An algorithm for PCT-guided antimicrobial therapy: a consensus statement by Japanese experts

Introduction

The number of deaths due to infectious diseases by multidrug-resistant (MDR) pathogens has been increasing. Overuse of antibiotics is one of the major risk factors for the increase in MDR pathogens [1]. In Japan, the government formulated an antimicrobial resistance (AMR) action plan in April 2016, including appropriate use of antimicrobial agents to reduce the spread of AMR pathogens. The AMR action plan contains concrete goals for reducing the rate of AMR pathogens and antimicrobial agent consumption by 2020. Unfortunately, however, the goals have not yet been achieved, specifically with respect to intravenous antibiotic usage [2], with a reduction rate of intravenous antibiotics of only 2.7% in comparison with the goal of 20% [2]. Therefore, further effort is definitely needed to decrease antibiotic consumption.

Procalcitonin (PCT) is a better inflammatory biomarker for diagnosing bacterial infection and sepsis compared with C-reactive protein [3]. Many randomized controlled trials (RCTs) of lower respiratory tract infections showed that PCT-guided antibiotic therapy significantly reduced antibiotic consumption, including the rate of antibiotic therapy commencement and duration of antibiotic usage, without increasing mortality in both mild to moderate and critically-ill patients [4, 5]. In 2017, based on the results of 32 RCTs, the Cochrane library showed that PCT-guided therapy significantly reduced antibiotic consumption and their side effects, without worsening prognosis, in acute respiratory tract infections, and supported the use of PCT-guided therapy to reduce antibiotic usage [6].

Studies on PCT-guided therapy have adopted optimal cut-off levels of PCT, with a value of <0.1 μg/L indicating no bacterial infection, 0.1 to <0.25 μg/L being less likely to indicate bacterial infection, 0.25 to <0.5 μg/L indicating a possible bacterial infection, and ≥0.5 μg/L as highly likely to indicate bacterial infection [4]. However, having many cut-off points increases the complexity in clinical practice. To resolve this problem, in September 2018, some experts in Europe and the USA held a meeting to discuss the appropriate way to use PCT-guided therapy and optimal PCT cut-off levels. They proposed several different algorithms for PCT-guided therapy in mild to moderate cases outside the intensive care unit (ICU) and severe cases in the ICU [7].

However, to date, only a few studies have investigated the usefulness of PCT-guided antimicrobial therapy in Japan [8, 9]. Additionally, PCT-guided antimicrobial therapy appears to be not as widespread in daily clinical practice in Japan due, in part, to the restriction of payment in the medical insurance system. Therefore, we considered that there is a room for establishing a modified Japan-specific algorithm for PCT-guided antimicrobial therapy that would fit Japanese medical circumstances.

Materials and methods

To create a modified version of a PCT-guided algorithm in Japan, five Japanese experts (NS, SF, SF, KK and AI) and Dr. Schuetz (the chairperson of the international experts consensus) held an online meeting on 8th February 2022 (Supplementary Table 1). The areas of expertise of the doctors covered intensive care medicine, emergency medicine, respiratory medicine and infectious diseases (with three intensive care medicine, two emergency medicine, three respiratory medicine and five infectious disease specialists, with overlap). We discussed the appropriate use of PCT-guided antimicrobial therapy based on the international experts consensus [7]. Clinical evidence on PCT-guided therapy, adherence to the algorithm, infectious diseases for which the algorithm can be adopted, and cut-off levels and optimal follow-up timing of PCT were discussed. After the meeting, we had follow-up discussions via e-mail and online meetings.

Results

We created a modified version of the PCT-guided antimicrobial therapy algorithm (Japanese experts consensus), as shown in Figure 1. The Japanese algorithm targets hospitalized patients stratified according to whether they are admitted to medical wards (non-ICU setting) or in the ICU. The optimal cut-off PCT level to determine the presence of bacterial infections is set at 0.25 μg/L in a non-ICU setting and 0.5 μg/L in ICU settings, as in the international experts consensus [7]. If the PCT level is lower than these cut-off levels, or there is more than an 80% decrease from peak levels, the algorithm recommends discontinuing antibiotic therapy and/or assessing for the possible presence of non-infectious illnesses.

