Methods

Randomised, multicentre, single‐blind clinical trial in 8 French ICUs

Participants

Inclusion criteria: Adults admitted to a participating ICU were eligible if they, < 48 h, had SIRS, acute dysfunction of at least 1 organ, absence of indisputable clinical infection, and negative microbial cultures
Exclusion criteria: Pregnancy, burns over ≥ 15% of body surface area, trauma, outpatient or inpatient cardiac arrest, post‐orthopaedic surgery status, drug‐related neutropenia, withdrawal of life‐supportive therapies or a decision to withhold them, indisputable clinical infection or antibiotic exposure ≥ 48 h during the time shortly before ICU admission
Included in this study: 62/1250 screened patients were eligible for the study, of whom 31 were randomised to each arm. 4 post randomisation exclusion (4 withdrew their consent)

Interventions

Guiding antibiotic decisions in ICU patients with non‐microbiologically proven apparent severe sepsis
Algorithm used in this study: In the experimental arm, both initiation and discontinuation of antibiotics were guided by a PCT‐based algorithm, applied at 6 h and on day 3 and day 5 post randomisation. Briefly, antibiotic therapy was not to be started or was to be halted when PCT was < 0.25 μg/L, was strongly discouraged when PCT was ≥ 0.25 to < 0.5 μg/L, was recommended when PCT was ≥ 0.5 to < 5 μg/L, and was strongly recommended when PCT was ≥ 5 μg/L. Owing to the fact that surgery can increase PCT levels, for 12 participants enrolled in the 48‐hour postoperative period, the respective PCT cut‐offs were < 4 μg/L, ≥ 4 to < 9 μg/L, and ≥ 9 μg/L. Investigators were strongly advised not to overrule the algorithm every day up to the study day 5. In the control arm, the decision to start or stop antibiotic therapy was at the discretion of the participant’s physician, without knowledge of the participant’s PCT concentrations.

Outcomes

• proportion of participants receiving antibiotics at day 5 post randomisation

• death at day 5, at ICU discharge, and at hospital discharge

• proportion of participants started on antibiotics post randomisation

• duration of antibiotic exposure

• SOFA score at day 3 and day 5

• proportion of participants with infection acquired between randomisation and day 3, day 5, and ICU discharge

• ICU and hospital length‐of‐stay

Notes

Funding: Research grant partly by Thermo Fisher B·R·A·H·M·S France. The sponsor had no input in study design, conduct, or reporting.
Follow‐up time: Until hospital discharge or 30 days' post randomisation, whichever came first
Registration: NCT01025180

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Participants were randomised in a 1:1 ratio according to a computer‐generated list. Randomisation was centralised through a secured web site and performed by an independent statistician.

Allocation concealment (selection bias)

Low risk

Centralised randomisation using permutation blocks

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

The control arm remained blinded to PCT levels. Masking of antibiotic therapy was not feasible in this study.

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Investigators remained blinded to PCT levels in the control arm, but no blinding of overall outcome assessment was mentioned in the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for participants on antibiotic on day 5 was: 58/62. 4 participants withdrew their consent.

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol. Trial registered (NCT01025180).

Other bias

Unclear risk

Moderate adherence in the PCT arm: physicians were non‐compliant with the PCT‐based algorithm in 19% of participants at 6 h, 17% on day 3, and 37% on day 5.
The study was stopped prematurely owing to the low incidence of eligible patients. As a consequence, the study population is small with low statistical power.

Methods

Randomised, investigator‐initiated, multicentre, partially blinded clinical trial, in 33 multidisciplinary ICUs across Germany

Participants

Inclusion criteria: Adults with severe sepsis or septic shock (severe sepsis was defined as SIRS caused by infection combined with acute organ dysfunction. Septic shock was defined as sepsis in combination with arterial hypotension or need for vasopressor therapy despite adequate fluid resuscitation)
Exclusion criteria: Pregnant or lactating women, patients with selenium intoxication, individuals with infections for which guidelines recommend a longer duration of antimicrobial therapy, immunocompromised patients, and those without commitment to full therapy or where death was imminent owing to coexisting diseases were excluded from the trial.
Included in this study: 1180 participants were randomised; 91 participants were excluded from the final analysis because informed consent was not obtainable in the deferred consent process, resulting in 1089 participants with valid data.

Interventions

Guiding antibiotic decisions and effect of sodium selenite administration in people with severe sepsis or septic shock

Algorithm used in this study: Using a 2 × 2 factorial design, participants were randomly assigned to receive intravenous sodium selenite or placebo as well as antimicrobial therapy guided by a PCT algorithm or conventional antimicrobial therapy.
In participants randomised to the PCT guidance arm, PCT was measured locally on days 0, 1, 4, 7, 10, and 14 after randomisation if the participant was still in the ICU. Procalcitonin concentration on day 0 or day 1 served as the baseline value. Depending on the PCT results, an algorithm provided recommendations to change or discontinue antimicrobial therapy or trigger diagnostic procedures to optimise source control. On day 4, no change in antimicrobial therapy was recommended if the PCT level dropped by at least 50% compared with the baseline value. Otherwise, change or optimisation of antimicrobial therapy or interventions regarding source control were recommended. On the other days, stopping antimicrobial therapy was recommended if the PCT level was 1 ng/mL or lower or if the PCT level dropped by at least 50% compared with the previous value. Otherwise, change or optimisation of antimicrobial therapy or interventions regarding source control were recommended. The treating physician was allowed to overrule the algorithm recommendation.
In the group without PCT guidance, no PCT measurements were obtained until day 14; changes in antimicrobial therapy were made at the discretion of the treating physician. Investigators agreed to treat all participants according to the Guidelines of the Germany Sepsis Society, which included recommendations to re‐evaluate antimicrobial therapy after 48 to 72 hours and to restrict duration of antimicrobial therapy to no more than 10 days.

Outcomes

• death from any cause by 28 days after inclusion

• 90‐day all‐cause mortality

• mean total SOFA score and its sub scores

• duration of ICU and hospital stay

• ventilator‐, vasopressor‐, and dialysis‐free days until day 90

• duration and costs of antimicrobial therapy

• duration until change of antimicrobial therapy

• antimicrobial exposure days

• days free of antimicrobial therapy

• frequency of surgical source control

• frequency of diagnostic procedures for localisation of the infection focus

• clinical and microbiologic treatment response

• secondary infections

• emergence of antibiotic‐resistant bacteria

Notes

Funding: The study infrastructure was partially funded by grant 01 KI 0106 from the German Federal Ministry of Education and Research. Biosyn (Germany) and Thermo Fisher (Germany) provided study medication and financial support via unrestricted grants.
Follow‐up: 90 days
Trial registration: NCT00832039

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Web‐based, centralised randomisation

Allocation concealment (selection bias)

Low risk

quote "Using a 2 × 2 factorial design, we randomly assigned patients to receive intravenous sodium selenite or placebo as well as antimicrobial therapy guided by a PCT algorithm or conventional antimicrobial therapy without PCT guidance with an allocation ratio of 1:1:1:1 by use of a central randomisation web server. Randomisation was stratified by study centre, sex, and sepsis severity"

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Due to the study design, blinding of physicians was not feasible. Procalcitonin values were only available in participants with study intervention

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment was mentioned in study

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Postrandomisation exclusions were about even among intervention arms with a very low dropout rate for mortality

Selective reporting (reporting bias)

Low risk

Outcomes correspond to trial registration (NCT00832039)

Other bias

Unclear risk

Overruling of the PCT algorithm was very high (adherence was lower than 50%)

Methods

Randomised, multicentre clinical trial in 9 French ICUs

Participants

Inclusion criteria: Patients with suspected bacterial infections during ICU stay without prior AB (> 24 h)

Exclusion criteria: Aged under 18 years; known pregnancy; expected stay in the ICU of less than 3 days; bone marrow transplant or chemotherapy‐induced neutropenia (< 500 neutrophils per mL); infections for which long‐term antibiotic treatment is strongly recommended (i.e. infective endocarditis, osteoarticular infections, anterior mediastinitis after cardiac surgery, hepatic or cerebral abscesses, chronic prostatitis, or infection with Mycobacterium tuberculosis, Pneumocystis jirovecii, or Toxoplasma gondii); poor chance of survival, defined as a simplified acute physiology score (SAPS II) of more than 65 points at screening; and do‐not‐resuscitate orders.

Included in this analysis: 394 participants with CAP and VAP out of 630 randomised participants; 9 post randomisation exclusions (8 withdrew consent, 1 randomised twice), and 227 not considered for this analysis due to diagnosis other than ARI

Interventions

Guiding antibiotic decisions in ICU patients with repeated PCT measurements

Algorithm used in this study: Investigators were encouraged to discontinue ABs when PCT concentration was less than 80% of the peak concentration or an absolute concentration of less than 0.5 μg/L was reached.

Outcomes

• all‐cause mortality at day 28

• all‐cause mortality at day 60

• antibiotic use

• relapse or superinfection (days 1 to 28)

• number of days without mechanical ventilation (days 1 to 28)

• SOFA score (days 1, 7, 14, and 28)

• length of stay in the ICU and hospital

Notes

Funding: Research grant from the Départment à la Recherche Clinique et au Développement, Assistance Publique‐Hopitaux de Paris (PHRC AOR06019), France. B·R·A·H·M·S, Germany (manufacturer of PCT assay) provided all assay‐related materials free of charge for the study (Kryptor machines if not already available on‐site and kits and maintenance required for study‐related measurements).

Follow‐up: Fixed period of 60 days for mortality

Registration: NCT00472667

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Independent, centralised, computer‐generated randomisation sequence (CleanWEB, Telemedicine Technologies, Boulogne, France)

Allocation concealment (selection bias)

Low risk

Centralised randomisation

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Quote: "Although treatment assignments were not masked, all investigators were unaware of aggregate outcomes during the study and primary endpoints were strictly defined and not patient‐reported."

Quote: "An adjudication committee comprised of 4 specialists in infectious diseases and critical care medicine who were masked to the randomisation assignment reviewed and validated all infectious episode classifications by consensus."

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 393/394 (100%)

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol. Trial registered (NCT00472667).

