Doctors aren't complying with the CMS sepsis quality measure

When the Centers for Medicare and Medicaid Services (CMS) released its 2015 performance measure for the treatment of sepsis -- called SEP-1 or the Severe Sepsis/Septic Shock Early Management Bundle, physicians responded with general befuddlement: the measure demanded they follow such unusual practices as giving 3-liter boluses of saline to anuric, hypertensive, hypoxemic patients with end-stage renal disease, based solely on the presence of lactatemia and a suspicion of sepsis.

As noncompliance could be met with financial penalties from the federal government or other unpleasantness, hospital administrators responded with a flurry of emails to their ICUs and emergency departments; new protocols and electronic medical record order sets for sepsis soon followed.

A new study serves as a sort of progress report, suggesting hospitals and emergency medicine physicians are getting into line with the new sepsis regulations, but not quickly or with a lot of gusto.

The American College of Emergency Physicians has its own quality network to monitor practice patterns and outcomes throughout the U.S. Their investigators reviewed self-reports of adherence with the SEP-1 bundle from 50 emergency departments at hospitals enrolled in the program.

Average hospital bundle adherence rate was 54%, but varied widely (between 10% and 100%). Compliance increased throughout 2016, from 39% to 57%. The largest variations were in intravenous fluid resuscitation and in rechecking lactate levels. The highest rate of compliance (~90%) was in initially checking lactate.

Although dubbed a "quality" measure, SEP-1 is not grounded in much evidence, and everyone subject to it should know something about its genesis.

The measure was introduced to the National Quality Foundation's decision-making process by the team that brought us early-goal-directed therapy (EGDT), and is largely based on those principles. A previous version of SEP-1 the NQF endorsed in ~2012 even contained a requirement that physicians use a catheter to continuously monitor central venous oxygenation and central venous pressure. This was a boon to catheter maker Edwards Lifesciences, which had also disclosed paying hundreds of thousands of dollars to the original EGDT proponent and his hospital, who together proposed the performance measure.

Along the way (in trials published 2014-2016), EGDT was debunked as not improving outcomes in sepsis, while adding to health care costs, ICU stays and subjecting patients to unnecessary procedures and risks of complications.  But by then the federal bureaucratic ship had sailed, and SEP-1 (minus the catheter requirement) became the law of the health care land in 2015.

A subsequent randomized trial in Africa (in very different patients and in centers with vastly fewer resources than the U.S.) suggested that fluid resuscitation of septic patients with volumes roughly comparable to those mandated by SEP-1 caused respiratory failure and death (largely because mechanical ventilation was not available).

In fact, no sepsis bundle elements have been well-tested in randomized trials. A large observational study in New York suggested quickly giving antibiotics as part of a legally-mandated sepsis bundle saved lives in New York. Any benefit from the other bundle elements could not be ascertained in that study.

There's no way to know the reasons for hospitals' and physicians' slow pace of adoption of the SEP-1 CMS sepsis performance measure to date. Ordinary institutional delays, resistance by health care teams, or both could be present. Considering that SEP-1 originated from a small group with established conflicts of interest, based on deficient evidence and a disproven study, and mandates treatments that frequently contradict good clinical judgment, public criticism (or even skepticism) by physicians has been limited to a surprisingly few well-informed voices.