How reliable is procalcitonin in identifying bacterial infections (primary and secondary) in patients admitted with influenza being treated for pneumonia?

Comment by InpharmD Researcher

The sensitivity of procalcitonin (PCT) when determining bacterial infection ranges from 38-91%, with PCT levels ≥ 0.25 µg/L indicative of bacterial pneumonia and PCT levels ≤ 0.1 µg/L suggesting viral infection. A 2013 meta-analysis derived a PCT sensitivity of 0.84 and specificity of 0.64 when identifying secondary bacterial infections among patients with influenza pneumonia, although sensitivity was greater in critically ill patients at 0.91. PCT levels are generally used to guide antibiotic therapy in patients with pneumonia. However, due to the wide range in sensitivity, PCT levels are not recommended to be used as the sole marker for guiding antibiotic therapy, especially in critically ill patients.

Background

Per the 2019 clinical practice guidelines of the American Thoracic Society (ATS) and Infectious Diseases Society of America (IDSA) for the diagnosis and treatment of adults with community-acquired pneumonia (CAP), empiric antibiotic therapy should be initiated if CAP is clinically suspected and radiographically confirmed, independent of serum procalcitonin levels. Studies have suggested that procalcitonin levels may be used to distinguish the etiology of respiratory infection, specifically that procalcitonin levels of ≥ 0.25 µg/L indicate a high likelihood of bacterial pneumonia, while procalcitonin levels ≤ 0.1 µg/L indicate a high likelihood of viral infection. However, a recent study of hospitalized CAP patients failed to identify a procalcitonin threshold that allows for determination between bacterial and viral pathogens. Furthermore, the reported sensitivity of procalcitonin in determining bacterial infection ranges from 38% to 91%, reinforcing the fact that procalcitonin levels alone should not be used to justify withholding antibiotics from patients with CAP. Further studies in patients with radiographically confirmed pneumonia are warranted to elucidate the validity and safety of using procalcitonin levels in the guidance of antibiotic therapy in patients with pneumonia. [1]

The IDSA guidelines state that in patients with CAP and a positive influenza test, no biomarkers (such as a low procalcitonin level), and early clinical stability, early antibiotic discontinuation at 48 to 72 hours can be considered. Studies have shown a procalcitonin-guided pathway and serial procalcitonin (PCT) measurement can reduce the duration of antibiotic therapy in CAP, but these cases average in a length of treatment greater than standards of practices in the U.S. Serial procalcitonin measurement is likely to be useful primarily in settings where the average length of stay for patients with CAP exceeds normal practice (e.g., 5–7 days). However, procalcitonin levels may not be elevated in concurrent viral and antibacterial infection or with important pathogens such as Legionella and Mycoplasma. [1]

Procalcitonin (PCT) has been broadly investigated for the identification and management of bacterial infections, as its levels typically rise in presence of these infections. Causative infectious pathogens generate an inflammatory response resulting in release of pro-inflammatory and anti-inflammatory cytokines, including PCT. A higher level of PCT is released due to gram negative bacterial infections compared to gram positive, due to the distinct pathways of activation of the inflammatory cascades. Prompt identification of such infections would allow for more proactive antibiotic treatment. However, utilization of PCT for bacterial infections, despite its high specificity, is complicated due to certain circumstances that may result in increased PCT levels. Still, evidence supports the use of PCT to achieve antibiotic stewardship for treatment of community acquired pneumonia, which multiplies early in cases of infection. Based on data derived from numerous trials, PCT levels > 0.25 ng/mL are indicative of high likelihood of bacterial respiratory infection. Use of PCT to initiate antibiotic treatment has been found to result in lower risk of mortality, lower antibiotic use, and lower risk for antibiotic-related side effects. Bacterial co-infections are common in patients with influenza pneumonia, ranging from 20% to 30%, frequently resulting in worse outcomes, yet studies specific to the use of PCT for identification of bacterial respiratory co-infection in patients with influenza pneumonia are limited. [2]