Other bias

Unclear risk

Low adherence to PCT algorithm in PCT group (47%)

Methods

Single‐centre, randomised, open‐label clinical trial in Rochester, NY, USA

Participants

Inclusion criteria: Adults ≥ 21 years of age with symptoms compatible with LRTI (i.e. admission diagnosis of pneumonia, acute exacerbations of COPD, bronchitis, asthma, influenza, viral syndrome, respiratory failure, and congestive heart failure CHF) were identified by reviewing the daily admission census.
Exclusion criteria: Patients with characteristics indicative of a high risk for bacterial infection (i.e. ICU requirement, active chemotherapy or radiation, immunosuppression, definitive infiltrate on chest radiograph, enrolment systolic blood pressure of < 90 mmHg, and ≥ 15% band forms in peripheral blood). Patients who had conditions known to increase PCT levels (i.e. trauma, renal failure, and pancreatitis) or who had received antibiotics prior to admission were also excluded.
Included in this study: 300/685 screened patients were eligible for the study, of whom 151 were randomised to the intervention group and 149 to the non‐intervention group

Interventions

Guiding antibiotic decisions in hospitalised patients with respiratory infections

Algorithm used in this study: In the intervention group serum PCT and viral/atypical pathogen PCR testing were performed on admission (in addition to all standard‐of‐care diagnostic tests). Antibiotic decisions were made based on a PCT algorithm (for PCT values of ≤ 0.1 ng/mL, initiation of antibiotic treatment is strongly discouraged; for values of 0.11 to 0.24 ng/mL, initiation is discouraged; for values of 0.25 to 0.49 ng/mL, initiation is encouraged; and for values of ≥ 0.5 ng/ mL, initiation is strongly encouraged). In the standard care group standard‐of‐care testing (bacterial and viral cultures of respiratory samples, hospital influenza/RSV duplex PCR, and urine Legionella antigen analysis) were obtained on admission and antibiotic decisions were made by the attending physician. Participants in the standard care group had PCT and viral testing samples frozen and tested at study termination.

Outcomes

• duration of antibiotic therapy

• length of hospital stay

• respiratory complications

• ICU care

• death

• healthcare utilisation

Notes

Funding: This work was supported by Rochester General Hospital (KIDD Fund); the National Institutes of Health, National Institute of Allergy and Infectious Diseases (contract HHSN27220120005C); and BioFire (FilmArray respiratory panel instrument).
Follow up time: Study personnel reviewed the EMR daily until hospital discharge. Participants were contacted by phone at 30 days and 3 months by personnel blinded to randomisation, who collected information about healthcare utilisation, antibiotic use and complications, and return to baseline health.
Registration: NCT01907659

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Participants were stratified by the presence of COPD and were randomly assigned at a ratio of 1:1, using blocks of 4, to receive standard care or the intervention.

Allocation concealment (selection bias)

Low risk

Small block size (4) for randomisation

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Blinding was not feasible due to the study design.

Although the control arm remained blinded, there was a risk for spillover care resulting from providers caring for participants in both arms.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Participants in the standard care group had blood samples frozen and tested at study termination, making the data only available at the end of the study. Phone follow‐up by personnel blinded to randomisation

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Over 3 months of follow‐up 36 participants were lost to follow‐up, 13 died, and 4 withdrew their consent (237/300 completed 3‐month follow‐up).

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol. Trial registered (NCT01907659).

Other bias

High risk

Low overall adherence to the PCT algorithm (64%)
Small sample size

Methods

Randomised clinical trial, multicentre in 53 primary care practices in northwest Switzerland

Participants

Inclusion criteria: People with upper or lower ARIs in primary care and the physician's intention to prescribe antibiotics on the basis of evidence‐based guidelines

Exclusion criteria: Antibiotic use within the previous 28 days, psychiatric disorders or inability to give written informed consent, not being available for follow‐up, not fluent in German, severe immunosuppression, cystic fibrosis, active tuberculosis, and the need for immediate hospitalisation

Included in this analysis: 458 out of 458 randomised participants

Interventions

Guiding antibiotic decisions in primary care with repeated measurements

Algorithm used in this study: In participants with PCT levels lower than 0.1 μg/L, a bacterial infection was considered highly unlikely and the use of ABs was discouraged. In participants with a PCT level higher than 0.25 μg/L, a bacterial infection was considered likely and the use of ABs was recommended. For PCT concentrations of 0.1 to 0.25 μg/L, a bacterial infection was considered unlikely and the use of ABs was not recommended. When ABs were withheld from participants, a second measurement of the PCT level was mandatory within 6 to 24 hours for safety reasons. The use of ABs was recommended if this second measurement was higher than 0.25 μg/L or if the PCT level had increased from the first measurement by more than 50% and the participant showed no clinical improvement. All participants given ABs based on PCT level were reassessed after 3 days. Discontinuation of AB treatment was then recommended in participants with a PCT level of 0.25 μg/L or lower.

Outcomes

• number of days, within the first 14 days after baseline, during which a participant’s daily activities (work or recreation) were restricted by a respiratory tract infection

• degree of discomfort from infection (scored on a scale from 0 (no discomfort) to 10 (a great deal of discomfort)) at 14 days

• days of work missed within 14 days

• days with adverse effects from medication (abdominal pain, diarrhoea, vomiting, skin rash) within 14 days

• antibiotic use

• participants with any symptoms of ongoing or relapsing infection at 28 days

• all‐cause mortality

• hospitalisation

Notes

Funding: Swiss National Science Foundation (Grant 3300C0‐107772) and Association for the Promotion of Science and Postgraduate Training of the University Hospital Basel, Switzerland. B·R·A·H·M·S, Germany provided assay and kit material related to the study.

Follow‐up: Fixed period of 28 days

Registration: ISRCTN73182671

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Independent statistician generated the randomisation sequence.

Allocation concealment (selection bias)

Low risk

Centralised randomisation communicated by phone to physician.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Blinded medical students performed interviews with participants at 14 and 28 days.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 454/458 (99%)

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol.

Other bias

Low risk

85% adherence to PCT algorithm in PCT group

Methods

Randomised clinical trial, multicentre in 15 primary care practices in the Hanover, Germany area

Participants

Inclusion criteria: Adults with upper or lower ARIs in primary care

Exclusion criteria: Treatment with antibiotics during the previous 2 weeks, chronic liver disease, major surgery that had required hospitalisation during the last 4 weeks, autoimmune or systemic disorders, dialysis, medullary C‐cell carcinoma and other inflammatory diseases

Included in this analysis: 550 out of 571 randomised participants; 21 post randomisation exclusions (2 withdrew consent, 1 due to loss of sample, 15 with autoimmune, inflammatory, or systemic disease, 2 with advanced liver disease, 1 with prior use of antibiotics)

Interventions

Guiding antibiotic decisions in primary care with initial measurement only

Algorithm used in this study: PCT value < 0.25 µg/L indicated that a relevant bacterial infection of the respiratory tract is unlikely.

Outcomes

• days with impairment during everyday life or leisure activities, or both, due to the infection of the respiratory tract within the first 14 days according to self assessment

• revisit to the physician's office with a respiratory tract infection within 28 days

• number of days with antibiotic‐induced side effects

• antibiotic use

• change of antibiotics within 28 days

• participants with any symptoms of ongoing or relapsing infection at 28 days

• all‐cause mortality

• hospitalisation

Notes

Funding: B·R·A·H·M·S AG, Germany

Follow‐up: Fixed period of 28 days

Registration: NCT00827060 and NCT00688610

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐generated

Quote: "Baseline adaptive randomisation was realised through a web‐based randomisation data bank (IOMTech GmbH, Berlin, Germany), which had been programmed specifically for that purpose."

Allocation concealment (selection bias)

Low risk

Central randomisation

Quote: "In the central laboratory, the web‐based randomisation of the patient into the PCT group or the control group took place."

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Structured interviews by blinded personnel

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 546/550 (99%)

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol.

Other bias

Low risk

87% adherence to PCT algorithm in PCT group

Methods

Randomised clinical trial, single‐centre, emergency department at the University Hospital Basel, Switzerland

Participants

Inclusion criteria: People with lower ARIs presenting at a medical emergency department

Exclusion criteria: Severely immunocompromised people, i.e. with HIV infection and a CD4 count less than 200 cells per mL, neutropenic patients, and stem cell transplant recipients; those with cystic fibrosis or active tuberculosis; and individuals with nosocomial pneumonia

Included in this analysis: 243 participants out of 243 randomised participants

Interventions

Guiding antibiotic decisions in emergency department patients with different ARIs with initial PCT values only

Algorithm used in this study: A PCT value of 0.1 to 0.25 µg/L was regarded as an indication that bacterial infection was unlikely and use of ABs was discouraged. A serum PCT between 0.25 and 0.5 g/L was deemed indicative of a possible bacterial infection, and the treating doctor was advised to initiate antimicrobial treatment. A PCT value of 0.5 µg/L or greater was judged suggestive of the presence of bacterial infection and AB treatment was strongly recommended. For participants on antimicrobial therapy at the time of admission, discontinuation of ABs was recommended if PCT concentrations were less than 0.25 µg/L.

Outcomes

• antibiotic use

• all‐cause mortality

• ICU admission

• frequency and length of hospital admission

• quality of life

• rate of re‐exacerbation in COPD patients

Notes

Funding: Freiwillige Akademische Gesellschaft Basel, Switzerland; Department of Internal Medicine and the Divisions of Endocrinology and Pneumology, University Hospital Basel; B·R·A·H·M·S AG, Germany and Orgenium Laboratories, Finland, provided assay material and partial support of the investigator‐initiated study.

Follow‐up: Fixed period of 10 to 14 days; in participants with acute exacerbations of COPD the follow‐up period comprised 4 to 6 months

Registration: None

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Quote: "We randomly assigned eligible patients either standard antimicrobial therapy (standard group) or PCT‐guided antimicrobial treatment (PCT group) according to a computer‐generated week wise‐randomisation scheme."

Allocation concealment (selection bias)

High risk

Recruiting physicians were aware of group allocation based on week‐wise randomisation.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Unblinded investigators

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 230/243 (95%)

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol.

Other bias

Low risk

83% adherence to PCT algorithm in PCT group

Methods

Randomised clinical trial, single‐centre, emergency department at the University Hospital Basel, Switzerland

Participants

Inclusion criteria: CAP with X‐ray confirmation in the emergency department

Exclusion criteria: People with cystic fibrosis or active pulmonary tuberculosis, people with hospital‐acquired pneumonia, and severely immunocompromised individuals

Included in this analysis: 302 out of 302 randomised participants

Interventions

Guiding antibiotic decisions in emergency department patients with CAP with repeated PCT measurements

Algorithm used in this study: a PCT level of less than 0.1 µg/L suggested the absence of bacterial infection, and the initiation or continuation of ABs was strongly discouraged. A PCT level between 0.1 and 0.25 µg/L indicated that bacterial infection was unlikely, and the initiation or continuation of ABs was discouraged. A PCT level from 0.25 to 0.5 µg/L was considered indicative of a possible bacterial infection, and the initiation or continuation of AB therapy was encouraged. A PCT level greater than 0.5 µg/L strongly suggested the presence of bacterial infection, and AB treatment and continuation was strongly encouraged. Re‐evaluation of the clinical status and measurement of serum PCT levels were recommended after 6 to 24 h in all participants from whom ABs were withheld. PCT levels were reassessed after 4, 6, and 8 d. Antibiotics were discontinued on the basis of the PCT cut‐offs defined above. In participants with very high PCT values on admission (e.g. greater than 10 µg/L), discontinuation of ABs was encouraged if levels decreased to less than 10% of the initial value (e.g. 1 µg/L, instead of less than 0.25 µg/L).

