How Procalcitonin Can Guide Early Antibiotic Discontinuation

Part of the procalcitonin series

During my medical training, procalcitonin was rarely used in practice, despite some of my teaching sessions presenting it as a tool for distinguishing viral from bacterial infections. Even now, with greater exposure, I find myself less familiar with procalcitonin than with other infectious markers such as CRP or white blood cell count. This series explores procalcitonin’s value in clinical practice. In this blog, how procalcitonin can guide early antibiotic discontinuation, focusing on three key questions:

1) What is the evidence for using procalcitonin-guided algorithms in antibiotic discontinuation?

2) What are the main points of criticism against procalcitonin-guided algorithms?

3) How to implement a procalcitonin-guided algorithm in clinical practice?

1) What is the evidence for using procalcitonin-guided algorithms in antibiotic discontinuation?

Procalcitonin-guided algorithms for antibiotic discontinuation have been investigated in over 20 randomized controlled trials, generally demonstrating positive outcomes (1). The following table provides an overview of some of the most important trials (2-8):

A recent meta-analysis, including 26 randomized controlled trials and 9048 patients, confirmed that procalcitonin guidance reduces antibiotic duration without affecting ICU or hospital length of stay and, possibly with a survival benefit (1). However, critics have raised concerns about the findings of these trials (8-12).

2) What are the main points of criticism against procalcitonin-guided algorithms?

I. Suboptimal standard of care in the control group.

Critics argue that control groups in above mentioned trials often received antibiotics for longer than recommended by guidelines (8,11). For instance, in ADAPT-sepsis, critically ill patients received an average of 10 days of antibiotics in the first 28 days after randomization (7), substantially exceeding antibiotic durations in reference trials such as BALANCE or STOP-IT (9). However, these 10 days include treatment for both initial and subsequent infections. Moreover, in the SAPS study, median antibiotic duration among its critically ill patients was only 5 days in the procalcitonin-guided group versus 7 days in the standard of care, demonstrating procalcitonin’s potential to reduce antibiotic use even in settings with stringent stewardship (4).

II. Higher recurrence of infections.

The aforementioned meta-analysis observed a significant increase in recurrent infections in the procalcitonin-guided group (with odds ratio of 1.36) (1). Since most included randomized trials were open-label, this finding is likely due to ascertainment bias for recurrent infections (1,4). No difference in mortality or ICU/ hospital length of stay was after all demonstrated (1).

III. Lack of clear long-term benefits

Some authors question the clinical significance of an average 0.88-days reduction in antibiotic duration (as in ADAPT-sepsis) or similar reductions in other trials (12). However, the PROGRESS trial showed a significant decrease in infections caused by multidrug-resistant organisms and C. Difficile (5). Notably, in this study, the procalcitonin group received 5 days less antibiotics than the standard-of-care group (5).

IV. Poor adherence to procalcitonin-guided algorithms

Apart from the ProGUARD trial, procalcitonin-guided algorithms were followed with only moderate adherence at best in the highlighted studies (2-6,8). For instance, in ADAPT-sepsis, a strong recommendation to stop antibiotics was issued after on average 3.5 days in the procalcitonin group, yet the initial antibiotic course was continued for 7 days (7). It remains an open question whether better adherence would lead to greater harm (7). However, overly strict adherence to procalcitonin-guided algorithms is also undesirable, given their moderate diagnostic accuracy (13).

3) How to implement a procalcitonin-guided algorithm in clinical practice?

Despite promising results in randomized trials, real-world implementation of procalcitonin-guided algorithms has yielded mixed outcomes.

• In a retrospective cohort study of 20750 critically ill patients, procalcitonin-guided algorithms did not reduce antibiotic prescription or C. difficile infections (14). However, only a third of patients with procalcitonin measurements underwent sequential testing, indicating poor implementation (14).

• Similarly, among 35 610 patients with lower respiratory tract infections in a retrospective, propensity score-matched cohort study, procalcitonin failed to significantly impact antibiotic duration (15). Again, implementation was poor with scarce serial monitoring (15).

• Nevertheless, some case-control studies found that complying with procalcitonin-guided algorithms was associated with shorter antibiotic duration (14, 16).

As correct implementation of procalcitonin-guided algorithms seems key, the following decision rules should be followed:

a. Not all infections can be monitored with procalcitonin. As mentioned in an earlier post, procalcitonin performs poorly in conditions such as cellulitis or subacute bacterial endocarditis. Additionally, baseline levels are often below the stopping threshold, making procalcitonin impossible to use (8). For example, in a meta-analysis of 6708 patients with respiratory infections, procalcitonin levels were below 0.1 µg/L and 0.5 µg/L in 30% and 60% of cases, respectively (17).

b. Serial procalcitonin measurements are necessary.

Procalcitonin trends over time - not single values - can be used to guide antibiotics discontinuation as demonstrated by the above highlighted randomized controlled trials (2-7).

c. A liberal procalcitonin-guided algorithm is more potent.

