Procalcitonin-guided antibiotic prescription in patients with respiratory syncytial virus and influenza virus Authors: Lisa Hessels, Didi Vossen, Miriam Leeuw, Nienke Paternotte, Dian Karssen, Tony Chiang, Lotte Terpstra, Wim Boersma Published: 2025 ABSTRACT AbstractBackgroundBacterial co-infections may occur in patients with influenza and respiratory syncytial virus (RSV), often leading to high unnecessary antibiotic exposure. Procalcitonin (PCT) may help reduce inappropriate antibiotic prescriptions in viral infections.MethodsIn this retrospective cohort study, data was analysed from 558 influenza A/B and 175 RSV patients (2017–2024). Patients were divided into PCT (antibiotic prescription guided by PCT levels) and control groups (antibiotic prescription based on clinical judgement). Outcomes included antibiotic use, defined daily dose (DDD), duration of antibiotic treatment (DOT) and secondary outcomes (readmissions, ICU admissions, mortality, mechanical ventilation).ResultsAt admission, 139 (49.6%) of influenza patients in the control group and 148 (53.2%) in the PCT group received antibiotics (adjusted OR 1.20, 95% CI 0.79 to 1.83, p=0.325), indicating no significant difference compared with the control group. For RSV patients, 45 control patients (41.7%) and 33 PCT patients (49.3%) received antibiotics (adjusted OR 0.98, 95% CI 0.44 to 2.14, p=0.490), also showing no significant difference. For influenza, antibiotics were initiated in 175 control patients (62.7%) and 187 PCT patients (67.3%) (adjusted OR 1.25, 95% CI 0.81 to 1.92, p=0.313) during hospitalisation. For RSV, 62 control patients (57.4%) and 58 PCT patients (71.6%) received antibiotics (adjusted OR 1.30, 95% CI 0.60 to 2.89, p=0.498). No significant differences in DOT or DDD were observed for either group. PCT testing showed no significant impact on secondary outcomes.ConclusionIn this retrospective design, PCT testing did not significantly reduce antibiotic use or dosage, suggesting limited utility for optimising antibiotic use in influenza and RSV infections.Keywords: Viral infection, Pneumonia, Respiratory Infection WHAT IS ALREADY KNOWN ON THIS TOPIC Procalcitonin (PCT)-guided antibiotic prescribing has been shown to reduce unnecessary antibiotic use in respiratory tract infections, but its role in confirmed influenza and respiratory syncytial virus (RSV) infections remains unclear. Procalcitonin (PCT)-guided antibiotic prescribing has been shown to reduce unnecessary antibiotic use in respiratory tract infections, but its role in confirmed influenza and respiratory syncytial virus (RSV) infections remains unclear. WHAT THIS STUDY ADDS In this retrospective cohort of hospitalised patients with influenza and RSV, PCT testing did not significantly reduce antibiotic prescriptions, dosage or duration. Antibiotic use remained common, even in patients with low PCT levels. In this retrospective cohort of hospitalised patients with influenza and RSV, PCT testing did not significantly reduce antibiotic prescriptions, dosage or duration. Antibiotic use remained common, even in patients with low PCT levels. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY These results of this retrospective indicate that PCT testing alone does not reduce antibiotic use in influenza and RSV, and clinical decisions should therefore not rely solely on PCT results. These results of this retrospective indicate that PCT testing alone does not reduce antibiotic use in influenza and RSV, and clinical decisions should therefore not rely solely on PCT results. Introduction Respiratory syncytial virus (RSV) and influenza are significant global contributors to respiratory illnesses.1 2 While RSV was once considered primarily a paediatric concern, it is now recognised as a significant cause of respiratory illness in adults over 65 and immunocompromised individuals.3 4 Similarly, influenza poses a substantial burden on the elderly, with hospitalisation rates as high as 96.8 per 100 000 and considerable seasonal mortality.5 6 Bacterial co-infections further complicate