Outcomes

• antibiotic use

• mortality

• ICU admission

• hospital readmission

• complications due to CAP

• cure defined as resolution of clinical, laboratory, and radiographic signs of CAP

• improvement was defined as reduction of clinical signs and symptoms, improvement of laboratory findings, and reduction of the number or intensity of radiographic signs of CAP

• treatment success represented the sum of the rates for cure and improvement. Treatment failure included death, recurrence, relapse, or persistence of clinical, laboratory, and radiologic signs of CAP and participants lost to follow‐up

Notes

Funding: Funding obtained from B·R·A·H·M·S (Hennigsdorf, Germany), Pfizer (Schweiz AG), and Mepha (Schweiz AG) was used for assay material and salaries of technical personnel involved in laboratory work and for shipping and handling of data and specimens and presentation of data at scientific meetings. Additional support, which provided more than two‐thirds of the total study costs, was granted by funds from the Departments of Internal Medicine and Emergency Medicine, the Stiftung Forschung Infektionskrankheiten (SFI), and mainly from the Departments of Endocrinology and Pulmonary Medicine, University Hospital Basel, Switzerland.

Follow‐up: Fixed period of 6 weeks

Registration: ISRCTN04176397

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Independent statistician created randomisation list.

Allocation concealment (selection bias)

Unclear risk

Quote: "On admission, patients were randomly assigned to one of the two groups by sealed, opaque envelopes." Envelopes were not numbered.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Non‐blinded study members

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 300/302 (99%)

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol.

Other bias

Low risk

87% adherence to PCT algorithm in PCT group

Methods

Randomised, single‐centre clinical trial in an ED of a university hospital in Denmark

Participants

Inclusion criteria: 1) 18 years old and 2) admitted with an AECOPD (clinician’s diagnosis at admission), defined according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD)
Exclusion criteria: 1) person unable to understand or respond to oral or written information; 2) previously been enrolled in the study; and 3) do‐not‐resuscitate order in place
Included in this analysis: 120/630 screened people with AECOPD were randomised and used for the ITT analysis (62 in the PCT group, 58 in the control group).

Interventions

The aim was to assess whether PCT‐guided antibiotic treatment could reduce the overall use of antibiotics among people hospitalised for AECOPD.
Algorithm used in this study: In the control group, antibiotic therapy followed treatment strategies for AECOPD according to GOLD guidelines. In the PCT group, initiation or continuation of antibiotics was strongly discouraged if PCT was 0.15 ng/mL or lower and discouraged if levels were between 0.15 ng/mL and 0.25 ng/mL. Initiation or continuation of antibiotics was encouraged if PCT was > 0.25 ng/mL. In participants with PCT over 5 ng/mL on admission, the algorithm recommended stopping antibiotics when PCT levels decreased by 80% of the peak value.

Outcomes

• fraction of participants using antibiotics for at least 5 days within 28 days after inclusion

• cumulative number of days with any antibiotic therapy within 28 days

• fraction of participants using antibiotics for a) at least 1 day, b) at least 3 days, c) at least 7 days

• length of hospital stay

• adverse events (composite endpoint of mortality, readmission, ICU admission all within 28 days

Notes

Funding: Thermo Fisher Scientific, MA, USA, and bioMérieux Denmark ApS supported the study non‐financially.
Follow‐up: 28 days
Registration: NCT01950936

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Participants were allocated according to the random part of the civil registration number in Denmark

Allocation concealment (selection bias)

Low risk

The randomisation algorithm was concealed to treating clinicians and participants

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Blinding was not feasible, but PCT was only measured in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No mentioning of blinding

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 120/120 (100%)

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol. Trial registered (NCT01950936)

Other bias

High risk

Moderate adherence to the PCT algorithm in PCT group (61.1%)

Methods

Randomised, multicentre, controlled, open‐label intervention trial in 15 hospitals in the Netherlands

Participants

Inclusion criteria: Eligible patients had to be at least 18 years of age, be admitted to the ICU, and have received their first dose of antibiotics no longer than 24 h before inclusion to the trial for an assumed or proven infection.
Exclusion criteria: Patients were excluded in cases of systemic antibiotics as prophylaxis only, antibiotics solely as part of selective decontamination of the digestive tract, prolonged therapy (e.g. endocarditis), expected ICU stay of less than 24 h, severe immunosuppression, severe infections (due to viruses, parasites, or Mycobacterium tuberculosis), and moribund patients. Patients who received corticosteroids were not excluded.
Included in this study: 1575/4507 screened patients were enrolled in the study (776 in the PCT group, 799 in the control group) for an intention‐to‐treat analysis.

Interventions

Guiding antibiotic decisions in critically ill ICU patients

Algorithm used in this study: Antibiotics in the standard‐of care group were stopped according to local or national guidelines and according to the discretion of attending physicians. Procalcitonin concentration was not measured in the standard‐of‐care group.
For participants randomly assigned to the PCT‐guided group, once‐a‐day measurements of PCT concentrations were taken and made available to the attending physicians, including a baseline measurement as close to initiation of antibiotics as possible, at least within 24 h. The study protocol advised stopping the prescribed antibiotics if PCT concentration had decreased by 80% or more of its peak value (relative stopping threshold) or when it reached a value of 0.5 μg/L or lower (absolute stopping threshold). The attending physician was free to decide whether to continue antibiotic treatment in participants who had reached these thresholds.

Outcomes

• consumption of antibiotics (expressed as defined daily doses)

• duration of antibiotic treatment (defined as the number of 24‐hour periods between start and end of antibiotic treatment)

• percentage of participants who had a recurrent infection

• length of stay in hospital and ICU

• costs of antibiotics

• costs of PCT tests

Notes

Funding: Thermo Fisher Scientific
Follow‐up: 1‐year follow‐up
Trial Registration: NCT01139489

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Randomisation was done centrally by use of a computer‐generated list produced by an independent research organisation (the Julius Centre for Human Research, Utrecht, Netherlands).

Allocation concealment (selection bias)

Low risk

Centralised randomisation by an independent organisation

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Participants and investigators were aware of treatment assignment.

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment mentioned in the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

1546 of 1575 participants included in the intention‐to‐treat population analysis

Selective reporting (reporting bias)

Low risk

The outcomes correspond to the trial registration (NCT01139489).

Other bias

High risk

Physicians did not adhere to the stopping advice in more than half of the participants.
About 30% of participants randomly assigned to the PCT group were discharged from ICU before the algorithm recommended stopping antibiotic treatment.

Methods

Randomised, single‐centre, controlled clinical trial at an ICU of a tertiary care, private hospital in São Paulo, Brazil

Participants

Inclusion criteria: People with microbiologically confirmed infections (blood, urine, tracheal aspirate, or bronchoalveolar lavage fluid cultures) with sepsis, severe sepsis, and septic shock
Exclusion criteria were: Onset of antibiotic therapy more than 48 hours before the date when the cultures were performed; people under 18 years old; known pregnancy; infections requiring prolonged antibiotic therapy, such as bacterial endocarditis, hepatic or brain abscess, deep abscess, mediastinitis, and osteomyelitis; severe infection caused by viruses, parasites, fungi, or mycobacteria; chronic localised infections, such as chronic osteomyelitis or chronic prostatitis; people without indication for ICU admission, as determined by the attending physician; and negative cultures (blood, urine, tracheal aspirate, or bronchoalveolar lavage fluid) in people with suspected sepsis, severe sepsis, or septic shock
Included in this study: 81/265 eligible patients randomised for a intention‐to‐treat analysis; after further exclusions, 51/265 patients remained for per‐protocol analysis.

Interventions

Guiding antibiotic decisions in ICU patients with proven bacterial infection

Algorithm used in this study: All participants received antibiotic therapy. For the control group, stopping antibiotic therapy was at the discretion of the attending physician. For the intervention group, physicians were guided by the PCT protocol to stop antibiotic treatment. The protocol stated 2 conditions: 1) PCT dropped more than 90% from the peak level, or 2) an absolute value < 0.5 ng/mL was reached.

Outcomes

• duration of antibiotic therapy

• in‐hospital mortality

• ICU mortality

• ICU length of stay

• recurrence of the initial infection

• analysis of the CRP levels along with the PCT protocol

• therapy costs

Notes

Funding: No funding declared in the main article.
Follow‐up time: Data were recorded from 2 days before the bacteraemia (when applicable), with day 0 defined as the day sepsis was diagnosed, until 14 days after or at ICU discharge, whichever came first.
Registration: NCT01494675

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

A blind randomisation scheme was used where a black box contained 100 folders, and 2 authors randomly drew 1 folder as soon as an informed consent was present.

Allocation concealment (selection bias)

Low risk

Folders were randomly and blindly assigned as “PCT group” or “standard group”. 2 of the authors would randomly draw 1 folder from a black box containing 100 folders (50 “PCT group” and 50 “control group”).

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Blinding of the treating physicians was not feasible in this study.

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of the outcome assessment mentioned in the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

In the intention‐to‐treat analysis the follow‐up for mortality was 81/81 (100%).

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol. Trial registered (NCT01494675).

Other bias

High risk

Low adherence to the PCT algorithm (47.6%)
Single‐centre study where not all attending physicians agreed to participate (also a reason for exclusions in the PCT arm in the per‐protocol analysis and low adherence)

Methods

Randomised, single‐centre, open‐label, controlled clinical trial in HeNan Hospital, China

Participants

Inclusion criteria: All patients with suspected AE‐IPF admitted to the respiratory department were assessed for eligibility from January 2009 to December 2011. Acute exacerbation of idiopathic pulmonary fibrosis was defined according to the criteria established by the Idiopathic Pulmonary Fibrosis Clinical Research Network: (1) previous or concurrent diagnosis of idiopathic pulmonary fibrosis; (2) unexplained worsening or development of dyspnoea within 30 days; (3) high‐resolution computed tomography with new bilateral ground‐glass; (4) abnormality and/or consolidation superimposed on a background reticular or honeycomb pattern consistent with usual interstitial pneumonia pattern; (5) no evidence of pulmonary infection by endotracheal aspirate or bronchoalveolar
lavage; (6) exclusion of left heart failure, pulmonary embolism, and identifiable cause of acute lung injury.
Exclusion criteria: Patients treated with antibiotics during the previous 2 weeks were excluded.
Included in this study: 68 of 78 randomised participants finished at follow‐up.

