. However, the question would then become – how low should the mortality rate be? I have a notion that whatever the number, the Sep-1 critics would find it unbearable.
There is a core principle embedded in the Sep-1 process measures, in SSC guidelines, and in the concept of early goal-directed therapy that preceded them: success is not dependent only on what we do but on when we do it. All of you have experienced this. Each of you has attended a professional school, whether medical, nursing, respiratory therapy, etc. None of you showed up unannounced on opening day of the semester and was admitted to that school. All of you garnered the grades, solicited the letters of recommendation, took the entrance exams, and submitted an application. Some of you went to an interview. All of these things were done in a timely fashion; professional schools do not accept incomplete applications or late applications. Doing the right things at the wrong time would have left us all pursuing different careers.
Very early in my career as an attending physician in the ICU, I found myself exasperated by the circumstances of many patients who we received in the ICU with sepsis. I would peruse their medical records and find that they had been septic, ie, had met criteria for severe sepsis, 1 to 2 days before their deterioration to septic shock, yet they had not been diagnosed with sepsis until shock developed. In the ICU, we began resuscitative fluids, ensured appropriate antibiotics, and started vasopressors, but it was often to no avail. The treatments we gave made no difference for many patients, because they were given too late. For me, this was career altering; much of my career since that time has focused on teaching medical personnel how to recognize sepsis, how to give timely and appropriate treatments, and how to keep the data to show when they have done that and when they have not.
Before Sep-1 many, if not most, of the hospitals in the United States had no particular strategy in place to recognize and treat patients with sepsis, even though it was and is the most common cause of death and the costliest condition in American hospitals. Now, most hospitals do have such strategies. Assertions by professional societies that it is difficult to collect the data for Sep-1 reporting are likely true. However, keeping patients safe and alive is a hospital’s primary reason for existing. As long as hospitals are tracking each antibiotic and every liter of fluid so that they can bill for them, my own ears are deaf to hearing that it is too difficult to make sure that we are doing our job. Modifying or eliminating Sep-1 for any reason except data that show we can clearly further improve the outcome for all patients with sepsis is the wrong move to make. So far, other professional societies want to remove elements of Sep-1 without evidence that it would improve our care for patients with sepsis or their outcomes. Thankfully, from the time we proposed the first criteria for diagnosing sepsis, CHEST has promoted what is best for patients, whether it is difficult or not.
Dr. Simpson is a pulmonologist and intensivist with an extensive background in sepsis and in critical care quality improvement, including by serving as a senior adviser to the Solving Sepsis initiative of the Biomedical Advanced Research and Development Authority (BARDA) of the U.S. Department of Health and Human Services and an author of the 2016 and 2020 updates of the Surviving Sepsis Campaign Guidelines. Dr. Simpson is the senior medical adviser for Sepsis Alliance, a nationwide patient information and advocacy organization. He is the immediate past president of CHEST.
Since its inception, the CMS Sep-1 Core Quality Measure has been unpopular in some circles. It is now under official attack by the American College of Emergency Physicians (ACEP) and the Infectious Diseases Society of America (IDSA), along with a handful of smaller professional societies. These societies appealed the National Quality Forum’s (NQF) 2021 recommendation that the measure be renewed. The NQF is the multidisciplinary and broadly representative group of evaluators who evaluate proposals for CMS-sponsored quality improvement on behalf of the American people and of the Centers for Medicare & Medicaid Services (CMS). Readers of CHEST Physician are likely familiar with core measures, in general, and with Sep-1, in particular. CMS requires hospitals to publicly report their compliance with several Core Quality Measures, and the failure to do so results in across the board reductions in Medicare payments. As of now, no penalties are levied for the degree of compliance but only for failure to report the degree of compliance.
The measure asks, in the main, for hospitals to perform what most physicians can agree should be standard care for patients with sepsis. Depending on whether shock is present, the measure requires:
1. Blood cultures before antibiotics
2. Antibiotics within 3 hours of recognition of sepsis
3. Serum lactate measurement in the first 3 hours and, if increased, a repeat measurement by 6 hours
4. If the patient is hypotensive, 30 mL/kg IV crystalloid within 3 hours, or documentation of why that is not appropriate for the patient
5. If hypotension persists, vasopressors within 6 hours
6. Repeat cardiovascular assessment within 6 hours for patients with shock
If I evaluate these criteria as a patient who has been hospitalized for a serious infection, which I am, they do not seem particularly stringent. In fact, as a patient, I would want my doctors and nurses to act substantially faster than this if I had sepsis or septic shock. If my doctor did not come back in less than 6 hours to check on my shock status, I would be disappointed, to say the least. Nevertheless, some physicians and professional societies see no reason why these should be standards and state that the data underlying them are of low quality. Meanwhile, according to CMS’ own careful evaluation, national compliance with the measures is less than 50%, while being compliant with the measures reduces absolute overall mortality by approximately 4%, from 26.3% to 22.2% (Townsend SR et al. Chest. 2022;161[2]:392-406). This would translate to between 14,000 and 15,000 fewer patients dying from sepsis per year, if all patients received bundled, measure-compliant care. These are patients I don’t care to ignore.
