Potential Confounders of Procalcitonin-Guided Antibiotic Therapy for Sepsis
Several factors that could potentially affect the results in this study (1), however, should be discussed. First, there was no precise information regarding the underlying pathogenesis of the severe sepsis or septic shock. Several treatment-associated factors, including the presence of drug-resistant bacteria and inadequate use of antibiotics, could have affected the outcomes of the patients with sepsis. Second, in studies showing the benefit of procalcitonin-guided antibiotic treatment in sepsis patients (2, 3), the primary site of infection in most of the patients (~70%) was the respiratory tract. Yan et al (4) showed that serum procalcitonin levels in patients with infections varied significantly according to the pathogenesis (i.e., abdominal infection, pneumonia, infective endocarditis, urinary tract infection, and catheter-related infection). Thus, subgroup analysis considering the infection site could be useful. Third, the interval between disease onset and enrollment of this study (1) could have affected the study results. More precise information would have allowed better understanding of the results.
Next, the authors evaluated the difference in longitudinal changes of procalcitonin concentration between baseline and day 4. Results of a previous study, however, suggested that decreased serum procalcitonin levels at an earlier stage (day 1) could predict mortality of patients with suspected sepsis (5). Furthermore, the kinetics of procalcitonin concentrations after day 1 in survivors nearly paralleled those of nonsurvivors in this study (Fig. 2 in ), suggesting that the presence of potential predictors of mortality could be identified earlier than day 4.
Finally, the study included a high percentage (~20%) of patients with baseline procalcitonin levels less than 0.5 µg/L, categorized as the “antibiotics discouraged group” in a landmark study (2). Whether a threshold of “decreased by 80%” could be applied to patients with low baseline procalcitonin is controversial. Excluding patients with baseline procalcitonin levels of less than 0.5 µg/L would be helpful for specific assessment of the procalcitonin-guided therapeutic strategy.
In conclusion, we appreciate the additional data provided by the authors for further verification of the association between procalcitonin-guided antibiotic therapy and sepsis-related mortality.