Interventions

Guiding antibiotic decisions in people with acute exacerbation of idiopathic pulmonary fibrosis

Algorithm used in this study: Serum PCT level was measured every 3 days. The first PCT measurement was available before the clinical decision to start antibiotics treatment. Participants whose serum PCT value exceeded the threshold of 0.25 ng/mL were administered antibiotics and were treated until PCT value fell to ≤ 0.25 ng/mL. In the routine treatment group, the decision to administer antibiotics was guided by the clinical experience of the clinician, typically by conventional laboratory tests such as sputum bacteriology and white blood cell count.

Outcomes

• length of hospitalisation

• the numbers of participants exposed to antibiotics treatment

• duration of antibiotic treatment

• cases of mechanical ventilation

Notes

Funding: No funding declared in the main paper.
Follow‐up: 30 days
Trial registration: No trial registration found for this clinical trial.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐generated randomisation

Allocation concealment (selection bias)

Low risk

Participants were randomly assigned to either PCT‐guided antibiotic treatment or a control group receiving routine antibiotic therapy by the statistician using computer‐generated random numbers.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

No blinding was performed. "In the routine treatment group, patients were treated by antibiotics according to the clinical experience of clinicians typically guided by conventional laboratory tests, such as sputum bacteriology and white blood cell count."

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment was mentioned in this study.

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

Postrandomisation exclusion rate was relatively high but similar in both study arms (6 in the intervention arm versus 4 in the control group).

Selective reporting (reporting bias)

Unclear risk

No trial registration was found for this trial.

Other bias

Low risk

100% adherence to PCT protocol in the per‐protocol analysed participants (all protocol violations were defined as "withdrawn").

Methods

Randomised clinical trial, single‐centre, ICU in Germany

Participants

Inclusion criteria: Patients in the surgical ICU with suspected bacterial infections and > 1 SIRS criteria

Exclusion criteria: Patients who refused study consent, whose antibiotic treatment had been initiated before intensive care admission, or who had therapy limitations

Included in this analysis: 43 (110); 67 not considered for this analysis due to diagnosis other than ARI

Interventions

Guiding antibiotic decisions in postoperative patients in a surgical ICU

Algorithm used in this study: Antibiotic therapy in the PCT‐guided group was discontinued if clinical signs and symptoms of infection improved and PCT decreased to less than 1 µg/L, or if the PCT value was more than 1 µg/L, but had dropped to 25% to 35% of the initial value over 3 days.

Outcomes

• antibiotic use

• mortality (ICU‐free days alive)

Notes

Funding: B·R·A·H·M·S AG

Follow‐up: Until hospital discharge

Registration: ISRCTN10288268

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

High risk

Unconcealed drawing of lots

Allocation concealment (selection bias)

High risk

Unconcealed drawing of lots

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Non‐blinded study members

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 393/394 (100%)

Selective reporting (reporting bias)

Low risk

No selective reporting (oral verification with first author)

Other bias

Unclear risk

Adherence to PCT protocol not reported/assessed

Methods

Randomised clinical trial, multicentre, 3 hospitals in Denmark

Participants

Inclusion criteria: Hospitalised patients with suspected pneumonia (no X‐ray confirmation); quote: "The assessment of eligibility (i.e. the clinical diagnosis) was made by the admitting physician and was based on medical history and physical examination."

Exclusion criteria: Not meeting the diagnostic criteria

Included in this analysis: 210 out of 223 randomised participants; 13 post randomisation exclusions (3 no PCT testing, 6 not meeting inclusion criteria, 4 withdrew informed consent)

Interventions

Guiding antibiotic decisions in CAP patients with initial values only

Algorithm used in this study: Physicians were not asked to wait for PCT results before initiating antimicrobial therapy, therefore PCT values were, in most cases, used to motivate either cessation or continuation of already initiated treatments. Discontinuation of AB treatment was recommended if PCT at admission was below 0.25 µg/L, despite delays in test results.

Outcomes

• antibiotic use

• mortality

• ICU admission

Notes

Funding: The Danish Medical Research Council and the Danish Lung Association provided financial support.

Follow‐up: Until hospital discharge

Registration: NCT00415753

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐generated randomisation scheme

Allocation concealment (selection bias)

Low risk

Central randomisation

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Non‐blinded study members

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 210/210 (100% until discharge)

Selective reporting (reporting bias)

Low risk

No selective reporting (oral verification with first author)

Other bias

High risk

59% adherence to PCT algorithm in PCT group

Methods

Randomised, single‐centre, prospective, controlled clinical study in 5 ICUs in Belgium

Participants

Inclusion criteria: Patients older than 18 yrs of age and hospitalised for > 2 days in 1 of the 5 ICUs
Included in this study: Of 509 randomised participants, 441 developed infection, and PCT was obtained and analysed in 389 of these participants.

Interventions

Guiding antibiotic decisions in ICU patients

Algorithm used in this study: According to the proposal by Mueller and colleagues, for participants in the PCT group, the use of antibiotics was more or less strongly discouraged if PCT level was < 0.25 μg/L or 0.50 μg/L, respectively, and more or less recommended if PCT level was above 1 μg/L or 0.50 μg/L, respectively. This strategy was applied to all infectious episodes encountered during participants' ICU stay.

Outcomes

• difference of antibiotic consumption between the PCT group and the control group

• usefulness of PCT levels in the ICU diagnostic algorithm in deciding whether or not to initiate antibiotics whenever infection was suspected and determination of concordance of the infection’s diagnostic ratings by the ICU physician and the infectious disease specialist, bearing in mind that the latter was blinded to PCT results in all of the cases

Notes

Funding: No information provided.
Follow‐up: During ICU stay
Trial registration: No trial registration found.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information on method of randomisation provided.

Allocation concealment (selection bias)

Unclear risk

No information on allocation concealment provided.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Procalcitonin levels in the control arm were blinded for treating physicians, but the study arm to which participants had been assigned was not blinded.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

At the end of ICU stay, participants’ charts were reviewed by infectious disease specialists blinded to PCT results, who classified them as confirmed, probable, possible, or no infection using all the clinical data and biological results including microbiological cultures and results from investigational procedures.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality 509/509 (100%)

Selective reporting (reporting bias)

Unclear risk

No trial registration found for this research article.

Other bias

High risk

Low adherence to PCT algorithm (46.3%)

Methods

Randomised, single‐centre, prospective, controlled clinical trial in Minas Gerais, Brazil

Participants

Inclusion criteria: Presence of febrile neutropenia (axillary temperature ˜37.8°C, neutrophils count < 500 cells/mm³) in people with diagnosis of haematological disease (except people undergoing allogeneic stem cell transplantation); with expected duration of neutropenia of more than 3 days; ongoing broad‐spectrum antibiotic therapy according to institutional guideline, based on Infectious Diseases Society of America; and no current use of therapeutic antibiotics or antifungals with the last day of use was > 14 days before inclusion in the study
Exclusion criteria: Severe organ dysfunction (e.g. hypotension, ICU admission, trans‐retinoic acid syndrome, respiratory insufficiency, and disseminated intravascular coagulation); previous proven or probable invasive fungal infection according to the Springer Ann Hematol (2016) European Organisation for Research and Treatment of Cancer‐Mycosis Study Group (EORTC‐MSG) criteria; infections due to Pseudomonas spp, Acinetobacter spp, Staphylococcus aureus, Mycobacterium tuberculosis, Pneumocystis jirovecii, Toxoplasma gondii, or HIV; infections requiring antibodies for a long time (e.g. infectious endocarditis, osteomyelitis); grade 3 or 4 oral mucositis, since this condition increases the risk of S aureus bacteraemia; and pregnancy
Included in this study: 62 randomised participants, 1 post randomisation exclusion due to withdrawal of informed consent

Interventions

Guiding antibiotic decisions in people with febrile neutropenia

Algorithm used in this study: Attending physicians were encouraged to discontinue antibiotics in participants when both of the following criteria were met: (i) no febrile episodes for 2 consecutive days (if no Ionger neutropenia) or 3 consecutive days (if still neutropenia), and (ii) PCT concentration at least 90% lower than highest baseline Ievels or lower than 0.5 ng/mL for 2 consecutive days, regardless of the initial Ievels. The final decision to discontinue antibiotics was left at the attending physician's discretion. Procalcitonin Ievels were measured for 2 additional days following antibiotics interruption to monitor a possible relapse of infection. For safety reasons, at least 5 days of antibiotic therapy were ensured for all included participants. Participants with bacteraemia were treated for at least 7 days. In the control group, duration of antibiotic therapy was based on institutional protocol, according to Infectious Diseases Society of America recommendations.

Outcomes

• antibiotic exposure, measured by the duration of the first course of antibiotic therapy (in days) and days without antibiotics during follow‐up (28 days)

• clinical cure rate

• infection relapse rate (infection diagnosed 48 h or more after antibiotic discontinuation)

• superinfection rate (defined as occurrence of infection due to 1 or more different pathogens, in the same or in another site, during the first antibiotic therapy)

• length of hospital stay from inclusion (in days)

• all‐cause 28‐ and 90‐day mortality

Notes

Funding: This work was supported by Funda of the de Amparo a Pesquisa do Eatado dc MiDaa Ocrais (.APQ‐01956‐10).
Follow‐up: 3 months
Trial registration: NCT00928291

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Participants were randomised on the third day of follow‐up, in a 1:1 ratio for the PCT group and the control group.

Allocation concealment (selection bias)

Low risk

Randomisation was performed using a table of random, computer‐generated numbers, and sealed, opaque envelopes were used.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

For participants randomised into the control group, results of PCT serum Ievels were kept concealed during the study and were only revealed for the final analysis. However, physicians were aware of the study group.

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment was mentioned in the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Most randomised participants finished follow‐up. Only 1 post randomisation exclusion

Selective reporting (reporting bias)

Low risk

Outcomes correspond to trial registration.