Although clinicians adhered better to the conservative algorithm of the ProGUARD-trial (2-6), this study was one of the few trials that failed to significantly reduce antibiotic duration (3). Moreover, a meta-analysis demonstrated the same ineffectiveness of conservative protocols (1).

The most widely studied stopping rule recommends considering antibiotic discontinuation when procalcitonin level drops by more than 80% from its peak or falls below 0.5 µg/L (6,7).

d. Procalcitonin-guided algorithms should be part of a broader assessment to stop antibiotics.

Procalcitonin has insufficient sensitivity and specificity to be used as a standalone test (13). As part of a broader assessment, however, there is an impressive amount of evidence supporting its use in antibiotic discontinuation (1).

References:

1. Papp M, Kiss N, Baka M, et al. Procalcitonin-guided antibiotic therapy may shorten length of treatment and may improve survival-a systematic review and meta-analysis. Crit Care. 2023 Oct 13;27(1):394.

 2. Bouadma L, Luyt CE, Tubach F, et al. Use of procalcitonin to reduce patients' exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial. Lancet. 2010 Feb 6;375(9713):463-74.

 3. Shehabi Y, Sterba M, Garrett PM, et al. Procalcitonin algorithm in critically ill adults with undifferentiated infection or suspected sepsis. A randomized controlled trial. Am J Respir Crit Care Med. 2014 Nov 15;190(10):1102-10.

 4. de Jong E, van Oers JA, Beishuizen A, et al. Efficacy and safety of procalcitonin guidance in reducing the duration of antibiotic treatment in critically ill patients: a randomised, controlled, open-label trial. Lancet Infect Dis. 2016 Jul;16(7):819-827.

 5. Kyriazopoulou E, Liaskou-Antoniou L, Adamis G, et al. Procalcitonin to Reduce Long-Term Infection-associated Adverse Events in Sepsis. A Randomized Trial. Am J Respir Crit Care Med. 2021 Jan 15;203(2):202-210.

 6. Dark P, Hossain A, McAuley DF, et al. Biomarker-Guided Antibiotic Duration for Hospitalized Patients With Suspected Sepsis: The ADAPT-Sepsis Randomized Clinical Trial. JAMA. 2025 Feb 25;333(8):682-693. doi: 10.1001/jama.2024.26458. Erratum in: JAMA. 2026 Jan 13;335(2):187.

 7. Gupta S, Klompas M, Rhee C. Reassessing Procalcitonin-Guided Antibiotic Therapy in Critically Ill Patients With Sepsis: Lessons From the ADAPT-Sepsis Trial. Clin Infect Dis. 2026 Mar 17;82(3):453-458.

 8. Lisboa T, Salluh J, Povoa P. Do we need new trials of procalcitonin-guided antibiotic therapy? Crit Care. 2018 Jan 27;22(1):17.

 9. Hsu CK, Lai CC. Are Procalcitonin Measures a Reliable Predictor of Stopping Antibiotics Among Patients With Sepsis? JAMA. 2025 May 20;333(19):1728-1729.

 10. Dark P, Hossain A, Lall R. Are Procalcitonin Measures a Reliable Predictor of Stopping Antibiotics Among Patients With Sepsis?-Reply. JAMA. 2025 May 20;333(19):1730-1731.

 11. Arrigo M, Huber LC. Are Procalcitonin Measures a Reliable Predictor of Stopping Antibiotics Among Patients With Sepsis? JAMA. 2025 May 20;333(19):1729.

 12. Bosch NA, Wilson KC, Law AC. Are Procalcitonin Measures a Reliable Predictor of Stopping Antibiotics Among Patients With Sepsis? JAMA. 2025 May 20;333(19):1728.

 13. van Oers JAH, Nijsten MW, de Lange DW. Do we need new trials of procalcitonin-guided antibiotic therapy? A response. Crit Care. 2018 Mar 23;22(1):83.

 14. Chu DC, Mehta AB, Walkey AJ. Practice Patterns and Outcomes Associated With Procalcitonin Use in Critically Ill Patients With Sepsis. Clin Infect Dis. 2017 Jun 1;64(11):1509-1515.

 15. Heren JE, Lund BC, Alexander B, et al. Procalcitonin Monitoring and Antibiotic Duration in Presumed Lower Respiratory Tract Infections: A Propensity Score-Matched Cohort Across the Veterans Health Administration. Open Forum Infect Dis. 2023 Oct 25;10(11):ofad520.

 16. Péju E, Dargent A, Roudaut JB, et al. Impact of Complying with a Procalcitonin-Guided Stopping Rule on the Duration of Antibiotic Therapy in Critically Ill Patients: A Real-Life Study. Antibiotics (Basel). 2025 Oct 11;14(10):1012.

 17. Schuetz P, Wirz Y, Sager R, et al. Effect of procalcitonin-guided antibiotic treatment on mortality in acute respiratory infections: a patient level meta-analysis. Lancet Infect Dis. 2018 Jan;18(1):95-107.