the management of patients hospitalised with either RSV or influenza, leading to increased mortality and adverse outcomes.7 8 Bacterial infections are reported in approximately 9%–13% of RSV cases and 11–53% of influenza cases, but they are difficult to distinguish from pure viral respiratory tract infections based on clinical symptoms.9,11 Inflammatory markers like C-reactive protein (CRP) lack specificity, leading to frequent overprescription of antibiotics.12 13 This practice contributes to antimicrobial resistance and immediate patient risks, such as adverse drug effects.14,16 Around 65%–75% of hospitalised patients with respiratory tract infections receive antibiotics, even when bacterial involvement is unconfirmed.17,19 Procalcitonin (PCT)-guided antibiotic prescribing has emerged as a promising strategy for reducing inappropriate antibiotic use. PCT, a precursor of the hormone calcitonin, is produced in response to bacterial infections and upregulated by cytokines such as interleukin-6 and TNF-α.20 21 Unlike CRP, PCT levels rise earlier in bacterial infections and remain relatively low in infections with viral and atypical pathogens, making it a more sensitive and specific marker for bacterial co-infections.22,24 Several studies have shown that PCT-guided protocols can safely reduce antibiotic use in lower respiratory tract infections.25,27 However, there is limited research evaluating the role of PCT-guided antibiotic prescribing specifically in patients with confirmed influenza and RSV, despite the inappropriate antibiotic use in this population. This study aims to evaluate the role of PCT measurements in reducing antibiotic use in adult patients hospitalised with RSV or influenza. We will explore whether PCT-guided prescribing can safely reduce antibiotic exposure in these viral infections without compromising patient outcomes. Methods Study design and setting This is a retrospective, single-centre, cohort study which investigates the role of PCT in reducing antibiotic prescription in hospitalised patients with Influenza A/B and RSV. The study was performed in Northwest Clinics Alkmaar, a large teaching hospital in the Netherlands. Data from patients with influenza and RSV between January 2019 and May 2024 were collected and used for this analysis. Patients were classified into a PCT group if a PCT test was performed on admission and into a control group if no PCT test was performed. The study was approved by the institutional research committee of Northwest Clinics. Since retrospective data was used only, the need for informed consent was waived. Study participants Patients were eligible for inclusion in the study if they were ≥18 years of age, had symptoms of respiratory tract infection, were admitted to the hospital and had a rapid positive PCR test for RSV or influenza A/B. The PCR was obtained from nasal secretions, nasopharyngeal secretions or oropharyngeal at admission. Patients with hospital-onset infections, those discharged within 24 hours and outpatients were not eligible. Study Group classification Group 1: PCT group For the PCT group, patients were included when the diagnosis of RSV or influenza A/B was confirmed at the emergency department and a PCT level was determined at the emergency department or at hospital admission (within 24 hours of admission). The diagnosis of influenza or RSV infection was made by the treating physician and was based on clinical symptoms, laboratory test results, radiology and ultimately confirmed by a positive PCR. Admitted patients with viral respiratory tract infections were treated in accordance with the Dutch guidelines.28,30 Antibiotics are not part of the standard treatment; however, patients were treated with antibiotics if there was a suspicion of a bacterial co-infection. In this group, the decision to treat patients with additional antibiotics was guided according to generally used cut-off results of PCT algorithm in respiratory tract infections (online supplemental table 1). PCT results were documented as numerical values in the