Table 2 Main characteristics of the included studies

Oliveira, C. F. et al., 2013, Crit Care Med Oct;41(10):2336–2343

Patients with severe sepsis or septic shock, admitted to the intensive care units at two university hospitals in Brazil; the authors compared the duration of antibiotic therapy between the group treated with the aid of a CRP-based protocol versus a procalcitonin-based protocol. A control group without biomarkers was lacking. They observed the PCT protocol was not superior to the CRP protocol level for reducing the use of antibiotics (Duration of antibiotic therapy in PCT protocol: 8.1 (3.7) days vs Duration of antibiotic therapy in CRP protocol: 7.2 (3.5) days; p = 0.25)

Mean age (years) CRP: 59.6 ± 18.5 Procalcitonin: 59.6 ± 13.3 Total: 59.8 ± 16.8 Male n(%): CRP: 26 (45,61%) Procalcitonin: 31 (54,39%) Total: 57 (60.6%) Females: n(%) CRP: 19 (42,2%) Procalcitonin: 18(48,8%) Total: 37 (39.4%)

Method: Discontinuation of antibiotic therapy based on serum levels of CRP (CRP < 25 mg/L or decrease ≥ 50%) and PCT (PCT < 0.1 ng/ml or decrease ≥ 90%) associated with SOFA score analysis

Borges, I. et al., 2020, Crit Care. Jun;24: 281

Critically-ill patients admitted to the ICU of a university hospital in Brazil; authors compared the duration of antibiotic therapy using a CRP-based protocol versus an evidence-based strategy without the use of biomarkers. They observed that the CRP-based strategy was viable in reducing antibiotics exposure (Intervention group median: 7 (5–8.8), Control group median: 7 (7–11.3), p = 0.011)

Median age (years) CRP: 62 (53–68) Control: 60 (49–70) years Total: 61 (51–68) years Male n(%): CRP: 34 (50%) Control: 34 (50%) Total: 68 (52,3%) Females n(%): CRP: 30 (48,4) Control: 32 (51,6%) Total: 62 (47.7%)

Method: Duration of antibiotic therapy based on serum levels of CRP (CRP < 35 mg/L or decrease ≥ 50%) versus a standard protocol without the use of biomarkers

von Dach, E. et al., 2020, JAMA Jun;323(21):2160–2169

Patients with gram-negative uncomplicated bacteremia in three Switzerland hospitals; authors compared the duration of antibiotic therapy guided by CRP levels versus two fixed-length antibiotic therapy groups (7 and 14 days). It was found that the CRP protocol was noninferior to the fixed-day treatment protocols treatment

Median age (years) CRP guided: 78 (69–86) 7 days: 78 (69–86) 14 days: 80 (67–85) Total: 79 (68–86) Male; n(%): CRP guided: 64 (32,5%) 7 days: 62 (31,5) 14 days: 71 (36%) Total: 197 (39%) Females n(%): CRP guided: 105 (34,3%) 7 days: 107 (35%) 14 days: 94 (30,7%) Total: 306 (61%)

Method: Duration of antibiotic therapy based on serum levels of CRP (decrease ≥ 75% from its peak associated with the absence of fever for the previous 48 h) versus a fixed-length of therapy (7 and 14 days)