Caring for sepsis survivors

Risks of sepsis, including mortality, functional and cognitive impairment, and mental health problems, continue long after discharge.


An 81-year-old man with hypertension and type 2 diabetes mellitus is hospitalized with Klebsiella urinary tract infection, bacteremia, acute renal injury, and lactic acidosis. Fluid and antibiotic therapy normalize his creatinine and lactic acid levels, vital signs, and blood pressure. He does not need physical restraints, a central line, or an indwelling urinary catheter. While preparing for discharge, he asks about his long-term prognosis.

Photo by Thinkstock
Photo by Thinkstock.

The treating hospitalist faces the issue of postsepsis sequelae—a syndrome of late mortality, functional and cognitive impairment, and mental health problems affecting the increasing number of patients who survive severe sepsis or septic shock.

“I see a lot of [sepsis survivors] with cognitive impairment—cloudy thinking or difficulty completing tasks—in addition to weakness or exercise limitations and recurrent sepsis or infection,” said ACP Member Hallie Prescott, MD, MSc, a medical intensivist at the University of Michigan and Veterans Affairs Hospital in Ann Arbor.

A review published in the Jan. 2 JAMA found that hospitalizations with a diagnosis of sepsis approximately tripled in the United States between 2005 and 2014. While survival rates also have gone up, only about half of survivors fully or near fully recover, while a third die within 12 months and about one in six develops severe, chronic functional, or cognitive impairments, the review noted.

In another study of national data, published in the March 10, 2015, JAMA, more than 40% of patients who survived severe sepsis were rehospitalized within 90 days. They were significantly more likely to be readmitted with recurrent sepsis, congestive heart failure, acute renal failure, or respiratory failure compared to similar patients who had not had sepsis. Dr. Prescott was a coauthor on both of these studies.

Additional research has linked sepsis to long-term mortality, increased health care utilization, new and often severe cognitive and functional impairments, lower quality of life, and increased rates of admission to long-term care facilities.

The problem doesn't seem to be just that patients are sicker before they develop sepsis. In a large study of older Medicare enrollees published May 17, 2016, by BMJ, more than one in five sepsis survivors died between a month and two years later as a direct consequence of sepsis.

Late mortality after sepsis “could be more amenable to intervention than previously thought,” Dr. Prescott and coauthors concluded.

“We will never be able to prevent all sequelae of sepsis, but I do believe we can reduce the rates of many of these sequelae,” Dr. Prescott said. She and other experts emphasized the need to diagnose and treat sepsis promptly, identify and prevent in-hospital complications such as delirium, facilitate early ambulation, and counsel patients at discharge, as well as screen them for new functional and cognitive impairments.

The problem of sepsis recovery starts with suboptimal identification and treatment at onset. “We know we give antibiotics too late, and then don't de-escalate them enough,” said Theodore Iwashyna, MD, PhD, a medical intensivist and health services researcher at the University of Michigan and Veterans Health Administration in Ann Arbor. “Sepsis care is hard, it is often low-tech, and it crosses professions and hospital divisions. As a result, it often falls through the cracks.”

CMS now requires all hospitals to report on measures for Severe Sepsis and Septic Shock Management (SEP-1), which went into effect with discharges beginning in October 2015. But the measures require extensive data collection and were recently shown not to measurably improve sepsis survival, according to a review published Feb. 20 by Annals of Internal Medicine (see sidebar).

Postsepsis care in the United States can be even more uneven and is often subpar. In a recent survey of nearly 1,500 sepsis survivors, most expressed only low to moderate satisfaction with postdischarge support services, researchers reported in a conference abstract published in the December 2016 Critical Care Medicine.

One problem is that sepsis guidelines and studies have tended to focus on short-term mortality, not cognitive or functional impairment or worsening comorbidities, Dr. Iwashyna said. “Yet we know half of the 90-day mortality attributable to sepsis happens after hospital discharge. There are really almost no large randomized trials that prove that what we do as inpatients or to help patients recover really works to help them afterwards.”

