More RN Hours Linked With Better Sepsis Outcomes

Hospitals providing more hours of nursing care may see reduced mortality from sepsis, according to a cross-sectional study.

In the analysis of more than 700,000 Medicare patients with sepsis in 2018, each increase in registered nurse (RN) hours per patient day was associated with a 3% decrease in the likelihood of mortality at 60 days (OR 0.97, 95% CI 0.96-0.99), according to a multivariable analysis that controlled for patient and hospital characteristics, including Severe Sepsis and Septic Shock Management Bundle (SEP-1) score.

"The study findings suggest that nurse workload is an overlooked and underused aspect of the treatment bundle for patients with a diagnosis of sepsis," wrote researchers led by Jeannie Cimiotti, PhD, RN, of the Nell Hodgson Woodruff School of Nursing at Emory University in Atlanta, in JAMA Health Forum.

More than 180,000 patients died of sepsis during the study period, but the researchers estimated that 1,266 fewer deaths would have occurred if all hospitals had added six more RN hours per patient day. And nine more RN hours per patient day would have resulted in an estimated 6,360 fewer deaths.

"The sepsis patient requires complex care, and if you don't have enough nurses on staff to address those patient care needs, unfortunately, your outcomes are going to be worse," Cimiotti told MedPage Today. "If they're short-staffed or overworked ... you're going to lose that window of opportunity."

The researchers also found that hospitals with an on-staff intensivist saw a 16% decrease in sepsis mortality at 60 days (OR 0.84, 95% CI 0.79-0.89). Yet more hospitalist and physician hours were linked with higher sepsis mortality, a finding that "could be a result of interprofessional communication ... and the likelihood that nurses might report findings to one group of physicians (e.g., intensivists) over another (e.g., hospitalist or physician)," the authors noted.

Additionally, each 10% increase in a hospital's SEP-1 score was associated with a 2% lower likelihood of sepsis mortality at 60 days (OR 0.98, 95% CI 0.97-0.99).

SEP-1 is an all-or-nothing core measure that was adopted by Centers for Medicare & Medicaid Services (CMS) in October 2015 and is endorsed by the National Quality Forum. It gauges a hospital's ability to provide "timely and effective sepsis care," Cimiotti group explained, but the protocol has been criticized as overly complex and subjective.

The care bundle has seven components, including steps such as taking serum lactate levels and blood cultures and delivering broad-spectrum antimicrobial therapy within 3 hours of sepsis onset, among others.

Past research has suggested that nursing workload stands in the way of "rapid initiation of antibiotic treatment" for patients with sepsis, noted Cimiotti and co-authors, which suggests that "an adequate number of nurses are necessary to improve SEP-1 bundle compliance." Communication between nurses and physicians has also been identified as a necessary component for sepsis care.

While the study was conducted pre-pandemic, Cimiotti stressed that the findings are even more relevant now with the "exodus of nurses from acute care."

"At the end of the day, acute care hospitals nationwide have to make it a priority that their units are adequately staffed with adequately trained RNs, and they have to be compensated in a way ... that will not only attract nurses but retain them," Cimiotti emphasized. "Otherwise every patient is at risk -- even those without sepsis."

For the current study, the researchers used 2018 data from the American Hospital Association (AHA) Annual Survey, CMS Hospital Compare, and Medicare claims data. It included a total of 702,140 Medicare beneficiaries with a primary diagnosis of sepsis on admission (182,346 of whom died) from 1,958 nonfederal general acute care hospitals that submitted SEP-1 data. Patients were 78 years on average, and 51% were women.

Mean SEP-1 score was 56.1, the average RN hours per patient day was 6.2, and two-thirds of the hospitals had an intensivist on staff. Just under a third of the hospitals had 100 or fewer beds, while 34% had 101-250 beds, 23% had 251-500 beds, and the rest had over 500. Roughly three-fourths were non-profit hospitals and in metropolitan areas, and 72% were non-teaching.

Study limitations included the cross-sectional design, which cannot prove causality between RN hours and sepsis-related death, although prospective studies have shown associations between nurse staffing and patient outcomes, including mortality, the study authors noted.

Comment