Compliance with the Centers for Medicare and Medicaid Services’ (CMS) SEP-1 quality measure does not improve sepsis mortality rates, according to results from a retrospective study.

“Our findings of similar adjusted mortality in cases that failed versus passed SEP-1 should not be interpreted as timely sepsis recognition and care being unimportant,” said Dr. Chanu Rhee from Brigham and Women’s Hospital, Harvard Medical School, in Boston.
“Rather, it likely reflects the overly rigid nature of the measure, since SEP-1 is an ‘all-or-nothing’ measure that requires perfect compliance to get any credit, and most failures occur on elements that are less clearly important to patient outcomes,” he told Reuters Health by email.

The severe sepsis bundle of SEP-1 requires lactate measurements, blood cultures and broad-spectrum antibiotics within three hours of sepsis onset, with repeat lactate measurements within six hours if the initial lactate level is elevated. The septic shock bundle adds three requirements: 30 mL/kg of IV fluids within three hours, vasopressors within six hours for persistent hypertension and a repeat volume assessment within six hours.

Dr. Rhee and colleagues in the Centers for Disease Control and Prevention (CDC) Prevention Epicenters Program evaluated the association between SEP-1 compliance and patient outcomes, taking into account patients’ clinical characteristics, for sepsis cases reported by seven academic and community hospitals to CMS during the first two years after SEP-1 implementation.

Of the 851 sepsis patients available for analysis, 33 percent passed and 67 percent failed SEP-1. The most common cause of failure was failure to draw an initial lactate or repeat lactate within six hours (40 percent), while 15.1 percent failed because of delayed administration of broad-spectrum antibiotics.

Unadjusted in-hospital mortality rates were higher for SEP-1 compliance failures (18.4 percent) than for SEP-1 compliance passes (11.0 percent), but the difference was no longer significant after adjusting for patients’ clinical characteristics (adjusted odds ratio, 1.36; P=0.205).

Time to antibiotics administration exceeding three hours was associated with 94 percent increased odds of death after adjustment, whereas failing SEP-1 for any other reason was not significantly associated with in-hospital mortality, the team reports in Critical Care Medicine.

Both SEP-1 failure and delayed antibiotic administration were associated with higher mortality in patients with infectious signs but not in those with vague presenting complaints.

“CMS should consider simplifying the SEP-1 measure to focus on the care elements that are most critical to patient outcomes,” Dr. Rhee said. “In particular, our study, and numerous others in the literature, strongly suggests that administration of appropriate antibiotics is the key time-dependent intervention in sepsis.”

“The other elements of the basic 3-hour bundle—drawing initial lactates and blood cultures before antibiotics—certainly seem like good practice as well,” he said. “However, most of the other bundle elements, such as 30 mL/kg fluid requirement for hypotension or lactate levels ≥4, repeat lactates for patients with initial lactate levels >2, and documentation of volume reassessment exams, seem overly prescriptive and are not well supported by evidence.”

Dr. Rhee added, “The other potential modification could be to change from an ‘all-or-nothing’ measure to allow partial credit for completion of certain aspects of the bundle, even if 100 percent of the bundle elements are not met.”

Dr. Arjun Venkatesh from Yale University School of Medicine, in New Haven, Connecticut, who recently evaluated emergency department performance on SEP-1, told Reuters Health by email, “Variation in the SEP-1 metric is broad and average performance is likely lower than many EDs or hospitals may observe in dashboard or local quality reports. This may be explained by nuances of this quality measure, and clinical leaders should use this data to reassure hospital leadership of boards that may perceive sepsis care to be inadequate based on this isolated snapshot of sepsis care.”

SEP-1 “is a process measure, and as such it will always drive improvements in documentation as much as it will drive improvements in actual care delivery,” said Dr. Venkatesh, who was not involved in the new work. “As sepsis care becomes increasingly standardized, the development of a risk-adjusted outcome measure (such as in-hospital mortality) may be a more worthy and powerful driver of practice change and improved outcomes.”

Dr. Ian J. Barbash of the University of Pittsburgh School of Medicine, in Pennsylvania, recently evaluated perceptions of hospital quality officials regarding SEP-1. He told Reuters Health by email, “The evidence from this study and others published since the beginning of SEP-1 suggest that a modified SEP-1 measure should retain a primary focus on timely sepsis recognition and antibiotic administration, but perhaps give hospitals more flexibility on other aspects of the bundle with less direct evidence that they improve patient outcomes.”

“In many ways, the message remains the same as it has for some time: physicians must be vigilant in identifying sepsis early and ensuring rapid administration of appropriate antibiotic therapy,” said Dr. Barbash, who also was not involved in the research. “This study also suggests that the problem of recognition and early treatment may be particularly vexing for patients with vague presenting symptoms or whose sepsis begins in the hospital.”