Sepsis drives far more readmissions than we realized

Sepsis may contribute to far more hospital readmissions than previously recognized -- more than any other monitored condition. Recognition of this by federal and private payers could result in increased intensity of oversight of health teams' care delivery for patients with sepsis. Hospital readmissions are squarely in the sights of the Centers for Medicare and Medicaid Services (CMS), the government behemoth that pays for the lion's share of health care delivered in the U.S. Penalties assessed on hospitals for readmissions will increase by ~25% to $528 million in 2017. CMS has monitored its "big four" conditions -- heart failure, myocardial infarction, COPD, or pneumonia, through its Hospital Readmissions Reduction Program (HRRP), created by the Affordable Care Act. Readmission for monitored conditions can generate financial penalties. But in a separate national quality monitoring program reviewing 14 million hospitalizations, sepsis accounted for more unplanned hospital readmissions than any of the (not-so-big-anymore) four. Sepsis was involved in 12.2% of 1 million hospitalizations with unplanned readmissions; heart failure was a distant second, accounting for a puny 6.7% of readmissions. Sepsis-associated readmissions were also the longest and most expensive. Results were reported to the Society of Critical Care Medicine meeting and published in the Journal of the American Medical Association. Investigators used the 2013 Nationwide Readmissions Database, covering about half the hospitalizations in the U.S.  They found the following proportions of conditions associated with unplanned readmissions:

• Sepsis (12.2%), mean length of stay 7.4 days

• Heart failure (6.7%), mean LOS 6.5 days

• Pneumonia (5.0%), mean LOS 6.9 days

• COPD (4.8%), mean LOS 6.3 days

• AMI (1.3%), mean LOS 6.0 days

The cost of sepsis-associated unplanned readmission was found to be significantly higher than any of the "big four" CMS conditions. Co-author Sachin Yende, MD of the University of Pittsburgh told MedPage Today,

For physicians and patients, I think the most important implication of this study is that sepsis is an acute illness with long-term consequences ... Most people think that once you get better from sepsis and are discharged from the hospital, you don't have to worry about any consequences. What this study shows is that many of the patients are likely to come back into the hospital within 30 days.

Could Sepsis Be Added to Medicare's "Black List" for Readmissions?

It's obvious to anyone on the front lines of health care delivery that not all readmissions for sepsis (or any other condition) are preventable. CMS's HRRP penalties are roundly criticized as being unfair by those on the losing end of the financial deal -- mainly hospitals. That being said, studies do suggest the HRRP has reduced readmission rates for some monitored conditions. CMS has already dipped its big fat toe into the waters of care delivery for sepsis, with its confusing and clunky sepsis core measure. With the annual costs of care delivery for sepsis at $3.4 billion in the U.S., it wouldn't be surprising to see CMS add sepsis to the HRRP list. That would be a giant cannonball dive into the middle of hospitals' ICU operations. An enormous proportion of ICU patients now trigger "sepsis alerts" based on organ dysfunction and the imprecision of sepsis case identification. Attaching a potential downstream financial loss to each of these patients is likely to bring more scrutiny, analytics, etc. to the care delivery in the ICU. The extra attention might help critically ill patients, whether or not it improves sepsis care per se. It could also be a half-baked, paper-pushing fad (see early goal-directed therapy "sepsis bundles", c. 2002), resulting in excessive and unhelpful testing and treatments, reducing complex medical decision making into administrator-pleasing metrics with a dubious relationship to reality. Trust that it will be both of these things. The balance of benefit from sepsis performance measures will depend on what proportion of patients bundled into the metric actually have sepsis (a metric we can't know, today). Read more: Mayr F, et al "Understanding the burden of unplanned sepsis readmissions" SCCR 2017; Abstract 1336. Mayr F, et al "Proportion and cost of unplanned 30-day readmissions after sepsis compared with other medical conditions" JAMA 2017; doi: 10.1001/jama.2016.20468.