Figure 1:

The proposed algorithm for PCT-guided antimicrobial therapy in Japan (Japanese experts consensus) in patients with (A) pneumonia of moderate severity outside the ICU and (B) critically-ill patients with infections in the ICU. (A) ^ad0, d1, d3-4 and d5-7 means on admission, 1 day, 3–4 days and 5–7 days after admission, respectively. ^bCaution is required in patients with immuno-suppression (including HIV), pancreatitis, trauma, pregnancy, high volume transfusion. ^cPCT-guided stewardship should not be used in patients with chronic infections, including lung abscess, empyema, infected lung bullae or recurrent aspiration pneumonia. Assessment of these conditions is needed if PCT-guided stewardship is applied. ^dPCT-guided stewardship should not be used for pneumonia due to Pseudomonas aeruginosa, Staphylococcus aureus and Legionella species because there might be recurrence of pneumonia if antibiotic therapy is discontinued within 7 days, based on PCT-guided stewardship. (B) ^ad0, d1, d3-4 and d5-7 means on admission, 1 day, 3–4 days and 5–7 days after admission, respectively. ^bCaution is required in patients with immuno-suppression (including HIV), pancreatitis, trauma, pregnancy, high volume transfusion. ^cPCT-guided stewardship should not be used in patients with chronic infections, including lung abscess, empyema, infected lung bullae or recurrent aspiration pneumonia. Assessment of these conditions is needed if PCT-guided stewardship is applied. ^dPCT-guided stewardship should not be used for pneumonia due to Pseudomonas aeruginosa, Staphylococcus aureus and Legionella species because there might be recurrence of pneumonia if antibiotic therapy is discontinued within 7 days, based on PCT-guided stewardship.

Discussion

There are three major differences between the Japanese and the international algorithm. First, in the Japanese algorithm, the target patients are limited to only hospitalized patients, and the non-ICU patients only include those with respiratory infections, including community-acquired pneumonia (CAP), nursing and healthcare-associated pneumonia, hospital-acquired pneumonia and aspiration pneumonia. A previous report showed that respiratory infections accounted for the greatest proportion of antibiotic consumption among other diseases in hospitalized patients [10]. In addition, many RCTs have shown that PCT-guided therapy specifically for respiratory tract infections is useful for reducing antibiotic consumption without worsening prognosis [4, 6]. Therefore, we believe that clinical application of a PCT-guided algorithm selectively for this population would efficiently lead to reduction in overall antibiotic consumption.

The latest CAP guidelines published by the American Thoracic Society and Infectious Diseases Society of America in 2020 recommend a short treatment duration of 5 days for CAP if clinical stability is achieved within the first 48–72 h [11]. In addition, the CAP guidelines mention that antibiotics should be administered if CAP is suspected clinically, regardless of PCT levels [11]. Therefore, the PCT-guided algorithm might not necessarily be adaptable for all CAP cases. However, studies from Japan have indicated that the mean duration of intravenous antimicrobial therapy for CAP ranges from 7.4 to 9.4 days [12], suggesting that there is room for reduction of the duration of antibiotic therapy for CAP.

When devising this algorithm, we considered that PCT-guided antibiotic therapy could not be applied for recurrent aspiration pneumonia, although a previous report in Japan indicated that, in aspiration pneumonia, PCT-guided antimicrobial therapy was able to reduce antibiotic duration without worsening prognosis, including relapse and death [8]. However, considering the frequent relapse of aspiration pneumonia [13], we believe that it might not be feasible to adapt the PCT-guided algorithm for recurrent aspiration pneumonia.

Second, we thought that PCT-guided antimicrobial therapy might not be applicable for pneumonia due to Pseudomonas aeruginosa, Staphylococcus aureus and Legionella species, because short-term antibiotic therapy might lead to recurrence of pneumonia due to these etiologies [14, 15]. Recent opinions recommended setting the treatment duration for pneumonia due to P. aeruginosa as 8–15 days, S. aureus as 14 days and Legionella species as 14–21 days [14, 15]. The safety and feasibility of reducing the treatment duration for these pneumonias based on PCT-guided algorithms remains uncertain.

Finally, we modified the timing and frequency of PCT follow-up considering the restrictions of the Japanese medical insurance system. The in-hospital cost of PCT measurements is included in the total admission cost in hospitals that have adopted the Diagnosis Procedure Combination (DPC) system, whereas PCT measurement costs need to be paid by pay-for-performance in non-DPC hospitals. In order to strike a balance between PCT measurement costs and the importance of frequent assessment of PCT values, we recommend the following PCT test schedule: (A) in DPC hospitals or ICU settings: on admission (day 0), 1 day after admission to accurately measure the peak level (day 1), 3–4 days after admission (days 3–4), and 5–7 days after admission (days 5–7) to assess antibiotic withholding (a total of four measurements), (2) in both non-DPC hospitals and non-ICU settings: on admission (day 0) and 3–5 days after admission (days 3–5) to assess antibiotic withholding (a total of two measurements).

The adapted PCT-guided algorithm has high potential to help clinicians safely discontinue antibiotic therapy, along with consideration of the patient’s clinical course, including symptoms, vital signs, radiological findings and microbiological findings. Implementation of the PCT-guided algorithm, together with clinician education and feedback, is expected to lead to appropriate antibiotic usage and reduction of AMR pathogens. In the future, the efficacy of the algorithm in Japanese clinical practice should be validated.