laboratory report. The PCT-guided protocol was advised to the treating physician to guide the decision to prescribe or withhold antibiotics and incorporated into departmental protocols. The ultimate decision was left to the treating physician. The PCT cut-off values were based on previous studies.31 PCT testing during admission was not routinely performed. Group 2: control group The control group consisted of patients admitted with influenza A/B or RSV in whom no PCT level was determined at hospital admission. The diagnosis of influenza or RSV infection was made by the treating physician and was based on clinical symptoms, laboratory test results, radiology and ultimately confirmed by a positive PCR. As stated above, antibiotics are not the appropriate treatment for viral infections unless a bacterial co-infection is suspected. The decision to prescribe or withhold antibiotics was left to the treating physician. The decisions were made based on the available vital signs, radiology images and laboratory results, including CRP as the most important inflammation parameter. Although antimicrobial stewardship programmes have been present in the hospital in the past, no programme was active during the study period; therefore, study results were not influenced by other stewardship interventions. Laboratory testing For the PCR test, a GeneXpert influenza/RSV reagent with nasal secretions, nasopharyngeal secretions, oropharyngeal secretions in an e-Swab or virus transport medium was used on a GeneXpert Dx System (Cepheid, France). The GeneXpert Xpress influenza/RSV test is an automated multiplex RT-PCR that distinguishes between the H1N1 Influenza subtype 2009 and the seasonal influenza A and B and RSV. The turnaround time from specimen collection to results is less than 2 hours. PCT testing was performed by the hospital laboratory (if ordered) within the same workflow as other standard laboratory tests. Results were available 1.5–2 hours after a blood sample was taken. PCT was measured using an Atellica IM BRAHMS PCT reagent on an Atellica IM analyser (Siemens Healthcare Diagnostics Inc. Tarrytown, NY 10591, USA). PCT reference values were established as <0.05 µg/L. Within-laboratory precision is determined as <7% coefficient of variation (CV) samples from 0.05 to 2.0 µg/L and <5% CV for samples >2.0 µg/L. Study outcomes The primary outcome was the proportion of patients receiving antibiotics at admission. Antibiotics at admission were defined as those decided and prescribed at the emergency department. The second primary outcome was the amount of antibiotics prescribed, by means of defined daily dose (DDD) and days of therapy (DOT). The primary outcomes were compared between patients with PCT levels available and patients without PCT levels available. The DDD is defined by the WHO as the assumed average maintenance dose of a specific antibiotic per day used for its main indication in adults.32 The total prescribed DDD per antibiotic per patient was calculated by dividing the total number of used milligrams by the recommended DDD. DOT was defined as the number of days a patient received each prescribed antimicrobial agent, regardless of dose, route or frequency of administration. All in-hospital prescribed antibiotics during the index admission were included, including prescriptions at discharge. Index admission was defined as the primary admission with clinical symptoms of RSV or Influenza as indication. Secondary outcomes included treatment failure, median length of hospital stay, ICU admission, all-cause mortality within 30 and 90 days of admission, readmissions and occurrence of adverse outcomes during admission. Treatment failure was defined as prescribing antibiotics for a lower respiratory tract infection within 7 days after initially withholding antibiotics. Adverse outcomes during admission were defined as mechanical ventilation, non-invasive ventilation (NIV), high flow nasal cannula (HFNO) and septic shock. Additionally, protocol adherence was determined in the PCT patient group. Protocol adherence was