|| Checking for direct PDF access through Ovid
Today, a total of 14 previous randomized controlled trials have demonstrated that clinical protocols integrating clinical assessment of patients and procalcitonin serum levels are efficient and safe for antibiotic stewardship of patients with respiratory infections and sepsis (1). Importantly, according to the risk of patients, different procalcitonin (PCT) algorithms were applied and, consequently, had different effects (2). Similar to other diagnostic tests such as D-dimers, PCT results need to be interpreted and applied in the context of the pretest probability for infection, overall morbidity and mortality and adapted to the clinical setting and the acuity. For example, in patients at low risk for severe bacterial infections (e.g., primary care patients with upper and nonpneumonic lower respiratory infections), PCT protocols were used to determine whether antibiotics should be initiated at all. Several studies showed a marked reduction of antibiotic prescription in this setting (3, 4). Conversely, in trials enrolling higher risk patients (ICU patients with sepsis  or hospitalized patients with pneumonia ), PCT was mainly used for monitoring of patients to determine when treatment could be safely discontinued. A recent meta-analysis showed that when these protocols were applied to ICU patients with respiratory infections, antibiotics were reduced by around 25% (median of 14 days vs. 11 days) without an excess increase in mortality (odds ratio 0.84, 95% confidence interval 0.54–1.31) (1). In light of these previous findings (1), the results of Layios and colleagues are not surprising, and we agree with their conclusion that “Procalcitonin measuring for the initiation of antimicrobials did not appear to be helpful in a strategy aiming at decreasing the antibiotic consumption in the intensive care unit” (7). We wonder, however, why such a design was chosen in this setting? Rather than relying on the initial PCT level only in these patients at high risk for complications, the focus should be on monitoring PCT levels for earlier discontinuation of antibiotics.PCT appeared to be increased in 33.8% of episodes with no confirmed infection. Importantly, previous research has shown that trauma and major surgery increase PCT due to the surgical stress per se. Thus the authors finding is not surprising in the light of around 40% surgical patients in their study population. Still, previous research focusing on the decrease of PCT postsurgery found lower antibiotic exposure (8), highlighting again that not initiation but rather discontinuation should be the target in this population.Still, the study by Layois et al provides some important new information. First, it is interesting to see that in patients with uncertain infection, initial PCT measurement reduced antibiotic prescription from 76.2% to 50.0% (p = 0.08) without apparent difference in outcomes. Second, initial PCT levels showed a stepwise increase with increasing likelihood for infection measured by the diagnosis confidence of infectious-disease specialists (Fig. 2 in ). Thus, initial PCT levels provide additional information about the risk for infection in ICU patients at high risk for adverse outcome. If PCT is low, infection becomes unlikely and other differential diagnoses should be considered; yet initial antibiotics may still be advised in a high mortality risk setting to minimize the risk for the patient. If circulating PCT levels steadily decline during follow up or remain low, early discontinuation is an evidence-based, effective and safe strategy to decrease antibiotic overconsumption in the ICU. If initial PCT levels are increased in nonsurgical/nontrauma patients, the risk for infection is increased and antibiotics should be started without delay.