Sepsis Patients Get Lost After Hospital Discharge, LDI Fellows Say

A patient survives sepsis, leaves the hospital, and the crisis is treated as over. But for many survivors, the danger still exists. A substantial number will return to the hospital, and many will suffer long-term cognitive, physical, or mental health problems.

A new implementation science study by LDI Senior Fellow Kathryn Bowles finds that getting these patients timely home health and outpatient care after discharge takes much more coordinated effort than health systems typically provide. Bowles initiated the I-TRANSFER study to address these gaps and document what successful post-acute care delivery looks like in that critical first week after discharge.

“Care transitions are a vulnerable time for older adults, “Bowles noted. “Timely response to problems can prevent readmissions” and other problems.

The study, published July 17 in BMC Health Services Research, followed five health systems and their partner home health agencies at 16 hospitals across four states. Implementation starts with a home health nursing visit within two days of hospital discharge, a second visit within the first week, and an outpatient follow-up appointment within seven days. Over nearly two years, the sites deployed more than 450 implementation strategies to make it happen.

Only about 28% of sepsis survivors nationwide currently receive this kind of timely post-acute follow-up, despite evidence that it lowers readmissions. Thirty-day readmission rates for sepsis survivors run as high as 25%, according to a 2020 systematic review and meta-analysis, and nearly one-third of sepsis readmissions from home health care occur within the first week home, according to a prior study conducted by Bowles.

“I was an ICU nurse and cared for many sepsis patients, knowing how serious the condition is and what they went through,” said Bowles, professor and van Ameringen Chair in Nursing Excellence at Penn Nursing. “I knew recovery must be tough.”

“Having an RN visit the patient’s home within two days of discharge assures that we have professional eyes on the patient to assess their status, educate them about their condition, and communicate with the providers for needed changes to the plan of care,” she said.

Sepsis is More a Communication Problem Than a Knowledge Problem

The study catalogued strategies using Expert Recommendations for Implementing Change (ERIC), a standard library of 73 named implementation tactics. Bowles’ team and the participating sites together used 45 of them, ranging from EHR sepsis alerts and discharge dashboards to phone calls talking reluctant patients into accepting a home health nurse.

What surprised her most was how little the acute and post-acute sides of care knew about each other’s workflows, Bowles said. A hospital marking sepsis as “resolved” in a chart before discharge, for example, could unintentionally erase it from the record that a home health nurse reviews days later. “It is important for us to work together. By us I mean acute and post-acute care,” she said. “There were so many issues each side was not aware of.”

That disconnect did not surprise Lynn Cardona, Penn Medicine Sepsis Alliance Program Coordinator for Quality in the Office of the CMO, who helps coordinate sepsis quality efforts across Penn’s seven hospitals and who consulted with Bowles’ team.

The field still lacks a shared, rapid way to identify sepsis in the first place. National and institutional definitions of sepsis have evolved. Like many health care institutions, Penn has implemented electronic health record and clinical decision support tools to assist clinicians in this regard. However, widespread adoption and utilization rely on how quickly and reliably these tools can identify sepsis. So, Penn’s working rule is pragmatic: “If it quacks like a duck, it is sepsis. I’m going down that path,” Cardona explained.

The urgency that clinicians feel early in a hospitalization tends to disappear by the time a patient is ready to leave, according to Cardona. “Once they get ready to go home, it’s less about that septic episode and more on, now let’s get you back to your life.”

That’s when the handoff to home health can fail. “How do we message that? How do we consistently communicate that in a way that provides the information, doesn’t create fear, but also creates the urgency?” asked Cardona. “That’s really, really hard to do, let alone [with] 40,000 clinicians.”

What Sepsis Care Looks Like on the Front Line

At New York City-based VNS Health, one of the study’s home health partners, that disconnect is all too common. Fortune Grunin, a clinical account manager who worked with NYU Langone’s Manhattan and Brooklyn hospitals on the study, said identifying sepsis patients in home care required significant effort, as agencies rely on hospital diagnosis and referral information. “Once sepsis is resolved, it may no longer flag the patient and appear among the primary diagnosis.”

Getting patient and family cooperation for the I-TRANSFER protocol was its own hurdle. Families are often overwhelmed by the precipitating health crisis, and don’t want a stranger in their home, especially within the first 24 to 48 hours. Or they are unable to follow up within the seven-day appointment window. Once that initial visit is pushed back, then the whole protocol falls apart, Grunin explained. If a start-of-care visit cannot be completed within 48 hours, the protocol criteria cannot be fully met.

Ensuring the patient or family member scheduled a physician follow-up appointment was also challenging. Patients would forget, or just not follow through. And even if they called, there was often no clinical availability.

“Timely physician follow-up after hospital discharge can be difficult for hospitals and the home care agencies to ensure and monitor.”

Still, the shared push changed how NYU and VNS worked together. The partnership has since expanded to a heart failure protocol modeled on the same I-TRANSFER approach. “It brought us more together because we were working towards a goal together,” she said.

A New Billing Code for Sepsis Survivors, and What’s Next

One concrete outcome of the study was the research team’s successful petition of the CDC for a dedicated ICD-10 code, Z51.A, “Encounter for Sepsis Aftercare.” It gives clinicians a standardized way to flag sepsis survivorship in the medical record after discharge, something that didn’t previously exist.

Cardona called it meaningful, but pointed out the code helps only if a patient is identified as a sepsis survivor before the home health referral is even sent. “If we do not identify that patient prior to the home care referral, then they lose all of their power in using that code,” she said.

Monitoring post-discharge sepsis survivors isn’t a one-size-fits-all solution. Each of the five sites chose its own path to the same eight objectives, or what Bowles called function versus form, because hospital size, technology, and staffing varied so widely.

This study documented how health systems implemented the I-TRANSFER protocol. A forthcoming publication will report its effect on readmissions.

The nonprofit Research Institute for Home Care calls the I-TRANSFER protocol “an exciting update” that might produce better outcomes for sepsis patients. In an emailed statement, the institute said it “applauds research efforts to improve clinical care in the home and formalize the information flow between home care providers and other clinics.”

Bowles is already looking beyond sepsis. Her team is considering a follow-up study applying the same model to discharged intensive care patients, and hopes wider adoption of the new aftercare code will make sepsis survivorship more visible in medical records.