James Ford, MD, MAS, discusses the evidence supporting SEP-1’s impact on sepsis mortality and suggests alternative approaches focused on outcome-based measures in sepsis management.
A recent systematic review found no strong evidence that compliance with or implementation of the Centers for Medicare & Medicaid Services (CMS) Severe Sepsis and Septic Shock Management Bundle (SEP-1) improves sepsis mortality. This raises concerns about the inclusion of SEP-1 in the Hospital Value-Based Purchasing (VBP) Program.
The review analyzed 17 observational studies examining the relationship between SEP-1 compliance or implementation and sepsis mortality. Five studies showed a statistically significant benefit from SEP-1 compliance, while seven did not. One study found a benefit from SEP-1 implementation, but it did not adjust for pre-existing trends in mortality before SEP-1’s introduction.
In an email interview with investigator James Ford, MD, MAS, assistant professor of emergency medicine at the University of California, San Diego Medical Center, he said, "Including the entire SEP-1 metric into the Hospital VBP is not supported by the available scientific evidence." He proposed one option: "One alternative would be to include only the antibiotic mandate in patients with septic shock, which has been shown consistently to be associated with mortality." He also suggested another approach: "Suspend inclusion of SEP-1 in the VBP altogether, and instead develop a measure that focuses on outcome measures (ie, mortality), rather than process measures (ie, bundle compliance)."
Ford mentioned that CMS is working on a new measure, "CMS is working on one such measure, the electronic clinical quality measure (e-CQM) for community-acquired sepsis, which tracks 30-day mortality."
The review also pointed out limitations, including the observational nature of all the included studies and a failure to control for confounders. Given these issues, the authors concluded there is insufficient evidence to support SEP-1’s impact on mortality rates.
Ford further addressed the challenges of relying on observational studies, stating, "Even the best-adjusted observational studies are still susceptible to bias from residual confounding. Historically, there have been many instances when observational studies found benefit with an intervention, only to be disproven by subsequent randomized controlled trials (RCTs)." He gave examples like hormone replacement therapy for cardiovascular risk in post-menopausal women and the use of drugs to prevent abnormal heart rhythms in patients who had heart attacks. "In both of these examples, observational studies showed benefit, but subsequent RCTs found no benefit or harm," Ford noted.
While no RCTs have assessed the efficacy of a 3-hour sepsis bundle, Ford referenced a recent RCT (Freund et al.) that found no difference in mortality between a 1-hour sepsis bundle and usual care.
The systematic review found no strong evidence that SEP-1 improves sepsis mortality, with mixed results across 17 studies.
Observational studies used in the review are prone to biases, raising concerns about their reliability in assessing SEP-1’s effectiveness.
Experts recommend reconsidering SEP-1’s inclusion in the Hospital VBP Program and focusing on outcome-based measures, with further research needed to refine sepsis management approaches.
The review also stresses the need for higher-quality studies to better evaluate the impact of SEP-1 on sepsis mortality. Given the current state of evidence, Ford noted, "Since SEP-1 is already standard of care in the United States, performing an RCT that randomizes patients to receive bundle components <3 hours or ≥ 3 hours would be unethical. This makes it challenging to truly study." He acknowledged that some studies have used propensity-matching to simulate randomized controlled trials (RCTs), but noted the limitations of these approaches, such as focusing only on Medicare beneficiaries or academic medical centers in California.
Ford suggested, "the next best approach may be to use propensity-matching from a large, nationally representative sample, with adequate adjustment and rebalancing for various important confounders." He also highlighted that future directions may move away from rigid treatment bundles and focus on "providing personalized care tailored to each patient’s specific needs." His team is currently exploring this with "machine learning and artificial intelligence to help predict how patients will respond to changes in treatment."
The findings of this systematic review suggest that CMS should reconsider the inclusion of SEP-1 in the Hospital VBP Program until further research provides stronger evidence. This article underscores the need for continued research and a more evidence-based approach to sepsis management.