Other bias

Low risk

High adherence to the PCT protocol (73.3%)

Methods

Randomised clinical trial, single‐centre, emergency department outpatients in China

Participants

Inclusion criteria: CAP with X‐ray confirmation

Exclusion criteria: Use of antibiotic therapy in 2 weeks before enrolment, systemic immune deficiency, organ dysfunction, tumour, mental illness, CAP onset ≥ 5 days, coexisting extrapulmonary infection requiring antibiotic therapy

Included in this analysis: 127 out of 127 randomised participants

Interventions

Guiding antibiotic decisions in CAP patients with repeated levels

Algorithm used in this study: A PCT level of less than 0.1 µg/L suggested the absence of bacterial infection, and the initiation or continuation of ABs was strongly discouraged. A PCT level between 0.1 and 0.25 µg/L indicated that bacterial infection was unlikely, and the initiation or continuation of ABs was discouraged. A PCT level of 0.25 µg/L or greater was considered indicative of a possible bacterial infection, and the initiation or continuation of AB therapy was encouraged. Re‐evaluation of the clinical status and measurement of PCT levels was recommended after 6 to 12 h in all participants from whom ABs were withheld.

Outcomes

• antibiotic use

• mortality

• ICU admission

• treatment success represented the sum of the rates for cure and improvement. Cure was defined as resolution of clinical, laboratory, and radiographic signs of CAP. Improvement was defined as reduction of clinical signs and symptoms, improvement of laboratory findings, and reduction of the number or intensity of radiographic signs of CAP.

• treatment failure included death, recurrence, relapse, or persistence of clinical, laboratory, and radiologic signs of CAP and participants lost to follow‐up

Notes

Funding: Training fund of Shanghai No. 5 Hospital

Follow‐up: 28 days

Registration: None

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

High risk

Odd and even patient ID numbers

Allocation concealment (selection bias)

High risk

Odd and even patient ID numbers

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Non‐blinded study members

Incomplete outcome data (attrition bias)
All outcomes

Low risk

210/210 (100% until discharge)

Selective reporting (reporting bias)

Low risk

No selective reporting (oral verification with first author)

Other bias

Unclear risk

Adherence to PCT protocol not reported/assessed.

Methods

Randomised clinical trial, single‐centre, emergency department outpatients in China

Participants

Inclusion criteria: CAP with X‐ray confirmation in an outpatient setting

Exclusion criteria: Pregnancy, commencement of antibiotic therapy ≥ 48 h before enrolment, systemic immune deficiency, withholding of life‐support, and active tuberculosis

Included in this analysis: 156 out of 172 randomised participants; 16 post randomisation exclusions (6 lost to follow‐up, 7 withdrew consent, 3 with final diagnosis other than CAP)

Interventions

Guiding antibiotic decisions in CAP patients with repeated levels

Algorithm used in this study: A PCT level of less than 0.1 µg/L suggested the absence of bacterial infection, and the initiation or continuation of ABs was strongly discouraged. A PCT level between 0.1 and 0.25 µg/L indicated that bacterial infection was unlikely, and the initiation or continuation of ABs was discouraged. A PCT level of 0.25 µg/L or greater was considered indicative of a possible bacterial infection, and the initiation or continuation of AB therapy was encouraged. Re‐evaluation of the clinical status and measurement of PCT levels was recommended after 6 to 12 h in all participants from whom ABs were withheld.

Outcomes

• antibiotic use

• mortality

• ICU admission

Notes

Funding: The study was sponsored by a grant from the Shanghai Fifth People's Hospital Science Foundation (09YRCPY11).

Follow‐up: Fixed period of 4 weeks

Registration: NA

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

High risk

Odd and even patient ID numbers

Allocation concealment (selection bias)

High risk

Odd and even patient ID numbers

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Non‐blinded study members

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 156 (156) (100%)

Selective reporting (reporting bias)

Low risk

No selective reporting (oral verification with first author)

Other bias

Unclear risk

Adherence to PCT protocol not reported/assessed.

Methods

Randomised, single‐centre, open‐label, controlled clinical trial in Shanghai, China

Participants

Inclusion criteria: People aged 18 to 65 years with severe acute exacerbations of asthma. A severe asthma exacerbation was defined as at least 1 of the following: need for systemic corticosteroids, or an increase from a stable maintenance dose, for at least 3 days and/or hospitalisation or ED visit because of asthma requiring systemic corticosteroids.
Exclusion criteria: People with antibiotic use within the previous 14 days, psychiatric disorders or other inability to give written informed consent, not being available for follow‐up, severe immunosuppression, heart failure, cystic fibrosis, active tuberculosis, pregnancy, and chest radiography–confirmed pneumonia
Included in this study: 180/216 screened individuals were eligible for the study (90 intervention group, 90 non‐intervention group); 169 finished the follow‐up.

Interventions

Guiding antibiotic decisions in people with acute severe exacerbation of asthma

Algorithm used in this study: Antibiotic treatment was strongly discouraged when serum PCT level was less than 0.1 μg/L; antibiotic treatment was discouraged when serum PCT level was less than 0.25 μg/L; and antibiotic treatment was encouraged when serum PCT level was higher than 0.25 μg/L. When antibiotics were withheld from participants, a second measurement of the PCT level was mandatory within 6 to 24 hours for safety reasons. The use of antibiotics was recommended if this second measurement was higher than 0.25 μg/L. Physicians were permitted to overrule the algorithm, but they had to indicate the reasons for overruling. The control group received antibiotic according to the discretion of the treating physician, who was unaware of the participant’s PCT levels.

Outcomes

• antibiotic use, expressed as rate of antibiotic prescriptions in percentage and relative risk of antibiotic exposure

• measures of treatment success

• length of hospital stay

• clinical, laboratory, and spirometry outcomes at discharge

• results of spirometry at the 12‐month follow‐up examination, as well as the results of the Asthma Control Test, the results of the Asthma Quality of Life Questionnaire at the 12‐month follow‐up visit

• clinical events during the 12‐month follow‐up period, including numbers of asthma exacerbations, ED visits, hospitalisations, and need for systemic corticosteroid use for treatment of asthma

Notes

Funding: The study was sponsored by a grant from the Shanghai Fifth People’s Hospital Science Foundation and Minhang District Natural Science Foundation of Shanghai. The funding bodies had no involvement in the design, collection, analysis, or interpretation of data; in the writing of the manuscript; or in the decision to submit the manuscript for publication.
Follow‐up time: 12 months
Registration: ChiCTR‐TRC‐12002534

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Allocation to either intervention was conducted according to computer‐generated random numbers produced by an independent statistician.

Allocation concealment (selection bias)

Low risk

After randomisation, an opaque, sealed, sequentially numbered envelope containing the PCT or control protocol was prepared for each participant according to group assignment.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label study with blinding of PCT level in the control group

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment mentioned in the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

169/180 participants completed 1‐year follow‐up visit (11 participants lost to follow‐up).

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol. Trial registered (ChiCTR‐TRC‐12002534).

Other bias

Low risk

High adherence to PCT algorithm (93.3%)

Methods

Randomised, single‐centre, open‐label, controlled clinical trial at a 200‐bed academic tertiary care hospital in Belgrade, Serbia

Participants

Inclusion criteria: People scheduled to undergo open heart surgery on cardio‐pulmonal bypass. We assessed people who were selected for elective cardiac surgery at the 200‐bed academic tertiary care hospital. The criterion for inclusion in the study was the type of operation: coronary artery bypass grafting (CABG) surgery, valve reconstruction, combined CABG and valve procedures. Entry criteria included stable and unstable angina pectoris, valve insufficiency, left ventricle ejection fraction (LVEF) above 30%, and epidemiological status with saprophyte bacteria.
Exclusion criteria: People selected for redo cardiac operations, thoracic aortic surgery, as well as people having active endocarditis and people with LVEF < 30%. People with preoperative signs of infection (leukocyte count > 12,000/L; body temperature > 38°C) were also excluded.
Included in this study: 205/205 included participants finished for follow‐up (102 PCT group/103 standard group).

Interventions

Guiding antibiotic decisions in patients after cardiac surgery

Algorithm used in this study: Antibiotic prophylaxis was performed in all participants. The participants were divided at the time of surgery into the standard group and the PCT group.
In the standard group, the antibiotic use was applied according to the criteria based on the laboratory and clinical signs; no antibiotic therapy was administrated routinely in the absence of clinical signs of infection or a bacteriologic positive sample.
In the PCT group, the use of antibiotics was encouraged or discouraged on the basis of serum PCT concentrations. A serum PCT concentration of 0.5 ng/mL or less indicated the absence of bacterial infection, at which point the use of antibiotics was discouraged.

Outcomes

• proportion of participants treated with antibiotics

• overall cost of antibiotics per 1 patient

• total cost of antibiotics per 1 hospital day after operation

• ICU stays and hospital stay

• rehospitalisation

• incidence of infections

• severe non‐infection complications

• mortality rate

Notes

Funding: No funding is mentioned in the main article.
Follow‐up: 1 year
Trial registration: No trial registration found.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐generated randomisation scheme

Allocation concealment (selection bias)

Low risk

The participants were divided at the time of surgery into the standard group and the PCT group by centralised randomisation.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

No blinding of physicians due to the study design

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment mentioned in the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 205/205 (100%)

Selective reporting (reporting bias)

Unclear risk

No trial registration found.

Other bias

Unclear risk

No information about adherence

Methods

Randomised, single‐centre, single‐blinded clinical trial in a 30‐bed ICU in Tehran, Iran

Participants

Inclusion criteria: Patients with at least 2 of 4 criteria including body temperature above 38°C or below 36°C; tachycardia > 90/min; tachypnoea > 20/min; and leukocytosis > 12 × 10⁹/L or a leftward shift with more than 10% band cells or leukopenia < 4 × 10⁹/L were defined as patient with SIRS.
Exclusion criteria: Documented infection, pus from wound or abscess, empyema, thrombophlebitis, infection due to viral or parasites, hypoxaemia (PO₂ < 60 mmHg), oliguria (urine output < 30 mL/h), Glasgow Coma Scale 3 without sedation, parenteral antibiotic usage 24 hours before admission to ICU, hospitalisation 48 hours before enrolment, conditions requiring prolonged antibiotic therapy such as endocarditis, chronic localised infection such as osteomyelitis, and severely immunocompromised patients
Included in this study: 60 participants were randomised (30 in the intervention group and 30 in the control group).