The handful of randomized trials available has shown mixed results. In a study of nearly 300 sepsis survivors, tracking postdischarge symptoms and encouraging patients to engage in physical activity and other self-care measures did not improve mental health compared with usual primary care. These results appeared in the June 28, 2016, JAMA.

What's more, strategies to prevent mortality and long-term impairment don't necessarily overlap. Among patients with sepsis-associated acute respiratory distress syndrome, conservative fluid therapy led to fewer ventilator and ICU days but was associated with a heightened risk of long-term cognitive problems, possibly because of decreased cerebral perfusion, researchers reported in the June 15, 2012, American Journal of Respiratory and Critical Care Medicine.

“As a practicing doctor, here's what I think: Great inpatient care matters, including prompt recognition [of sepsis], prompt antibiotics, and appropriate resuscitation,” said Dr. Iwashyna. “Preventing organ failure matters. Getting everybody—whether on a vent or not—walking to prevent deconditioning matters.” Ventilator care should emphasize low tidal volumes with minimal sedation, he added. “Make sure people go home on the right medicines, and not on residual antipsychotics from their delirium treatment.”

Exercise is vital, experts said. Although no clinical trials have favored specific postdischarge rehabilitation programs, a study of more than 30,000 sepsis survivors found that patients referred to rehabilitation within 90 days of discharge had a significantly lower risk of dying in the next decade compared with matched controls. These findings were published in the Nov. 1, 2014, American Journal of Respiratory and Critical Care Medicine.

Highly self-motivated patients might be able to follow a structured exercise plan that increases their activity day by day, but patients who are weaker or who are reluctant to exercise because of new pain, dyspnea, or limitations are more likely to benefit from formal rehabilitation—which could be physical therapy, occupational therapy, pulmonary rehabilitation, or cardiac rehabilitation depending on the situation, Dr. Prescott said.

Sepsis survivors often face new weakness, fatigue, cognitive impairment, and income loss. Caregivers also can be inordinately stressed. For these reasons, it's often best to start with just one or two referrals that cover the worst symptoms, said Dr. Prescott.

“I think it's important to understand how sepsis fits into a patient's broader clinical course,” she added. “In a patient with long-standing severe comorbidities, or progressively declining health status and poor quality of life leading into sepsis hospitalization, it may be time to discuss a palliative focus. But for a previously healthy patient who took a big hit during sepsis hospitalization, we should do everything possible to support their recovery.”

Dr. Iwashyna invests considerable energy in discharge counseling. “I think you can prevent problems at home by honestly talking with people about the problems they are likely to have,” he said. “I spend a lot of time talking about what the weeks going home are going to be like.” He often refers patients to the Society of Critical Care Medicine's video, “THRIVE: Discharge from the ICU.” “I am also increasingly trying to get people into peer support groups for ICU and sepsis recovery so they can help each other navigate the pitfalls of our uncoordinated non-system for care after sepsis,” he said.

Researchers continue to seek mechanisms by which sepsis increases the risk of serious morbidity and late mortality. But for now, these remain unknown, said Christopher Sankey, MD, FACP, a hospitalist at Yale School of Medicine in New Haven, Conn.

“Is it related to the site of infection, the kind of bacteria, [or] ICU factors, such as length of stay or number of procedure? It is likely a combination of all these, in addition to host factors, including comorbidities and genetic predispositions to sepsis severity,” he said.

Severity of initial organ dysfunction does not predict later disability or cognitive impairment, Dr. Iwashyna said. “The organ failures that happen to be hard for us to manage in inpatients, such as shock and respiratory failure, are not necessarily the drivers of long-term problems. Instead, keeping infection from progressing to sepsis by prompt recognition and early antibiotics, early resuscitation is likely key for everybody. Everybody is at risk of sepsis sequelae.”