defined as treated in line with the available protocol. Data collection Data was collected using the electronic case report form (Castor EDC 2021.5.0, Amsterdam, the Netherlands). Data were extracted from patient’s medical electronic files. For all included patients, the following data were collected from the medical records: demographics, comorbidities, symptoms before admission, duration of symptoms, vital signs, radiological findings on admission, laboratory results on admission, outcomes of bacterial cultures during admission, antibiotic therapy prescribed during admission or at discharge and in-hospital outcomes. Mortality at day 30 and 90 was checked in hospital records and the municipal administration. Statistical analysis The characteristics of the patients were described using standard descriptive statistics and compared by PCT status. Variables were compared between study groups using the Wilcoxon Rank Sum test for continuous variables. For categorical variables, χ² test or Fisher’s exact test was used. Logistic regression was used to evaluate the association between PCT testing and antibiotic prescription at admission and during the total admission, adjusting for potential confounder (infiltrate on chest X-ray). Assumptions of logistic regression were assessed using the Box-Tidwell test and visual inspection of the residuals. To model DDD and DOT, a two-part hurdle model was applied. This model is suited for situations where the outcome variable is zero for a significant portion of the data, such as when patients have received no antibiotics. The first part of the model used logistic regression to predict the probability of non-zero DDD and DOT values, with confounders including infiltrate on chest X-ray. The second part applied Gamma regression with a log-link to estimate the magnitude of DDD and DOT among patients with non-zero values. Model assumptions were validated through residual simulations. The association between PCT testing and ICU admission and NIV/HFNO use was analysed using logistic regression models, adjusting for age and Charlson Comorbidity Index (CCI). Survival analysis was performed for comparison of 30- and 90-day mortality outcomes between PCT status. Kaplan-Meier survival curves were plotted, and log-rank tests were used for group comparisons. Cox proportional hazards models were used for adjusted analysis, adjusting for age and CCI. The proportional hazards assumption was tested using Schoenfeld residuals. For the multivariate models, covariates were chosen based on clinical relevance. Missing data were accepted after maximal efforts to retrieve them from available sources. All analyses were performed with R Statistical Software (V.4.2.1, R Core Team 2021). The sample size calculation is included in online supplement 1. Results Baseline characteristics Between January 2019 and May 2024, 550 patients were admitted with influenza A, eight patients with influenza B and 175 patients with RSV. A detailed overview of patient screening, exclusions and final enrolment is presented in online supplemental figure 1. Table 1 shows the baseline characteristics of the study population, grouped based on whether a PCT test was performed or not. The median age was 75 years and 49% of participants were male, with a higher proportion of males in the PCT group (183 patients (53%) vs 175 patients (45%), p=0.035). The median time from symptom onset to admission was similar in both groups (4.0 vs 4.5 days, p=0.11). At admission, respiratory rate (p=0.003) and the need for oxygen supplementation (p=0.014) were higher in the PCT group, while rates of dyspnoea, cough, sputum and fever were similar between both groups. CRP levels were also higher in the PCT group (84 vs 54, p<0.001). Infiltrates on chest X-ray or CT scan were more frequently observed in the PCT group compared with the control group (154 (45%) patients vs 154 (32%) patients, p<0.001). Influenza A detection was more frequent in the PCT group (274 (79%) patients vs 276 (71%) patients), while RSV detection was higher in the