Interventions

Guiding antibiotic decisions in critically ill ICU patients

Algorithm used in this study: In case group, according to serum level of PCT, participants were divided into 3 groups as: PCT level 0.5 ng/mL or less (group A), PCT value of 0.5 to 2 ng/mL (group B), and PCT level 2 ng/mL or more (group C). Group A indicated a low probability of bacterial infection; use of antibiotics was discouraged, and PCT level was rechecked after 12 hours. In group B, with a medium probability of infection, antibiotic therapy was not administered, and PCT level was rechecked after 8 hours. In group C, with a high probability of bacterial infection, participants underwent antibiotic treatment. If the PCT level was higher than 2 ng/mL after recheck in group A and B, antibiotics therapy was administered; if the PCT level was lower than 2 ng/mL, participants underwent close observation, and PCT was rechecked until culture results were obtained.

Outcomes

• use of antibiotic treatment

• change in clinical status

• early mortality

Notes

Funding: No funding declared in the original research article.
Follow‐up: No precise follow‐up found in the original research article.
Trial registration: No trial registration found in the original research article.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

All participants were randomly divided into 2 groups by computer‐based random number generation.

Allocation concealment (selection bias)

Low risk

All participants were randomly divided into 2 groups by computer‐based random number generation.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Procalcitonin was only measured in the intervention arm, but blinding was not feasible due to the study design.

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment was mentioned in the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

All randomised participants finished the study and were assessed for the primary outcome.

Selective reporting (reporting bias)

Unclear risk

No trial registration was found for this study.

Other bias

Unclear risk

Adherence to the PCT algorithm was not known.

Methods

Randomised clinical trial, single‐centre, medical ICU in Switzerland

Participants

Inclusion criteria: Suspected severe sepsis or septic shock in the ICU

Exclusion criteria:

• microbiologically documented infections caused by Pseudomonas aeruginosa,Acinetobacter baumannii,Listeria spp,Legionella pneumophila,Pneumocystis jirovecii, orMycobacterium tuberculosis, for which a prolonged duration of antibiotic therapy is standard of care

• severe infections due to viruses or parasites (e.g. haemorrhagic fever, malaria)

• infectious conditions requiring prolonged antibiotic therapy (e.g. bacterial endocarditis, brain abscess, deep abscesses)

• antibiotic therapy started 48 hours or more before enrolment

• chronic, localised infections (e.g. chronic osteomyelitis)

• severely immunocompromised patients, such as patients infected with HIV and with a CD4 count < 200 cells/mm³, neutropenic patients (0.500 neutrophils/mm³), or patients on immunosuppressive therapy after solid organ transplantation

• withholding of life‐support

• absence of antimicrobial treatment despite clinical suspicion of sepsis

Included in this analysis: 52 out of 79 randomised participants; 27 not considered for this analysis due to a diagnosis other than RTI

Interventions

Guiding antibiotic decisions in ICU patients with repeated measurements

Algorithm used in this study: Procalcitonin levels measured at baseline and daily. For participants presenting a favourable clinical course, investigators used predefined “stopping rules” based on circulating PCT levels to encourage physicians to discontinue ABs. Participants with baseline PCT level ≥ 1 µg/L were re‐evaluated at day 5. Investigators encouraged treating physicians to discontinue ABs when:

1. PCT dropped more than 90% from the baseline peak level; or

2. an absolute value below 0.25 µg/L was reached.

Participants with PCT level below 1 µg/L at baseline were re‐evaluated at day 3; treating physicians were encouraged to discontinue ABs when PCT level was below 0.1 µg/L and careful clinical evaluation ruled out severe infection.

Outcomes

• all‐cause mortality at day 28

• clinical cure defined as clinical signs and symptoms present at baseline that had resolved by the final clinical assessment

• reoccurrence of the initial infection

• length of ICU stay

Notes

Funding: B·R·A·H·M·S AG

Follow‐up: Fixed follow‐up period of 28 days

Registration: NCT00250666

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐based random number generation

Allocation concealment (selection bias)

Low risk

Sequentially numbered, opaque, sealed envelopes

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Non‐blinded study members

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 52/52 (100%)

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol.

Other bias

Low risk

81% adherence to PCT algorithm in PCT group

Methods

Randomised, single‐centre, prospective, open‐label, non‐inferiority clinical trial in Shizuoka, Japan

Participants

Inclusion criteria: Patients at risk for aspiration, who had been hospitalised after developing pneumonia, were enrolled. Aspiration pneumonia was clinically diagnosed on the basis of the findings on computed tomography (e.g. bronchopneumonia in the dorsal lower lobes), combined with a history of aspiration pneumonia, stroke or dementia, poor systemic condition, or any combination of these (e.g. bedridden patients or patients fed by a nasogastric tube or percutaneous endoscopic gastrostomy). Selection criteria included the following: at least 1 month had elapsed since the last treatment for relapsed pneumonia, and ventilator use was not scheduled for the pneumonia treatment
Exclusion criteria: Patients with a known severe allergy to any drugs; patients with sepsis or a severe infectious disease; patients with severe underlying diseases (e.g. malignancy, COPD, heart failure) that affected the prognosis; and patients who could not safely have cessation of oral intake or hydration as a treatment for aspiration pneumonia because of dementia
Included in this study: The study enrolled 105 participants; 2 participants withdrew their informed consent, 5 were excluded because of other final diagnoses, and 1 was excluded because of a defect in the PCT data. The ITT population thus comprised 96 participants: 48 in the PCT group and 48 in the control group

Interventions

Guiding antibiotic decisions in patients with aspiration pneumonia and assessment of the continuation of oral intake

Algorithm used in this study: Procalcitonin levels were measured via outsourcing to SRL (Tokyo, Japan); the results were obtained 2 or 3 days after admission. In the PCT group, if the PCT levels upon admission were < 0.5 ng/mL, 0.5 to 1.0 ng/mL, or > 1.0 ng/mL, the duration of antibiotic therapy was determined to be 3, 5, or 7 days, respectively. If the PCT level upon admission was 45.0 ng/mL, antibiotic treatment was continued until it was less than 10% of the peak PCT level reached.
In the control group, antibiotic therapy followed the recommendations of the Japanese Respiratory Society guideline for management of community‐acquired pneumonia in adults. Antibiotic therapy was discontinued if 3 of the following 4 criteria were met: fever declined (body temperature < 37.0°C), normalisation of leukocyte count, decrease in the CRP level to 30% of the maximum, and an obvious improvement as observed by chest radiography. In both groups, the choice of antibiotic regimen was left to the discretion of the treating physician.

Outcomes

• primary non‐inferiority endpoint was a composite of a relapse of aspiration pneumonia and death from any cause occurring within 30 days of admission

• antibiotic exposure

• adverse events from antibiotic therapy

Notes

Funding: No funding declared in the main research paper.
Follow‐up: 30 days
Trial registration: UMIN000004800

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Randomly allocated in 1:1 ratio

Allocation concealment (selection bias)

Unclear risk

No clear mention of allocation concealment is made in the main article. Citation "Following enrolment, the patients were randomly allocated in a 1:1 ratio to groups assigned different durations of antibiotic therapy…"

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Procalcitonin was only measured in the intervention arm, but study design was open‐label.

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment mentioned in the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Most participants were followed for 30 days. Reasons for post randomisation exclusions are clearly reported in the article.

Selective reporting (reporting bias)

Low risk

Outcomes correspond to trial registration.

Other bias

Unclear risk

As cited by authors under limitations, time until analysis of PCT concentrations was available was relatively long (3 days).

Adherence to PCT algorithm not known

Methods

Randomised, multicentre, open‐label, controlled clinical trial in the ICUs of 2 university hospitals in Brazil

Participants

Inclusion criteria: All adult patients 18 years of age or older with suspected severe sepsis or septic shock were assessed for potential inclusion.
Exclusion criteria: Confirmed microbiological infection by Pseudomonas aeruginosa,Acinetobacter baumannii, Listeria spp,Mycobacterium tuberculosis, or fungi; Staphylococcus aureus bacteraemia; suspected or confirmed severe infections caused by viruses or parasites; infections that required long‐term treatment, regardless of the aetiologic agent (e.g. bacterial endocarditis); localised chronic infections (e.g. chronic osteomyelitis); > 48 hours of antibiotic treatment; immunosuppressed patients (such as those diagnosed with HIV), patients with neutropenia (< 500 neutrophils/mm³), patients post‐solid organ transplant, patients under immunosuppressive therapy, and patients who received more than 1 mg/kg of prednisone or equivalent; patients under palliative care; patients who suffered multiple trauma, burns, or major surgery in the previous 5 days; patients diagnosed with pulmonary neoplasias, carcinoid tumours, or medullary tumours of the thyroid; and patients who remained in the ICU for 24 hours or less
Included in this study: 94/355 patients assessed for eligibility were included in the final analysis (49 in the PCT group and 45 in the CRP group).

Interventions

Guiding antibiotic decisions with CRP versus PCT in septic patients

Algorithm used in this study: Antibiotic therapy was discontinued following a protocol based on serum levels of these markers, according to the allocation group. For both groups, at least 7 full days of antibiotic therapy were ensured in participants with SOFA score greater than 10 and/or bacteraemia at inclusion, and participants with evident resolution of the infectious process had antibiotics stopped after 7 days, despite biomarker levels.

Outcomes

• duration of antibiotic therapy for the first episode of infection

• total number of days on antibiotic therapy

• days off antibiotic therapy

• death from any cause during the 28 days of follow‐up in the hospital

• length of stay (LOS) in the ICU and LOS in the hospital

• clinical cure, recurrent infection, and nosocomial infection

Notes

Funding: Supported in part by the Minas Gerais Research Foundation (Fundação de Amparo à Pesquisa do Estado de Minas Gerais, FAPEMIG)
Follow‐up: 28 days or until death or hospital discharge
Trial registration: NCT00934011

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Randomisation was performed using a computer‐generated random number table.

Allocation concealment (selection bias)

Low risk

Sealed, opaque envelopes were used for the randomisation.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

No blinding due to the study design

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment mentioned in the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

All participants were included in the ITT analysis.

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol. Trial registered (NCT00934011).

Other bias

Low risk

High adherence to the PCT protocol (87.8%)

Methods

Randomised clinical trial, single‐centre, surgical ICU in Germany

Participants

Inclusion criteria: Patients after abdominal surgery with antibiotic treatment because of severe sepsis in the surgical ICU

Exclusion criteria: Patients were excluded if they did not meet the respective inclusion criteria, refused informed consent, or had already received antibiotic treatment prior to admission to the ICU.

Included in this analysis: 8 out of 27 randomised participants; 19 not considered for this analysis due to diagnosis other than RTI

Interventions

Guiding antibiotic decisions in postoperative patients in a surgical ICU

Algorithm used in this study: In the PCT‐guided group, antibiotic therapy was discontinued if clinical signs and symptoms of sepsis improved and PCT values had either decreased to 1 µg/L or less or had dropped to 25% to 35% of the initial PCT concentration over 3 consecutive days.

