According to new research, changes in procalcitonin (PCT) serum levels can guide duration of intravenous antibiotic use in patients who are critically ill with suspected sepsis more effectively than standard care or changes in C-reactive protein (CRP) levels. The ADAPT-Sepsis trial found that PCT guidance reduced overall antibiotic use by about 10% vs usual care, but CRP guidance did not. There was no significant mortality difference between PCT-guided care and standard care, while all-cause CRP mortality was inconclusive.

“The UK government, through its National Institute for Health and Care Research, solicited a biomarker trial in patients who were critically ill with suspected sepsis because data for and against biomarker use to guide antibiotic treatment in these patients were inconclusive,” said Paul Dark, MD, PhD, Professor of Critical Care Medicine and Vice Dean for Health and Care Partnerships at the University of Manchester, Faculty of Biology, Medicine, and Health, Manchester, UK. “Both PCT and CRP can be used, but the evidence was weak.”

The ADAPT-Sepsis trial randomized 2,760 adult patients across 41 ICUs in the United Kingdom to daily PCT guidance for antibiotic use (918 patients), CRP-guided antibiotic use (924 patients), or usual care (918 patients). All patients required intravenous antibiotics for suspected sepsis, defined as “acute organ dysfunction associated with suspected infection.” Although it is worth noting that patients who are severely immunocompromised were not included in this trial, so results are not applicable to them.

Patients had a mean age of 60.2 years, and 60.4% were male. Nearly all met Sepsis-3 criteria for the diagnosis of sepsis (average Sequential Organ Failure Assessment scores of 7 in the study), and the study groups were evenly divided between those with sepsis and septic shock.

Clinical teams received daily advice on stopping or continuing antibiotic treatment based on changes in PCT, CRP, or usual care but were blinded to the treatment group and source of the recommendations.

The primary outcomes were total antibiotic duration and all-cause mortality from randomization to 28 days. Secondary outcomes included antibiotic duration and dose for the initial sepsis duration, total antibiotic dose, unscheduled escalation or readmission, infection relapse or recurrence, suspected antibiotic adverse reactions, time to discharge, length of stay for critical care and total inpatient care, and all-cause mortality at 90 days.

There was a significant reduction in total duration of antibiotic treatment for PCT-guided therapy, 9.8 days vs 10.7 days for standard care for a difference of 0.88 days (95% CI, 0.19-1.58; P = .01). There was no difference between standard care and CRP-guided care at 10.6 days, a difference of 0.09 days (95% CI, -0.60 to 0.79; P = .79). Results were similar by site of infection, surgical vs medical admission, septic shock vs no shock, and other subgroups.

All-cause mortality at 28 days for PCT-guided therapy was noninferior to standard care, 20.9% vs 19.4%, respectively, for an absolute difference of 1.57 days (95% CI, -2.18 to 5.32; P = .02). Noninferiority for CRP-guided therapy was inconclusive.

There was also a biomarker-driven reduction in the duration of antibiotics for the initial sepsis period, a difference of 1.13 days for PCT and 0.71 days for CRP vs usual care. There were no differences in other secondary outcomes between the three groups.

The current weak recommendation for the use of PCT-guided antibiotic discontinuation in sepsis care could be revised later this year, Dr. Dark said. The National Institute of Health and Care Excellence will explore the clinical and cost effectiveness of PCT-guided antibiotic therapy and may issue its own recommendations.

“A 10% reduction in duration of antibiotic use may seem small, but it could provide significant cost and labor savings,” Dr. Dark said. “In the context of antibiotic stewardship, a 10% reduction in use is important. We have to guard the antibiotics we have.”