A 2014 prospective cohort study and individual patient data meta-analysis evaluated the diagnostic utility of PCT in critically ill patients with pneumonia during the 2009 H1N1 influenza pandemic. The prospective cohort component involved 46 intensive care unit (ICU) admissions with pneumonia at a tertiary care hospital in Cologne, Germany, during the 2009 and 2010 influenza seasons. PCT and clinical characteristics were assessed within 24 hours of ICU admission. Additionally, data from five previously published studies were systematically reviewed and included in a meta-analysis, resulting in a pooled cohort of 161 patients with confirmed H1N1 influenza or bacterial pneumonia, or both. PCT levels were significantly elevated in patients with bacterial pneumonia (median 6.2 mcg/L, interquartile range [IQR] 0.9–20) compared to those with isolated H1N1 pneumonia (median: 0.56 mcg/L, IQR 0.18 to 3.33; p<0.0001). The area under for PCT in detecting bacterial pneumonia was 0.72 (95% confidence interval [CI] 0.64 to 0.80), increasing to 0.76 (95% CI 0.68 to 0.85) when patients with hospital-acquired pneumonia and immune-compromising disorders were excluded. At a 0.5 mcg/L PCT cutoff, the sensitivity and negative predictive value for bacterial pneumonia were 80.5% (95% CI 69.9 to 88.7) and 73.2% (95% CI 59.7 to 84.2), respectively, improving to 85.5% (95% CI 73.3 to 93.5) and 82.2% (95% CI 68.0 to 92.0) in patients with community-acquired pneumonia and no immune suppression. Although PCT demonstrated reasonable accuracy in distinguishing bacteria from viral pneumonia, the authors cautioned against its use as a sole marker for guiding antibiotic therapy in critically ill patients. [3]

A 2013 systematic review and meta-analysis evaluated the diagnostic accuracy of PCT in identifying secondary bacterial infections among patients with influenza pneumonia. The analysis included six studies from Korea, France, Italy, and Australia, involving a total of 416 patients, of whom 137 had confirmed bacterial coinfections. The findings indicated that PCT has a sensitivity of 0.84 (95% CI 0.75 to 0.90) and a specificity of 0.64 (95% CI 0.58 to 0.69), with an area under the summary receiver operating characteristic (ROC) curve of 0.68 (95% CI 0.64 to 0.72). The positive likelihood ratio of 2.31 (95% CI 1.93 to 2.78) was deemed insufficient for PCT to serve as a rule-in test, while the negative likelihood ratio of 0.26 (95% CI 0.17 to 0.40) suggested utility as a rule-out test. Subgroup analysis demonstrated improved diagnostic performance among intensive care unit (ICU) patients, with a sensitivity of 0.91 (95% CI 0.82 to 0.97) and a lower negative likelihood ratio of 0.14 (95% CI 0.06 to 0.31). The findings suggest that while PCT cannot be used as a standalone rule-in test for bacterial pneumonia in influenza patients, its strong rule-out capability, particularly in ICU settings, may help guide clinical decision-making and antimicrobial stewardship strategies. [4]

References:

[1] Metlay JP, Waterer GW, Long AC, et al. Diagnosis and Treatment of Adults with Community-acquired Pneumonia. An Official Clinical Practice Guideline of the American Thoracic Society and Infectious Diseases Society of America. Am J Respir Crit Care Med. 2019;200(7):e45-e67. doi:10.1164/rccm.201908-1581ST
[2] Carbonell R, Moreno G, Martín-Loeches I, Bodí M, Rodríguez A. The Role of Biomarkers in Influenza and COVID-19 Community-Acquired Pneumonia in Adults. Antibiotics (Basel). 2023;12(1):161. Published 2023 Jan 12. doi:10.3390/antibiotics12010161
[3] Pfister R, Kochanek M, Leygeber T, et al. Procalcitonin for diagnosis of bacterial pneumonia in critically ill patients during 2009 H1N1 influenza pandemic: a prospective cohort study, systematic review and individual patient data meta-analysis. Crit Care. 2014;18(2):R44. Published 2014 Mar 10. doi:10.1186/cc13760
[4] Wu MH, Lin CC, Huang SL, et al. Can procalcitonin tests aid in identifying bacterial infections associated with influenza pneumonia? A systematic review and meta-analysis. Influenza Other Respir Viruses. 2013;7(3):349-355. doi:10.1111/j.1750-2659.2012.00386.x

Literature Review

A search of the published medical literature revealed 3 studies investigating the researchable question:

How reliable is procalcitonin in identifying bacterial infections (primary and secondary) in patients admitted with influenza being treated for pneumonia?

Level of evidence

A - Multiple high-quality studies with consistent results  Read more→



Please see Tables 1-3 for your response.