Outcomes

• antibiotic use

• mortality (ICU‐free days alive)

Notes

Funding: NA

Follow‐up: Until hospital discharge

Registration: None

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

Unconcealed drawing of lots

Allocation concealment (selection bias)

High risk

Unconcealed drawing of lots

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Non‐blinded study members

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 8/8 (100% until discharge)

Selective reporting (reporting bias)

Low risk

No selective reporting (oral verification with first author)

Other bias

Unclear risk

Adherence to PCT protocol not reported/assessed.

Methods

Randomised clinical trial, multicentre, 6 sites in Switzerland

Participants

Inclusion criteria: Clinical diagnosis of CAP, ECOPD, bronchitis with X‐ray confirmation

Exclusion criteria: People with active intravenous drug use, severe immunosuppression other than corticosteroid use, life‐threatening medical comorbidity leading to possible imminent death, hospital‐acquired pneumonia (development of pneumonia 48 hours after hospital admission or if they were hospitalised within 14 days before presentation), and chronic infection necessitating antibiotic treatment

Included in this analysis: 1359 out of 1381 randomised participants; 22 post randomisation exclusions due to withdrawal of consent

Interventions

Guiding antibiotic decisions in emergency department patients with different ARIs with repeated measurements

Algorithm used in this study: Initiation or continuation of ABs was strongly discouraged if PCT was less than 0.1 µg/L and discouraged if levels were 0.25 µg/L or lower. Initiation or continuation of ABs was strongly encouraged if PCT was higher than 0.5 µg/L and encouraged if levels were higher than 0.25 µg/L. If ABs were withheld, hospitalised patients were clinically re‐evaluated and PCT measurement was repeated after 6 to 24 hours.

Outcomes

• composite of overall adverse outcomes including death from any cause, ICU admission for any reason, disease‐specific complications, and recurrence of LRTI in need of ABs

• any of above outcomes

• length of stay

• side effects from antibiotics

Notes

Funding: This work was supported in part by grant SNF 3200BO‐116177/1 from the Swiss National Science Foundation and contributions from santésuisse and the Gottfried and Julia Bangerter‐Rhyner‐Foundation, the University Hospital Basel, the Medical University Clinic Liestal, the Medical Clinic Buergerspital Solothurn, the Cantonal Hospitals Muensterlingen, Aarau and Lucerne, respectively, the Swiss Society for Internal Medicine, and the Department of Endocrinology, Diabetology and Clinical Nutrition, University Hospital Basel. B·R·A·H·M·S Inc, the major manufacturer of the PCT assay, provided all assay‐related material, Kryptor machines if not already available onsite, and kits and maintenance required for 10,000 measurements related to the study.

Follow‐up: Fixed follow‐up period after 30 days and 180 days

Registration: ISRCTN95122877

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Independent statistician created randomisation scheme.

Allocation concealment (selection bias)

Low risk

Central randomisation using a study web site

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Interviews by blinded medical students, data safety monitoring board

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 1358/1359 (100%)

Selective reporting (reporting bias)

Low risk

Outcomes match previously published protocol.

Other bias

Low risk

91% adherence to PCT algorithm in PCT group

Methods

Randomised, multicentre, single‐blind, controlled clinical trial in 11 ICUs in Australia

Participants

Inclusion criteria: Patients older than 18 years of age, admitted to ICU within the previous 72 hours, receiving parenteral and/or enteral antibiotics for a suspected bacterial infection (with 2 or more SIRS criteria) and expected to remain in the ICU for longer than 24 hours were eligible.
Exclusion criteria: Patients receiving antibiotics for surgical prophylaxis or with proven bacterial infection requiring more than 3 weeks’ antibiotic therapy, isolated systemic fungal or systemic viral infection in the absence of bacterial infection, neutropenia with a neutrophil count less than 1000 cells/mL, receiving immunosuppressive agents, cardiac surgery or trauma or heat stroke within 48 hours, medullary thyroid or small cell lung cancer, patient not expected to survive to hospital discharge, or known pregnancy
Included in this study: Of 400 randomised participants, 394 finished 90 days' follow‐up for survival. 6 withdrew their consent.

Interventions

Guiding antibiotic decisions in critically ill patients in the ICU with undifferentiated infection or suspected sepsis

Algorithm used in this study: Cessation of antibiotics was recommended if initial or any subsequent PCT was negative (<0.10 ng/mL) or if initial or any subsequent PCT was between 0.10 to 0.25 ng/mL, and infection was highly unlikely; Subsequent PCT level declined more than 90% from baseline, and 2. Assess antibiotic appropriateness and/or adequacy of source control if PCT level at 48 hours is 70% of baseline value. Daily PCT results were made available to the treating clinician for participants randomised to the PCT group. Antibiotic prescription in both the standard care and PCT groups was according to the Australian Antibiotics Therapeutic Guidelines and the antimicrobial stewardship (implemented by infectious diseases twice‐weekly rounds and on‐need consultations). The algorithm was implemented only in the ICU.

Outcomes

• time to antibiotic cessation at 28 days, hospital discharge, or death, whichever came first after randomisation

• antibiotic‐free days at day 28 after randomisation

• number of antibiotic daily defined doses at day 28

• ICU and hospital length of stay

• mortality and 90‐day all‐cause mortality

• additional a priori outcomes included the relationship between baseline (taken at randomisation) PCT and sepsis severity, microbiologically confirmed infections within 72 hours, and the predictive value of baseline and serial PCT of mortality

• readmission

• emergence of resistant micro‐organisms

• number of algorithm violations

Notes

Funding: Funded by a competitive grant from the Intensive Care Foundation of Australia and New Zealand. Material support was provided by Roche Diagnostics, Thermo Fisher Scientific, and bioMérieux. Roche Diagnostics and Thermo Fisher Scientific provided additional unrestricted grant funding.
Follow‐up: 90 days' post randomisation for survival
Trial registration: ACTRN12610000809033

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Patients were variable block randomised 1:1 via a secured central study web site into either a PCT‐guided or clinician‐guided group.

Allocation concealment (selection bias)

Low risk

Central randomisation with variable blocks

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

For the standard care group, clinicians were blinded to the PCT levels, but the physicians were aware of the participants' study group due to the study design.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Data were collected by professional research personnel at each site and entered into a central secured database at the Clinical Informatics and Data Management Unit, Department of Epidemiology and Preventive Medicine, Monash University and analysed by a blinded biostatistician at Monash University, Melbourne, Australia. The study was monitored by an independent data safety and monitoring committee, with no interim analysis performed.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

The follow‐up for mortality was 394/394 (100%).

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol. Trial registered (ACTRN12610000809033).

Other bias

Low risk

High adherence to the PCT algorithm (97%). "The proportion of study days where the PCT algorithm was not followed was less than 3%, the majority of which was due to missed PCT sampling."

Methods

Randomised clinical trial, single‐centre, University Hospital Basel, Switzerland

Participants

Inclusion criteria: Clinical diagnosis of COPD exacerbation

Exclusion criteria: People who were considered to be vulnerable study participants (i.e. those with psychiatric comorbidity) were excluded from the study. Other exclusion criteria were immunosuppression, asthma, cystic fibrosis, and the presence of infiltrates on chest radiographs on hospital admission.

Included in this analysis: 208 out of 226 randomised participants; 18 post randomisation exclusions due to absence of COPD according to GOLD criteria

Interventions

Guiding antibiotic decisions in COPD patients with repeated measurements

Algorithm used in this study: Procalcitonin level of 0.1 µg/L was considered indicative of the absence of bacterial infection, and the use of ABs was discouraged. A level of 0.1 to 0.25 µg/L indicated possible bacterial infection, and the use of ABs was discouraged or encouraged, respectively, based on the stability of the participant’s clinical condition. A PCT level of 0.25 µg/L was considered suggestive of the presence of bacterial infection, and AB treatment was encouraged.

Outcomes

• antibiotic use

• “clinical success” defined as improvement of symptoms compared to exacerbation status

• “clinical failure” defined as the absence of the attenuation of symptoms, the worsening of symptoms, or death

• mortality

• ICU admission

• hospital readmission after 30 days and 6 months

Notes

Funding: This study was funded by the Clinic of Pulmonary Medicine; the Clinic of Endocrinology, Diabetes and Clinical Nutrition; and the Emergency Department of the University Hospital Basel. B·R·A·H·M·S provided PCT assays for this investigator‐driven study.

Follow‐up: Short‐term follow‐up visit after 14 to 21 days; long‐term follow‐up visit at 6 months

Registration: ISRCTN77261143

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Independent statistician created randomisation list.

Allocation concealment (selection bias)

High risk

Sealed envelopes, not numbered

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

Blinded personnel

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 208/208 (100%)

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol.

Other bias

Unclear risk

Adherence to PCT protocol not reported/assessed.

Methods

Randomised clinical trial, multicentre with 7 European and US intensive care units

Participants

Inclusion criteria: VAP when intubated for > 48 h

Exclusion criteria: Patients were excluded it they 1) were pregnant; 2) were enrolled in another trial; 3) had received immunosuppressants or long‐term corticosteroid therapy (> 0.5 mg/kg per day for > 1 month); 4) were severely immunosuppressed, including AIDS; or 5) had a coexisting extrapulmonary infection diagnosed between day 1 and 3 requiring antibiotic therapy for > 3 days.

Included in this analysis: 101 (101) (100%)

Interventions

Guiding antibiotic decisions in VAP patients with repeated measurements

Algorithm used in this study: A PCT level of < 0.25 µg/L suggested the absence of VAP, and discontinuation of ABs was strongly encouraged. A PCT level between 0.25 µg/L and 0.5 µg/L or a decrease by ≥ 80% as compared to day 0 indicated that bacterial infection was unlikely, and reduction or discontinuation of ABs was encouraged. A PCT level ≥ 0.5 µg/L or decrease by < 80% as compared to day 0 was considered indicative of unresolved bacterial infection, and reduction or discontinuation of AB was discouraged. A PCT level of > 1 µg/L strongly suggested unresolved bacterial infection, and AB discontinuation was strongly discouraged.