ACEP and IDSA point specifically to the new Surviving Sepsis Campaign Guidelines (SSC) recommendations as evidence that the antibiotic measure is based on low quality evidence (Evans L et al. Crit Care Med. 2021;49[11]:1974-82). In this regard, they are technically correct; the system of evidence review that the SSC panel uses, Grading of Assessment, Recommendations, and Evaluation (GRADE), considers that retrospective analyses, which nearly all of these studies are, can be graded no higher than low quality. Clearly, retrospective studies will never achieve the level of certainty that we achieve with randomized controlled trials, but the NQF, itself, typically views that when a number of well-performed retrospective studies point in the same direction, the level of evidence is at least moderate. After all, just as it would be inappropriate to randomize participants to decades of smoking vs nonsmoking in order prove that smoking causes lung cancer, it is not appropriate to randomize patients with sepsis to receive delayed antibiotics before we accept that such delays are harmful to them.
ACEP and IDSA also assert that the association of early antibiotics with survival is “stronger” for septic shock than for sepsis. In fact, the association is quite strong for both severities of illness. Until it progresses to septic shock, the expected mortality of sepsis is lower, and the percent reduction in mortality is less than for septic shock. However, the opportunity for lives preserved is quite large, because the number of patients with sepsis at presentation is approximately 10 times higher than the number with septic shock at presentation. Antibiotic delays are also associated with progression from infection or sepsis to septic shock (Whiles BB et al. Crit Care Med. 2017;45[4]:623-29; Bisarya R et al. Chest. 2022;161[1]:112-20). Importantly, SSC gave a strong recommendation for all patients with suspected sepsis to receive antibiotics within 3 hours of suspecting sepsis and within 1 hour of suspecting septic shock, a recommendation even stronger than that of Sep-1.
Critics opine that CMS should stop looking at the process measures and focus only on the outcomes of sepsis care. There is a certain attractiveness to this proposition. One could say that it does not matter so much how a hospital achieves lower mortality as long as they do achieve it. However, the question would then become – how low should the mortality rate be? I have a notion that whatever the number, the Sep-1 critics would find it unbearable.
There is a core principle embedded in the Sep-1 process measures, in SSC guidelines, and in the concept of early goal-directed therapy that preceded them: success is not dependent only on what we do but on when we do it. All of you have experienced this. Each of you has attended a professional school, whether medical, nursing, respiratory therapy, etc. None of you showed up unannounced on opening day of the semester and was admitted to that school. All of you garnered the grades, solicited the letters of recommendation, took the entrance exams, and submitted an application. Some of you went to an interview. All of these things were done in a timely fashion; professional schools do not accept incomplete applications or late applications. Doing the right things at the wrong time would have left us all pursuing different careers.
Very early in my career as an attending physician in the ICU, I found myself exasperated by the circumstances of many patients who we received in the ICU with sepsis. I would peruse their medical records and find that they had been septic, ie, had met criteria for severe sepsis, 1 to 2 days before their deterioration to septic shock, yet they had not been diagnosed with sepsis until shock developed. In the ICU, we began resuscitative fluids, ensured appropriate antibiotics, and started vasopressors, but it was often to no avail. The treatments we gave made no difference for many patients, because they were given too late. For me, this was career altering; much of my career since that time has focused on teaching medical personnel how to recognize sepsis, how to give timely and appropriate treatments, and how to keep the data to show when they have done that and when they have not.
Before Sep-1 many, if not most, of the hospitals in the United States had no particular strategy in place to recognize and treat patients with sepsis, even though it was and is the most common cause of death and the costliest condition in American hospitals. Now, most hospitals do have such strategies. Assertions by professional societies that it is difficult to collect the data for Sep-1 reporting are likely true. However, keeping patients safe and alive is a hospital’s primary reason for existing. As long as hospitals are tracking each antibiotic and every liter of fluid so that they can bill for them, my own ears are deaf to hearing that it is too difficult to make sure that we are doing our job. Modifying or eliminating Sep-1 for any reason except data that show we can clearly further improve the outcome for all patients with sepsis is the wrong move to make. So far, other professional societies want to remove elements of Sep-1 without evidence that it would improve our care for patients with sepsis or their outcomes. Thankfully, from the time we proposed the first criteria for diagnosing sepsis, CHEST has promoted what is best for patients, whether it is difficult or not.
Dr. Simpson is a pulmonologist and intensivist with an extensive background in sepsis and in critical care quality improvement, including by serving as a senior adviser to the Solving Sepsis initiative of the Biomedical Advanced Research and Development Authority (BARDA) of the U.S. Department of Health and Human Services and an author of the 2016 and 2020 updates of the Surviving Sepsis Campaign Guidelines. Dr. Simpson is the senior medical adviser for Sepsis Alliance, a nationwide patient information and advocacy organization. He is the immediate past president of CHEST.