control group (108 (28%) patients vs 67 (19%) patients (p=0.027). Infiltrates on chest X-ray or CT scan were present in 224/558 (40%) of influenza patients and 56/175 (32%) of RSV patients. Table 1Baseline characteristics, overall and by PCT groupOverall n=733No PCT test performed n=388PCT test performed n=342P Value*Age†75 (66–81)75 (66–82)74 (65–80)0.13Sex (male)358 (49%)175 (45%)183 (53%) 0.035 BMI<2020–2525–30>3070 (11%)222 (35%)202 (32%)145 (23%)40 (12%)127 (37%)103 (30%)73 (21%)30 (10%)95 (32%)99 (33%)72 (24%)0.4SmokingFormer smokerActive smoker292 (41%)144 (20%)138 (37%)78 (21%)154 (46%)66 (20%) 0.026 Antibiotic pretreatment‡152 (21%)86 (22%)67 (19%)0.4ComorbiditiesCOPDAsthmaDiabetesCongestive heart failureMalignancyCharlson Comorbidity Index†230 (31%)90 (12%)148 (20%)115 (16%)124 (17%)7 (5–8)113 (29%)47 (12%)85 (22%)60 (15%)66 (17%)7 (5–8)117 (34%)43 (13%)63 (18%)55 (16%)58 (17%)6 (5–8)0.20.90.20.9>0.90.3Time from onset of symptoms to admission, days†4.0 (3.0–7.0)4.5 (3.0–6.0)4.0 (3.0–7.0)0.11Symptoms at admissionDyspnoeaCoughSputum productionRespiratory rate†O2 suppletionFever554 (87%)629 (93%)348 (67%)22 (18–28)429 (59%)368 (66%)273 (86%)314 (93%)166 (67%)22 (18–27)210 (55%)181 (63%)281 (89%)306 (94%)182 (67%)24 (20–28)219 (64%)187 (68%)0.20.8>0.90.0030.0140.2Infiltrate on chest X-ray or CT scan280 (38%)126 (32%)154 (45%) <0.001 Laboratory findings at admission†CRP (mg/L)WBC (x10ˆ9 /L)PCT (µg/L)67 (30–147)8.9 (6.7–11.9)0.22 (0.10–0.69)54 (21–119)8.7 (6.4–12.0)-84 (42–166)9.4 (7.0–11.9)0.22 (0.10–0.69)<0.0010.3PCR positive for:Influenza AInfluenza BRSV550 (75%)8 (1.1%)175 (24%)276 (71%)4 (1.0%)108 (28%)274 (79%)4 (1.2%)67 (19%) 0.027 Bold values indicate statistical significance.*Wilcoxon Rank Sum test; Pearson’s χ² test.†Median (95% IQR), other values are presented as N (%).‡Antibiotic pretreatment = antibiotic use prior to hospital admission. BMI, body mass index; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; PCT, procalcitonin; RSV, respiratory syncytial virus; WBC, white blood cell. Baseline characteristics, overall and by PCT group <20 <20 20–25 20–25 25–30 25–30 >30 >30 Former smoker Former smoker Active smoker Active smoker COPD COPD Asthma Asthma Diabetes Diabetes Congestive heart failure Congestive heart failure Malignancy Malignancy Charlson Comorbidity Index† Charlson Comorbidity Index† Dyspnoea Dyspnoea Cough Cough Sputum production Sputum production Respiratory rate† Respiratory rate† O2 suppletion O2 suppletion Fever Fever CRP (mg/L) CRP (mg/L) WBC (x10ˆ9 /L) WBC (x10ˆ9 /L) PCT (µg/L) PCT (µg/L) Influenza A Influenza A Influenza B Influenza B RSV RSV Bold values indicate statistical significance. Wilcoxon Rank Sum test; Pearson’s χ² test. Median (95% IQR), other values are presented as N (%). Antibiotic pretreatment = antibiotic use prior to hospital admission. BMI, body mass index; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; PCT, procalcitonin; RSV, respiratory syncytial virus; WBC, white blood cell. Antibiotic treatment at admission The proportions of antibiotic prescriptions at admission and during the total admission are displayed in figure 1. For influenza A/B patients, antibiotic treatment was started at admission in 139 patients (49.6%) of the control group and 148 patients (53.2%) of the PCT group (table 2). The adjusted OR for starting antibiotics in the PCT group compared with the control group was 1.20 (95% CI 0.79 to 1.83, p=0.391). For RSV patients, antibiotics were initiated in 45 patients (41.7%) of the control group and 33 patients (49.3%) of the PCT group. The adjusted OR for the PCT group compared with the control group was 0.98 (95% CI 0.44 to 2.15, p=0.968). Proportion of antibiotic prescription at admission and during total admission. PCT, procalcitonin; RSV, respiratory syncytial virus. Table 2Antibiotic treatment started in procalcitonin (PCT) group vs control groupNObserved (%)Unadjusted OR*P valueAdjusted OR†P value Antibiotic treatment started at admission Influenza A/B Control, n=280PCT, n=278139 14849.6%53.2%1.32 (0.86–1.98)0.1721.20 (0.79–1.83)0.391 RSV Control, n=108PCT, n=67453341.7%49.3%1.38 (0.67–2.81)0.3790.98 (0.44–2.14)0.968 Antibiotic treatment started during admission Influenza A/B Control, n=280PCT, n=278175 18762.7%67.3%1.36 (0.89–2.06)0.1521.24 (0.81–1.92)0.313 RSV Control, n=108PCT, n=67625857.4%71.6%1.65 (0.79–3.52)0.1831.31 (0.60–2.89)0.498*Pearson’s χ² test.