Can procalcitonin help identify associated bacterial infection in patients with severe influenza pneumonia? A multicentre study

Design

Retrospective observational multicentre study

N= 103 

Objective

To determine whether procalcitonin (PCT) levels could help discriminate isolated viral from mixed (bacterial and viral) pneumonia in patients admitted to 23 French ICUs during the A/H1N1v2009 influenza pandemic

Study Groups

Bacterial co-infection (n= 48) 

Influenza only (n= 55)

Inclusion Criteria Confirmed diagnosis of H1N1 influenza infection, clinical pattern of community-acquired pneumonia (CAP), no prior antibiotics before ICU admission
Exclusion Criteria Suspected hospital-acquired influenza infection, documented non-pulmonary bacterial infection, severely immunocompromised patients
Methods

Procalcitonin (PCT) levels were measured using time-resolved amplified cryptate emission technology; microbiological investigations through blood cultures, urinary antigen tests, and respiratory tract secretions culture. The levels of PCT at admission were compared between patients with confirmed influenzae A pneumonia associated or not associated with a bacterial co-infection.

Duration Data collected within the first 48 hours of ICU admission
Outcome Measures

PCT levels to discriminate between viral and mixed pneumonia, sensitivity and specificity of PCT levels, association with clinical outcomes

Baseline Characteristics Characteristic Bacterial co-infection (n= 48) Influenza only (n= 55) p-Value OR (95% CI)
Male 26 (54.2%) 27 (49.1%) 0.74 0.61 (0.56 to 2.7)
Age, years, median (IQR) 43 (27 to 56) 43 (25 to 55) 0.87  0.99 (0.97 to 1.02)
No comorbidity 21 (43.7%) 12 (21.8%) 0.02  2.8 (1.15 to 6.8)
SAPS 3 score, median (IQR) 54 (42 to 70) 44 (34 to 54) 0.006  1.03 (1.007 to 1.05)
CURB-65 ≥3 21 (43.8%) 14 (25.5%) 0.051  2.3 (0.97 to 5.3)
Peak temperature, C (mean ± SD) 39 (38.8 to 39.6) 39 (38.5 to 40) 0.76
Alveolar infiltrates on chest X-ray 40 (83.3%) 29 (52.7%) 0.001 4.5 (1.7 to 12.4)
PaO2/FiO2 ratio, median (IQR) 116 (80 to 214) 178 (137 to 252) 0.04 -
PaO2/FiO2 <200 27 (56.2%) 20 (36.4%) 0.04 2.3 (1.0 to 5.1)
Admission WBC (x109/L), median (IQR) 8.7 (3.3 to 12.0) 8.4 (5.0 to 13.3) 0.38
Admission CRP (mg/L), median (IQR) 260 (110 to 347) 95 (57 to 161) 0.002
Results   Bacterial co-infection (n= 48) Influenza only (n= 55)    
Admission PCT*, µg/L, median (IQR) 29.5 (4.0 to 45.4) 0.5 (0.1 to 1.8)  <0.001
Mechanical ventilation 36 (75.0%) 26 (47.3%) 0.01 3.3 (1.4 to 8.1)
Length of ICU stay, days, median (IQR) 12.5 (4.5 to 29) 5 (3 to 16) 0.009
Death 10 (20.8%) 8 (14.6%) 0.4  1.5 (0.5 to 4.3)

*PCT levels were available for 52 patients (19 with and 33 without bacterial co-infection)

For a cut-off of 0.8 µg/L or more, the sensitivity and specificity of PCT for distinguishing isolated viral from mixed pneumonia were 91 and 68%, respectively. Alveolar condensation combined with a PCT level of 0.8 µg/L or more was strongly associated with bacterial co-infection (OR 12.9, 95% CI 3.2–51.5; p<0.001). 

Abbreviations: CURB-65= confusion, urea ≥7 mmol/l, respiratory rate ≥30, systolic blood pressure <90 mmHg or diastolic <60 mmHg, and age ≥65 years; WBC= white blood cells count; IQR= interquartile (25th–75th percentile) range; OR= odds ratio; SAPS 3= simplified acute physiology score 3; ICU= intensive care unit

Adverse Events No specific adverse events reported related to the study procedures
Study Author Conclusions PCT levels can help discriminate between viral and mixed pneumonia during influenza season. Low PCT levels combined with clinical judgment suggest bacterial infection is unlikely, potentially guiding antibiotic therapy decisions.
InpharmD Reseracher Critique The exclusion of prior antibiotic use enhances accuracy, but small sample size, lack of repeated PCT measurements, and observational design limit causal inference.
References:

Cuquemelle E, Soulis F, Villers D, et al. Can procalcitonin help identify associated bacterial infection in patients with severe influenza pneumonia? A multicentre study. Intensive Care Med. 2011;37(5):796-800. doi:10.1007/s00134-011-2189-1

 

Procalcitonin (PCT) Levels for Ruling-out Bacterial Coinfection in ICU Patients with Influenza: A CHAID Decision-Tree Analysis

Design

Secondary analysis of a prospective, multicentre, observational study

N= 972

Objective

To define which variables upon intensive care unit (ICU) admission could be related to the presence of community-acquired respiratory coinfection (CARC) 

Study Groups

Influenza A(H1N1)pdm09 without CARC (n= 776)

Influenza A(H1N1)pdm09 with CARC (n= 196)

Inclusion Criteria

Patients admitted to the ICU with PCR confirmed influenza A(H1N1)pdm09 virus infection and procalcitonin (PCT) measurements upon admission

Exclusion Criteria

Patients <15 years of age

Methods

Data was collected and analyzed for patients admitted to the ICUs in Spain with confirmed influenza A(H1N1)pdm09 infections, with or without concurrent CARC. Serum PCT was not standardized and measured at admission per physician discretion. Data analysis included using a Chi-squared Automatic Interaction Detection (CHAID) classification tree to identify clinical variables to determine the liklihood of CARC.

Duration

June 2009 to April 2014

Outcome Measures

PCT and CRP values in identifying CARC

Baseline Characteristics

 

Without CARC (n= 776)

With CARC (n= 196)

p-Value

Age, years

50.8 ± 14.9 52.7 ± 14.0 0.120

Male

59.1% 66.3% 0.030

Severity of illness

APACHE II score (IQR)

SOFA score (IQR)

 

15.0 (10-20)

6.0 (3-8)

 

18.0 (13-22)

7.0 (4-10)

 

<0.001

<0.001

Comorbidities*

Obesity

 

3.2%

 

26.5%

 

<0.020

Complications

Shock

Mechanical ventilation

Acute renal injury

 

53.6%

80.4%

22.6%

 

70.9%

83.2%

38.8%

 

<0.001

0.410

<0.001

Clinical biomarkers

PCT, ng/ml (IQR)

CRP, mg/dl (IQR)

 

0.5 (0.2-2.0)

25 (13.2-88.2)

 

2.4 (0.6-11.7)

38.5 (19.5-154)

 

<0.001

0.620

Abbreviations: IQR, interquartile range; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, sequential organ failure assessment; PCT, procalcitonin; CRP, C-reactive protein

*No difference in remainder of comorbidities (p> 0.05)

Results

Endpoint

Without CARC (n= 776)

With CARC (n= 196)

p-Value

ICU mortality 

23.1% 32.1% <0.010

PCT ​<0.29 ng/mL: high sensitivity (88.2%), low specificity (33.2%), and high negative predictive value (91.9%) with post-test probability of CARC (6%)

Adverse Events

N/A

Study Author Conclusions

In patients admitted to the ICU with confirmed influenza A(H1N1)pdm09 infection and without shock, low serum levels of PCT may be a good marker for ruling out the presence of CARC. However, while PCT can assist physicians in developing patient-specific therapeutic plans, it is important to highlight that biomarkers are tools that should never replace physician decision-making.

InpharmD Researcher Critique

Limitations include an unblinded study design, lack of multiple or serial PCT values, and non-standardized PCT lab measurement as this was up to the discresion of the provider.



References:

Rodríguez AH, Avilés-Jurado FX, Díaz E, et al. Procalcitonin (PCT) levels for ruling-out bacterial coinfection in ICU patients with influenza: A CHAID decision-tree analysis. J Infect. 2016;72(2):143-151. doi:10.1016/j.jinf.2015.11.007

 

Prognostic Value of Procalcitonin and C-Reactive Protein in 1608 Critically Ill Patients with Severe Influenza Pneumonia

Design

Multicenter, retrospective cohort analysis

N= 1608

Objective

To evaluate the association of baseline serum procalcitonin (PCT) and C-reactive protein (CRP) levels and mortality, among patients with sepsis and septic shock, with a respiratory source of infection

Study Groups

Primary viral pneumonia (PVP; n= 1186)

Bacterial co-infection (BC; n= 422)

Inclusion Criteria

Adult patients admitted to the ICU with pneumonia due to influenza virus, with or without bacterial co-infection, in whom PCT and CRP were determined upon admission

Exclusion Criteria

Missing PCT and CRP values at admission or in ICU mortality data; non-respiratory source of infection

Methods

Data that were collected at ICU admission included demographic and general characteristics, underlying diseases, microbiological results, laboratory data, and complications during ICU stay. The severity of illness was evaluated by the APACHE II score calculated for all patients within the first 24 hrs of ICU admission and organ failure was assessed using the SOFA score.