Outcomes

• antibiotic‐free days alive

• any antibiotic exposure after inclusion, i.e. total antibiotic exposure days and total antibiotic agent days, regardless of indication

• the number of mechanical ventilation‐free days

• the number of ICU‐free days alive

• the evolution of the signs and symptoms potentially linked to pulmonary infection

• SaO₂, PaO₂/FiO₂

• the evolution of the SOFA, ODIN, and CPIS scores

• length of hospital stay

• the VAP‐related clinical deterioration rate and overall mortality at 28 days

Notes

Funding: Funding was granted by the Clinic of Pulmonary Medicine, University Hospital Basel. Funding obtained from B·R·A·H·M·S AG (Hennigsdorf, Germany)

Follow‐up: Fixed follow‐up period of 28 days

Registration: ISRCTN61015974

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Independent statistician created randomisation list.

Allocation concealment (selection bias)

Low risk

Quote: "Randomisation was through arbitrary allocation to one of the two treatment assignments based on sealed, opaque envelopes. Block size was 20 envelopes. Treating physicians were not aware of envelope contents before randomisation"

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Open‐label trial where physicians knew to which group participants had been assigned and where PCT levels were only communicated in the intervention arm

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Unblinded study member

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Follow‐up for mortality: 101/101 (100%)

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol.

Other bias

Unclear risk

Adherence to PCT protocol not reported/assessed.

Methods

Randomised, single‐centre, single‐blinded, controlled clinical trial in the emergency department of the Fifth People’s Hospital of Shanghai, China

Participants

Inclusion criteria: 1) ≥ 18 years old; 2) has any, or all, of the following clinical features as defined by the Global Initiative for National Asthma (GINA) Guidelines: dyspnoea, wheeze, acute cough, increased work of breathing, increased requirement for beta2‐agonist from baseline use, O₂ saturation < 95%, a peak expiratory flow (PEF) at randomisation ≤ 80% of their known best (within the last 12 months) or, in the absence of this information, of their predicted PEF
Exclusion criteria: 1) treatment with antibiotics within 2 weeks prior to recruitment; 2) bacterial infection in other parts of body than the respiratory system; 3) chest X‐ray‐confirmed pneumonia; 4) suffering from other chronic respiratory diseases; 5) suffering from severe organ dysfunction
Included in this study: 265 people were eligible, and 255 participants completed the study.

Interventions

Guiding antibiotic decisions in patients with acute exacerbation of asthma

Algorithm used in this study: Participants in the PCT group were treated with antibiotics based on their PCT serum level according to the following guidelines: antibiotics treatment was strongly discouraged when serum PCT level was < 0.1 μg/L; antibiotics treatment was discouraged when serum PCT level was < 0.25 μg/L; antibiotics treatment was encouraged when serum PCT level was > 0.25 μg/L.

Outcomes

• antibiotic prescription rate and the relative risk of antibiotic exposure

• clinical, laboratory, and lung function outcomes at the follow‐up visit (6 weeks)

• secondary ED visits, hospital readmissions, repeated need for steroids or dosage increase, need for antibiotics, white blood cell count, PCT levels, and FEV1% were assessed during the 6‐week follow‐up period

Notes

Funding: The study was sponsored by a grant from the Shanghai Fifth People’s Hospital Science Foundation and Minhang District Natural Science Foundation of Shanghai.
Follow‐up: 6 weeks
Trial registration: ICTRP ChiCTR‐TRC‐12002534

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Allocation to either intervention was conducted according to computer‐generated random numbers produced by an independent statistician.

Allocation concealment (selection bias)

Low risk

After randomisation, an opaque, sealed, and sequentially numbered envelope containing the PCT or control protocol was prepared for each participant according to the group.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Attending physicians responsible for participants in the control group remained unaware of the participants’ PCT concentrations throughout the study, but blinding was not feasible due to the study design.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

All participants, laboratory technicians, investigators, and research designers were blinded to participant assignments until the data analysis was completed. There were no protocol violations during the study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

265/265 participants completed follow‐up for mortality.

Selective reporting (reporting bias)

Low risk

Outcomes correspond to study protocol: ICTRP ChiCTR‐TRC‐12002534.

Other bias

Low risk

100% adherence to the PCT algorithm. "There were no protocol violations during the study."

Methods

Randomised, multicentre, open, controlled, parallel‐group, non‐inferiority trial involving 18 university/city hospital pulmonary departments in Italy

Participants

Inclusion criteria: Study participants were male or female, 18 years of age, current or former smokers, and diagnosed with COPD stages I‐IV as defined by GOLD guidelines available at the time the study was designed, with protocol deviation. Participants were hospitalised for severe ECOPD requiring antibiotic treatment, i.e. type 1 exacerbation (increased dyspnoea, sputum volume, and sputum purulence verified by the attending clinician) according to Anthonisen, and/or characterised by respiratory failure. Exacerbation of chronic obstructive pulmonary disease was defined as “an acute event characterised by a worsening of the patient’s respiratory symptoms that is beyond normal day‐to‐day variations and leads to a change in medication”.
Exclusion criteria: Bronchial asthma, unstable concomitant disease (cardiovascular, renal, hepatic, gastrointestinal, neurological, metabolic, musculoskeletal, neoplastic, respiratory, or other disease), pregnancy and breastfeeding, clinically significant laboratory abnormalities suggestive of unstable concomitant disease, survival for 1 year unlikely, and inability to give written consent. Antibiotic pretreatment before hospital admission and radiographic signs of pneumonia did not preclude eligibility for the study.
Included in this study: 183 participants were randomised, of which 178 participants, 88 in the PCT group and 90 in the standard care group, were analysed.

Interventions

Guiding antibiotic decisions in people with severe exacerbations of COPD

Algorithm used in this study: On admission, all patients received a 3‐day course of antibiotics (either amoxicillin plus clavulanate or quinolones) according to 2005 international guidelines. Procalcitonin levels were measured on hospital admission, on day 1, and on day 2. On day 2 each eligible patient was randomly assigned to 1 of the 2 treatment plans.
quote "Participants randomised to the standard group continued antibiotic therapy for 10 days, whereas participants randomised to the PCT group either continued treatment for 10 days or stopped on day 3, depending on their PCT levels, according to previously recommended cut‐off values. Specifically, participants continued antibiotic treatment for 10 days if 1 or more of the PCT values on the first 3 days of hospitalisation were 0.25 μg/L. When PCT values were < 0.25 μg/L but > 0.1 μg/L on any occasion, antibiotic treatment was continued for 10 days if participants were clinically unstable or had acute respiratory failure; otherwise, treatment was stopped on day 3. If all PCT values were consistently < 0.1 μg/L, treatment was stopped on day 3."
All participants were also treated with systemic corticosteroids for 14 days, plus regular inhaled short‐acting or long‐acting bronchodilators.

Outcomes

• percentage of participants with at least 1 exacerbation within 6 months after the index exacerbation

• hospital readmission

• admission to the ICU

• change in lung function (ΔFEV1)

• length of hospital stay

• death from any cause

Notes

Funding: The trial was approved and funded by the Agenzia Italiana del Farmaco (AIFA), the Italian agency for drugs, which is an official body of the Italian Ministry of Health.
Follow‐up: Follow‐up visits were scheduled on day 1, day 3, and 6 months after discharge; telephone interviews were conducted at 2, 4, and 5 months after discharge.
Trial registration: NCT01125098

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

The randomisation was web based, and only statisticians and the web site administrator knew the randomisation sequence

Allocation concealment (selection bias)

Low risk

Eligible patients were randomly assigned to receive standard antibiotic therapy (standard group) or PCT‐guided antibiotic treatment (PCT group) according to a 1:1 permuted block computer‐generated scheme, stratified according to hospital

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

Blinding was not feasible due to the study design

Blinding of outcome assessment (detection bias)
All outcomes

High risk

No blinding of outcome assessment mentioned in the study. The authors state that: "... because we anticipated that the primary outcome (exacerbations of COPD) would be strong and easy to identify, and thus unlikely to be biased by investigator influence, we did not adopt any procedure to reduce bias during the follow‐up part of the trial."

Incomplete outcome data (attrition bias)
All outcomes

Low risk

187 of 192 randomised participants were included in final analysis.

Selective reporting (reporting bias)

Low risk

Results correspond to the trial registration.

Other bias

Low risk

High adherence to the PCT protocol (95.5%)

The study planned to enrol 400 participants to have enough statistical power for the primary endpoint, but randomised only 183.

Methods

Randomised, single‐centre, single‐blinded, controlled clinical trial in the Beijing Luhe Hospital, China

Participants

Inclusion criteria: Patients with AECOPD who were 40 years of age, had sound understanding and language abilities, and who had a PCT level < 0.1 ng/mL were included.
Exclusion criteria: Fever (38°C), tracheal intubation within 24 h after hospital admission, a PCT level of 0.1 ng/mL on admission, pneumonia, chronic renal failure, history of malignant disease, immunosuppressive therapy, and refusal to participate
Included in this study: 194 randomised participants. 191 finished the 30‐day follow‐up.

Interventions

Guiding antibiotic decisions in patients with acute exacerbation of chronic obstructive pulmonary disease

Algorithm used in this study: Patients with a PCT concentration < 0.1 ng/mL were randomised. Antibiotics were withheld from participants in the control group. However, antibiotics could be administered later for participants whose clinical condition was unstable or who had a worsening of symptoms and signs, and for those with positive evidence of bacteria as assessed by the attending physicians. In the antibiotic group, antibiotics were administered routinely.

Outcomes

• treatment success rate on day 10 after admission

• symptoms assessed by visual analogue scale (at hospital admission, 3 days after hospitalisation, and on the day of hospital discharge)

• length of hospital stay

• intubation rate

• mortality during hospitalisation and the 30‐day follow‐up period

• rate of antibiotic use

• readmission due to AECOPD in the 30‐day follow‐up period

Notes

Funding: The study was sponsored by the National Science Fund for Distinguished Young Scholars (81425001/H0104) for Dr Bin Cao.
Follow‐up: 30 days after hospital discharge
Trial registration: ChiCTR‐TRC‐14004726

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

A computer digital table method was used to generate randomisation numbers.

Allocation concealment (selection bias)

Low risk

Researchers in this study had 24‐hour access to randomisation numbers, allowing immediate and concealed allocation to the trial.

Blinding of participants and personnel (performance bias)
All outcomes

Unclear risk

No blinding feasible due to the study protocol.

Blinding of outcome assessment (detection bias)
All outcomes

High risk

Individuals responsible for allocation concealment were not allowed to take part in the measurement of results, but no overall blinding of the outcome assessment was mentioned in this study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Only 3 participants were excluded during the 30‐day follow‐up, due to a diagnosis of pneumonia according to chest X‐ray.

Selective reporting (reporting bias)

Low risk

Outcomes correspond to the trial registration (ChiCTR‐TRC‐14004726).

Other bias

Low risk

High adherence to the PCT protocol (82.3%)