†Multivariate logistic regression using infiltrate on chest X-ray/CT and sputum production as covariates.RSV, respiratory syncytial virus. Antibiotic treatment started in procalcitonin (PCT) group vs control group Pearson’s χ² test. Multivariate logistic regression using infiltrate on chest X-ray/CT and sputum production as covariates. RSV, respiratory syncytial virus. Antibiotic treatment during total admission During the entire admission period, 175 patients (62.7%) in the control group and 187 patients (67.3%) in the PCT group with influenza A/B were initiated on antibiotic treatment (table 2). The unadjusted odds showed no significant difference in antibiotic initiation between the groups (OR 1.36, 95% CI 0.89 to 2.10, p=0.152). Similarly, the adjusted analysis did not reveal a statistically significant difference (adjusted OR 1.25, 95% CI 0.81 to 1.92, p=0.313). Among RSV patients, antibiotics were started in 62 individuals (57.4%) in the control group and 58 individuals (71.6%) in the PCT group during total admission. Neither the unadjusted (OR 1.65, 95% CI 0.80 to 3.52, p=0.183) nor the adjusted (adjusted OR 1.30, 95% CI 0.60 to 2.89, p=0.498) analyses showed significant differences between groups. DDD during admission For influenza A/B patients, the regression analysis revealed that the PCT group had a 1.07-fold non-significant increase in the expected DDD compared with the control group (estimate 0.07, 95% CI −0.22 to –0.36, p=0.637) (table 3). For RSV patients, no significant difference in DDD was observed between groups, with an estimate of −0.06 (95% CI −0.39–0.27, p=0.730) and a relative effect of 0.94. Table 3Antibiotic prescription during total admission in procalcitonin (PCT) vs control group*Mean±SDMedian (IQR)Estimate(95% CI)SEp valueRelative effect†Defined daily dose (DDD)‡ Influenza A/B Control, n=590PCT, n=3045.1±6.87.0 ±11.94.0 (0–7.5)6.0 (0–8.6)0.07 (-0.22 to 0.36)0.150.6371.07 RSV Control, n=157PCT, n=884.6±6.56.1 ± 5.52.0 (0–7.3)7.0 (0–8.8)−0.06 (-0.39 to 0.27)0.170.7300.94Days of therapy (DOT) Influenza A/B Control, n=590PCT, n=3044.3±5.35.2 ± 5.83.5 (0–7.0)5.0 (0–7.0)0.05 (-0.13 to 0.22)0.090.5791.05 RSV Control, n=157PCT, n=883.7±4.74.9 ± 4.12.0 (0–6.0)5.0 (0–7.0)−0.02 (-0.30 to 0.26)0.140.8660.98*Results presented from Gamma regression part of hurdle model.†Relative effect=exp(estimate) represents the multiplicative change in the expected outcome for a one unit increase in the predictor. ‡DDD: The assumed average maintenance dose per day for a drug used for its main indication in adults.DDD per patient = (total amount of drug consumed) / (number of days of drug use) x DDD.PCT, procalcitonin ; RSV, respiratory syncytial virus. Antibiotic prescription during total admission in procalcitonin (PCT) vs control group* Results presented from Gamma regression part of hurdle model. Relative effect=exp(estimate) represents the multiplicative change in the expected outcome for a one unit increase in the predictor. DDD: The assumed average maintenance dose per day for a drug used for its main indication in adults.DDD per patient = (total amount of drug consumed) / (number of days of drug use) x DDD. PCT, procalcitonin ; RSV, respiratory syncytial virus. DOT during admission For influenza A/B patients, the regression analysis indicated a 1.05-fold non-significant increase in the expected DOT in the PCT group compared with the control group (estimate 0.05, 95% CI −0.13 to –0.22, p=0.579) (table 3). For RSV patients, there was no significant difference in DOT, with an estimate of −0.02 (95% CI −0.30 to –0.26, p=0.866) and a relative effect of 0.98. Protocol adherence Antibiotic prescriptions at admission varied by PCT level (online supplemental