Move to Take CMS Sepsis Mandates Nationwide Debated

Marcia Frellick

February 25, 2019

SAN DIEGO — The case for the large-scale adoption of the Centers for Medicare and Medicaid (CMS) SEP-1 Early Management Bundle for sepsis is "compelling," according to Laura Evans, MD, director of critical care at Bellevue Hospital Center in New York City.

The need for bundles is not because physicians lack skills or motivation to implement quality-control measures. "These are errors of omission," said Evans, who argued the merits of the bundle in a debate here at the Society of Critical Care Medicine 2019 Congress.

"We have so many competing priorities that we need tools to help us implement the measures," she explained.

In 2015, the CMS mandated that all hospitals document compliance with the core measure. But a recent study showed that between October 2015 and September 2017, only 33% of cases "passed" the bundle requirements (Crit Care Med. 2018;46:1585-1591). The all-or-nothing approach means that every bundle component requirement must be met for a case to pass.

On the other side of the debate was Michael Klompas, MD, associate hospital epidemiologist at Brigham and Women's Hospital in Boston, who is not convinced by the evidence.

The bar for mandates should be extremely high and reserved only for unambiguous directives "that ought to be done for every single patient," he said. Guidelines are different, in that they are recommendations and physicians can weigh the quality of the evidence.

Mandates are a double-edge sword. Although they build awareness about serious public health problems, they can lead to ascertainment bias. With more awareness, more cases are recognized as sepsis, but when the more subtle and more marginal cases are added into the mix, mortality trends can easily be misrepresented, Klompas explained.

Numerous studies have demonstrated the link between adherence to bundles and better outcomes, but wide practice variation in patient care has also been documented.

Pro: Adherence Lowers Mortality

Low rates of influenza vaccination, for example, and of eye exams in patients with diabetes indicate that even routine quality healthcare measures are not being met in the United States, Evans pointed out.

"We don't perform a great deal better in critical care," she added. "There's a lot of variability in what we do from provider to provider at the bedside."

Numerous studies have proven that when compliance with sepsis bundles improves, so do outcomes. One recent study looked at the effect of the nation's first mandated public-reporting initiative for sepsis, which was put in place in New York after the death of a 12-year-old boy, as reported by Medscape Medical News.

As might be expected, compliance with sepsis performance measures improved and mortality rates decreased. The sepsis protocol was initiated in 81.3% of the 91,357 patients in the study cohort. In that group, 3-hour bundle compliance increased from 53.4% to 64.7% (P  <  .001), and risk-adjusted mortality decreased from 28.8% to 24.4% ( < 0.001). In addition, stays were shorter in hospitals with better compliance.

Barriers to implementation include time pressure, perceived threat to physician autonomy, and physicians weighing their own experience more heavily than population-based data, said Evans.

When physicians determine that a patient does not fit the bundle's parameters, they should document the exception for that particular patient and provide the appropriate care, she added.

"But most of our failure to meet these measures is probably based on omission, not an active decision," she said.

CON: The Evidence Isn't Convincing

Klompas argued that the evidence is limited for individual components of the SEP-1 bundle. He questioned whether the measurement of lactate, for example, is necessary for every patient.

"Lactate itself is very nonspecific for infection," he said, pointing out that there is a long list of alternate diagnoses that could cause elevated lactate levels.

Another part of the bundle calls for immediate broad-spectrum antibiotics, but often what seems like sepsis is later found not to be sepsis, he noted.

"The most common cause of sepsis is pneumonia, but one-third of pneumonia is caused by viruses, not bacteria," he said, adding that even severe infections in patients in the intensive care unit can be caused by viruses or fungi.

Klompas also questioned the requirement to give at least 30 cc/kg of crystalloid fluids to patients in septic shock.

Data are emerging that indicate that excess fluids could be associated with higher mortality rates. "We might be hurting our patients," he said, when we do not weigh other information, such ejection fraction and renal function.

One big question is the ability of physicians to determine at the bedside who has sepsis and at what level. The bundle lays out strict timeframes for the administration of antibiotics and fluids, but more flexible timelines might benefit some patients, Klompas asserted.

"If you are uncertain about infection and the patient doesn't have severe disease, then take the time to evaluate your patient further," he advised. "Take the time to see the impact of nonantibiotic administration, such as diuretics and bronchodilators."

Some Common Ground

Although Evans and Klompas had very different points of view on the merits of the mandate, they did find some common ground in its potential, noted panel moderator Phillip Dellinger, MD, from the Cooper Research Institute in Camden, New Jersey.

"They both seem to agree that mandates are potentially good if the metrics are appropriate," he told Medscape Medical News. "Klompas's position was much more that the metrics aren't there yet."

And Evans, even though she was thoroughly supportive of the mandate, "still feels that we have a way to going in refining the metrics," Dellinger said.

Evans, Klompas, and Dellinger have disclosed no relevant financial relationships.

Society of Critical Care Medicine (SCCM) 2019 Congress. Presented February 19, 2019.

Follow Medscape on Twitter @Medscape and Marcia Frellick @mfrellick

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....