Procalcitonin-Based Algorithm Does Not Reduce Antibiotic Use in the ICU

Clinical question

Can a procalcitonin-based algorithm reduce antibiotic use in critically ill patients?

Bottom line

A procalcitonin-based algorithm using a 0.1 ng/mL cutoff does not significantly decrease the duration of antibiotic treatment in critically ill patients nor does it reduce length of stay or number of deaths. The rate of decline in the procalcitonin level over the first 72 hours, however, does serve as an independent predictor of short-term and long-term all-cause mortality. (LOE = 1b-)

Reference

Shehabi Y, Sterba M, Garrett PM, et al, for the ProGUARD Study Investigators and the ANZICS Clinical Trials Group. Procalcitonin algorithm in critically ill adults with undifferentiated infection or suspected sepsis. Am J Respir Crit Care Med 2014;190(10):1102-1110.

Study design: Randomized controlled trial (nonblinded)

Funding source: Foundation

Allocation: Concealed

Setting: Inpatient (ICU only)

Synopsis

Procalcitonin (PCT) is a sepsis biomarker that has been utilized to guide antibiotic use in different patient populations. In this study, the authors tested a PCT-algorithm using a 0.1 ng/mL cut-off to reduce antibiotic exposure in critically ill patients. Patients newly admitted to intensive care units (ICUs) who were receiving antibiotics for suspected infections were randomized, using concealed allocation, to receive PCT-guided care (n = 196) or standard care (n = 198). All patients had PCT levels drawn daily until discharge from the ICU or up to a maximum of 7 days. In the PCT group, antibiotics were stopped if PCT levels were negative (< 0.1 ng/mL), if PCT levels were borderline (0.1 - 0.25 ng/mL) and infection was unlikely, or if PCT levels decreased more than 90% from baseline values. In the standard care group, the treating clinician determined antibiotic use without knowledge of the PCT results. Baseline characteristics of the 2 groups were similar with regard to severity-of-disease scores and baseline PCT values. There was high compliance with the PCT algorithm, with less than 3% of study days when the algorithm was not followed. There was no significant difference detected between the 2 groups for the primary outcome of time to antibiotic cessation. However, duration of antibiotic use was longer than expected in the control group (11 days actual vs 9 days expected), so the study may have been underpowered to detect an expected 25% reduction. Nevertheless, the 2 groups were similar with regard to ICU and hospital lengths of stay, as well as ICU, hospital, and 90-day mortality rates. Of note, the rate of decline in PCT level over the first 72 hours was an independent predictor of hospital mortality and 90-day mortality, with a slower decline corresponding to a higher mortality.

Dr. Kulkarni is an assistant professor of hospital medicine at Northwestern University in Chicago.

Clinical question

Can a procalcitonin-based algorithm reduce antibiotic use in critically ill patients?

Bottom line

A procalcitonin-based algorithm using a 0.1 ng/mL cutoff does not significantly decrease the duration of antibiotic treatment in critically ill patients nor does it reduce length of stay or number of deaths. The rate of decline in the procalcitonin level over the first 72 hours, however, does serve as an independent predictor of short-term and long-term all-cause mortality. (LOE = 1b-)

Reference

Shehabi Y, Sterba M, Garrett PM, et al, for the ProGUARD Study Investigators and the ANZICS Clinical Trials Group. Procalcitonin algorithm in critically ill adults with undifferentiated infection or suspected sepsis. Am J Respir Crit Care Med 2014;190(10):1102-1110.