table 2). Among influenza A/B patients, antibiotics were prescribed to 37/145 (26%) of those with PCT <0.25 µg/L, 24/39 (62%) with PCT 0.25–0.50 µg/L and 87/93 (94%) with PCT >0.50 µg/L. For RSV patients, antibiotics were prescribed to 9/39 (23%) with PCT <0.25 µg/L, 9/10 (90%) with PCT 0.25–0.50 µg/L, and 15/18 (83%) with PCT >0.50 µg/L. The main reason for prescribing antibiotics in the lower PCT groups (<0.25 µg/L and 0.25–0.50 µg/L) was a high suspicion of bacterial superinfection, while for the highest PCT group (>0.50 µg/L), high PCT values were the predominant reason (online supplemental table 3). Treatment failure occurred in 36 influenza patients in the control group (12.8%) and 39 in the PCT group (14.0%) and in 17 RSV patients in the control group (15.7%) and 25 in the PCT group (37.3%). Microbiology Results of microbiology showed similarities and differences between PCT and control groups (online supplemental tables 4 and 5). For influenza patients, blood cultures showed low positivity rates, with 10 positive cultures (3.6%) in the control group and nine positive cultures (3.2%) in the PCT group. Sputum cultures had higher positivity results, with 29 positive cultures (10%) in the control group and 37 positive cultures (13%) in the PCT group. Streptococcus pneumoniae was the most frequently isolated pathogen from both blood and sputum cultures. Five patients had a positive sputum culture for Aspergillus fumigatus in the control group and four patients in the PCT group. None of them had an invasive aspergillosis. For RSV patients, blood cultures also showed low positivity rates, with two positive cultures (1.9%) in the control group and two positive cultures (3.0%) in the PCT group. Sputum culture positivity was higher, with 21 positive cultures (8.3%) in the control group and 26 positive cultures (25%) in the PCT group (p=0.002). S. pneumoniae was again the most common pathogen identified. No co-infections with influenza A/B, RSV or SARS-CoV-2 were observed In-hospital outcomes and mortality The in-hospital outcomes are summarised in tables4 5. For patients with influenza A/B, ICU admissions were significantly less frequent in the PCT group compared with the control group (p<0.005). No significant differences were observed for mechanical ventilation, NIV and HFNO use, or readmission rates. For patients with RSV, no statistically significant differences were found in ICU admissions, mechanical ventilation, NIV and HFNO use, or readmissions between the groups. Regarding mortality outcomes, no significant differences were observed in 30-day or 90-day mortality for either influenza A/B or RSV groups. Table 4In hospital outcomes in procalcitonin (PCT) versus control group Groups Observed, N (%)Unadjusted OR*P valueAdjusted OR†p value Influenza A/B ICU admissionControlPCT63 (11%)15 (4.9%)0.43 (0.24–0.76) 0.005 0.42 (0.23–0.73) <0.005 Mechanical ventilationControlPCT27 (4.6%)8 (2.6%)0.56 (0.24–1.20)0.160.54 (0.23–1.16)014NIV and HFNOControlPCT59 (10%)38 (13%)1.29 (0.83–1.98)0.261.28 (0.83–1.97)0.26ReadmissionHospitalControlPCT43 (7.3%)26 (8.6%)1.19 (0.71–1.96)0.501.20 (0.71–1.99)0.48 RSV ICU admissionControlPCT10 (6.4%)6 (6.8%)1.10 (0.36–3.00)0.891.13 (0.37–3.20)0.82Mechanical ventilationControlPCT3 (1.9%)3 (3.4%)1.81 (0.33–9.98)0.472.28 (0.39–13.73)0.34NIV and HFNOControlPCT19 (12%)13 (15%)1.26 (0.58–2.67)0.551.29 (0.59–2.75)0.52ReadmissionControlPCT10 (6.4%)8 (9.1%)1.47 (0.54–3.87)0.441.40 (0.51–3.70)0.50Bold values indicate statistical significance.*Pearson’s χ² test.