Procalcitonin reduces antibiotic prescribing for acute bronchitis but not other causes of lower respiratory tract infection

Reference

Huang DT, Yealy DM, Filbin MR, et al, for the ProACT Investigators. Procalcitonin-guided use of antibiotics for lower respiratory tract infection. N Engl J Med 2018 May 20. doi: 10.1056/NEJMoa1802670. [Epub ahead of print]

Synopsis

PCT is a biomarker associated with bacterial infection. The PCT test has been used with mixed results as a way to influence antibiotic prescribing for infections, with most studies conducted in the inpatient setting. This US study, set in 14 emergency depatments, identified patients presenting with symptoms of a lower respiratory tract infection for whom the physician was uncertain about whether antibiotics were indicated. The criteria for diagnosing lower respiratory tract infection were not reported. Patients were randomized to receive PCT-guided care or usual care. The former included measurement of PCT using a rapid point-of-care test with the following guidance: < 0.1, antibiotics strongly discouraged; 0.1 - 0.25, antibiotics discouraged; 0.25 - 0.5 antibiotics recommended; and > 0.5 antibiotics strongly recommended. A total of 1664 patients were randomized to receive care, and after 8 withdrew from the study, a total of 1656 were included (mean age 53 years). The final diagnosis was community-acquired pneumonia in 20% of the patients, COPD exacerbation in 32%, and acute bronchitis in only 25%; 45% of patients were hospitalized. This was a fairly sick group, with more than half having a clear indication for antibiotics. Clinicians were adherent to the PCT protocol 73% of the time: Adherence was highest when the final diagnosis was acute bronchitis ("Now I can justify no antibiotic for the patient!") and lowest when the diagnosis was community-acquired pneumonia ("If I think they have pneumonia, of course they need an antibiotic regardless of PCT level!"). Overall, the intervention had no effect on antibiotic prescriptions, with 4.2 vs 4.3 days of antibiotics consumed in the PCT and usual care groups, respectively. Although fewer patients in the PCT group were prescribed an antibiotic (57% vs 61.8%), this was not statistically significant using the prespecified 99.86% confidence interval (a Bonferroni correction was used because there were 36 secondary comparisons). In the subgroup of patients with a final clinical diagnosis of acute bronchitis, far fewer received an antibiotic in the PCT-guided care group (17.3% vs 32.1%; risk difference -14.8%; 95% CI -28.5 to -1.1; number needed to treat to avoid one antibiotic prescription = 7).