Study design: Randomized controlled trial (nonblinded)

Funding source: Foundation

Allocation: Concealed

Setting: Inpatient (ICU only)

Synopsis

Procalcitonin (PCT) is a sepsis biomarker that has been utilized to guide antibiotic use in different patient populations. In this study, the authors tested a PCT-algorithm using a 0.1 ng/mL cut-off to reduce antibiotic exposure in critically ill patients. Patients newly admitted to intensive care units (ICUs) who were receiving antibiotics for suspected infections were randomized, using concealed allocation, to receive PCT-guided care (n = 196) or standard care (n = 198). All patients had PCT levels drawn daily until discharge from the ICU or up to a maximum of 7 days. In the PCT group, antibiotics were stopped if PCT levels were negative (< 0.1 ng/mL), if PCT levels were borderline (0.1 - 0.25 ng/mL) and infection was unlikely, or if PCT levels decreased more than 90% from baseline values. In the standard care group, the treating clinician determined antibiotic use without knowledge of the PCT results. Baseline characteristics of the 2 groups were similar with regard to severity-of-disease scores and baseline PCT values. There was high compliance with the PCT algorithm, with less than 3% of study days when the algorithm was not followed. There was no significant difference detected between the 2 groups for the primary outcome of time to antibiotic cessation. However, duration of antibiotic use was longer than expected in the control group (11 days actual vs 9 days expected), so the study may have been underpowered to detect an expected 25% reduction. Nevertheless, the 2 groups were similar with regard to ICU and hospital lengths of stay, as well as ICU, hospital, and 90-day mortality rates. Of note, the rate of decline in PCT level over the first 72 hours was an independent predictor of hospital mortality and 90-day mortality, with a slower decline corresponding to a higher mortality.

Dr. Kulkarni is an assistant professor of hospital medicine at Northwestern University in Chicago.

Clinical question

Can a procalcitonin-based algorithm reduce antibiotic use in critically ill patients?

Bottom line

A procalcitonin-based algorithm using a 0.1 ng/mL cutoff does not significantly decrease the duration of antibiotic treatment in critically ill patients nor does it reduce length of stay or number of deaths. The rate of decline in the procalcitonin level over the first 72 hours, however, does serve as an independent predictor of short-term and long-term all-cause mortality. (LOE = 1b-)

Reference

Shehabi Y, Sterba M, Garrett PM, et al, for the ProGUARD Study Investigators and the ANZICS Clinical Trials Group. Procalcitonin algorithm in critically ill adults with undifferentiated infection or suspected sepsis. Am J Respir Crit Care Med 2014;190(10):1102-1110.

Study design: Randomized controlled trial (nonblinded)

Funding source: Foundation

Allocation: Concealed

Setting: Inpatient (ICU only)

Synopsis

Procalcitonin (PCT) is a sepsis biomarker that has been utilized to guide antibiotic use in different patient populations. In this study, the authors tested a PCT-algorithm using a 0.1 ng/mL cut-off to reduce antibiotic exposure in critically ill patients. Patients newly admitted to intensive care units (ICUs) who were receiving antibiotics for suspected infections were randomized, using concealed allocation, to receive PCT-guided care (n = 196) or standard care (n = 198). All patients had PCT levels drawn daily until discharge from the ICU or up to a maximum of 7 days. In the PCT group, antibiotics were stopped if PCT levels were negative (< 0.1 ng/mL), if PCT levels were borderline (0.1 - 0.25 ng/mL) and infection was unlikely, or if PCT levels decreased more than 90% from baseline values. In the standard care group, the treating clinician determined antibiotic use without knowledge of the PCT results. Baseline characteristics of the 2 groups were similar with regard to severity-of-disease scores and baseline PCT values. There was high compliance with the PCT algorithm, with less than 3% of study days when the algorithm was not followed. There was no significant difference detected between the 2 groups for the primary outcome of time to antibiotic cessation. However, duration of antibiotic use was longer than expected in the control group (11 days actual vs 9 days expected), so the study may have been underpowered to detect an expected 25% reduction. Nevertheless, the 2 groups were similar with regard to ICU and hospital lengths of stay, as well as ICU, hospital, and 90-day mortality rates. Of note, the rate of decline in PCT level over the first 72 hours was an independent predictor of hospital mortality and 90-day mortality, with a slower decline corresponding to a higher mortality.

Dr. Kulkarni is an assistant professor of hospital medicine at Northwestern University in Chicago.