Regarding co-infection, the isolated pathogens were divided into gram-positive cocci (GPC), gram-negative bacilli (GNB), polymicrobial (PM) infection, and Aspergillus spp. The serum levels of both biomarkers were also determined for the most frequent microorganisms.

Duration

June 2009 to April 2018

Outcome Measures

PCT and CRP levels, ICU mortality, duration of mechanical ventilation, length of ICU stay

Baseline Characteristics

 

PVP (n = 1186)

BC (n = 422)

 p-value

Age, years

55 (44 to 66) 60 (48 to 72) 0.001

Male

58.8% 62%  0.24 

APACHE II Score

16 (11 to 21)  19 (14 to 25) 0.001

SOFA Score

6 (4 to 8) 7 (4 to 10) 0.001

Comorbidities

        COPD

        Asthma

        Chronic Heart Failure

        Chronic Kidney Disease

        Hematologic Disease

        Pregnancy

        Obesitya

 

6.5 %

17.1%

12.6%

9.2%

8.4%

14.2%

34.3%

 

7.3%

29.4%

14.5%

8.8%

9%

17.1%

5.5%

 

0.45

0.001

0.34

0.77

0.73

0.16

0.01

Laboratory findings

        PCT, ng/mL

        CRP, mg/dL

        White Blood Cell Count, 109/L

 

0.6 (0.2 to 2.31)

28.1 (13.3 to 109)

7.1 (4 to 11.6)

 

4.35 (0.6 to 19.7)

36.8 (20.4 to 118)

8.3 (2.9 to 15.2)

 

0.001

0.001

0.06 

Complications

        Acute Kidney Failure

        CRRT

        Mechanical Ventilation

        Shock on Admission

        MODS

 

28.4% 

11.8%

82.1%

53.5%

65.3%

 

45.7%

17.8%

81%

67.3%

76.5% 

 

0.001

0.002

0.65

0.001

0.001

Microorganisms Isolated

       Gram Positive Cocci

       Gram Negative Bacilli

       Polymicrobial

       Aspergillus spp.

 

-

-

-

-

 

67.3%

23.7%

5.7%

3.3%

 

-

-

-

-

Abbreviations: PVP, primary viral pneumonia; BC, bacterial coinfection; APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, sequential organ failure assessment; COPD, chronic obstructive pulmonary disease; PCT, procalcitonin; CRP, C-Reactive protein; CRRT, Continuous renal replacement therapy; MODS, Multiple organ dysfunction syndrome

a Defined as body mass index >30 kg/m2

Results

Clinical Outcomes

PVP (n = 1186)

BC (n = 422)

p-Value

Mechanical Ventilation, days

10 (4-19) 10 (4-18) 0.85

ICU Length of Stay, days

10 (5-21) 10 (5-19) 0.73 

ICU Mortality

22.3% 27.7% 0.02 

Adverse Events

N/A

Study Author Conclusions

A single initial serum measurement of the biomarkers (PCT and CRP) have a limited value to predict mortality in critically ill patients with influenza pneumonia. In this population, none of the biomarkers presented an adequate predictive value to determine the causative pathogen that is responsible for co-infection, so it should not be used to tailor empiric antimicrobial therapy.

InpharmD Researcher Critique

This study was limited by its retrospective study design and included critically ill patients with sepsis or septic shock, resulting in limited generalizability. The study groups were unbalanced in severity of illness as the patients in the BC group had higher APACHE II and SOFA scores at baseline, limiting the utility of PCT and CRP in identifying bacterial co-infection in critically ill patients with viral pneumonia. 



References:

Carbonell R, Moreno G, Martín-Loeches I, et al. Prognostic Value of Procalcitonin and C-Reactive Protein in 1608 Critically Ill Patients with Severe Influenza Pneumonia. Antibiotics (Basel). 2021;10(4):350. Published 2021 Mar 26. doi:10.3390/antibiotics10040350