†Multivariate logistic regression using age and Charlson Comorbidity Index as covariates NIV.HFNO, high flow nasal cannula; NIV, non-invasive ventilation; RSV, respiratory syncytial virus. In hospital outcomes in procalcitonin (PCT) versus control group Bold values indicate statistical significance. Pearson’s χ² test. Multivariate logistic regression using age and Charlson Comorbidity Index as covariates NIV. HFNO, high flow nasal cannula; NIV, non-invasive ventilation; RSV, respiratory syncytial virus. Table 5Mortality in procalcitonin (PCT) versus control groupGroupsObserved, N (%)Unadjusted HR*P valueAdjusted HR†p value Influenza A/B 30-day mortalityControlPCT50 (8.5%)19 (6.3%)0.72 (0.43–1.22)0.230.76 (0.45–1.28)0.390-day mortalityControlPCT62 (10.5%)23 (7.6%)0.63 (0.20–1.96)0.430.67 (0.21–2.10)0.48 RSV 30-day mortalityControlPCT16 (10%)9 (10%)0.99 (0.44–2.25)0.991.00 (0.44–2.27)0.9990-day mortalityControlPCT20 (12.7%)10 (11.4%)0.44 (0.049 0 3.95)0.460.37 (0.04–3.54)0.39*Univariate cox proportional hazards model†Multivariate cox proportional hazards model with age and Charlson Comorbidity Index as covariatesRSV, respiratory syncytial virus. Mortality in procalcitonin (PCT) versus control group Univariate cox proportional hazards model Multivariate cox proportional hazards model with age and Charlson Comorbidity Index as covariates RSV, respiratory syncytial virus. Discussion This study evaluated the impact of PCT testing on antibiotic use in hospitalised patients with Influenza A/B and RSV infections. The primary findings revealed that while PCT testing was associated with variations in antibiotic prescribing based on PCT levels, it did not significantly reduce overall antibiotic use at admission or during hospitalisation after adjusting for infiltrate(s) on chest X-ray. Additionally, PCT use showed no impact on DDD or DOT prescribed. These results suggest that PCT testing alone may have limited utility in guiding antibiotic stewardship in proven influenza and RSV infections. Our results differ from previous research, which primarily examined broader populations with acute respiratory infections rather than focusing specifically on confirmed viral infections like influenza and RSV. A Cochrane review found that PCT-guided protocols reduced antibiotic use and duration in acute respiratory infections.33 Similarly, the ProHOSP trial showed an average reduction of 2.15 days of antibiotic use without harming patient safety.26 A recent meta-analysis supported these findings, showing that PCT protocols reduce antibiotic exposure while keeping patients safe.34 These differences in findings may be due to variations in study populations, adherence to protocols and the focus on proven viral infections in our study. For example, the ProACT trial found no reduction in antibiotic use in patients with respiratory infections presenting to the emergency department.35 These findings highlight how patient selection and following protocols affect outcomes. A French multicentre study showed that PCT might help identify bacterial co-infections in severe influenza pneumonia, but it did not investigate reduction in antibiotic use.36 Similarly, the EPIC study found that bacterial co-infections occurred in approximately 11% of patients hospitalised with influenza and RSV, with S. pneumoniae and Staphylococcus aureus being the most common pathogens.37 The study also noted that PCT levels were often elevated in patients with bacterial co-infections compared with those with viral infections alone. However, it did not directly assess the utility of PCT in guiding antibiotic use or its effectiveness as a stewardship tool in this context. These gaps show the need for more research into how PCT can guide antibiotic decisions in proven viral respiratory infections. This study is the first to examine the impact of PCT determination in adults with a proven RSV and influenza infection. The detailed dataset made it possible to account for confounding, improving the reliability of the findings. However, because the study is retrospective, it cannot establish cause and effect accurately. Residual confounding due to differences in baseline may have influenced the results and this should be considered when interpreting the findings. Decisions about prescribing antibiotics were based on clinician judgement, which might have led to inconsistencies in following PCT protocols, as seen in the number of antibiotic prescriptions in patients with low PCT levels (23% vs 26%). A potential source of bias in this study arises from the selective determination of PCT. PCT testing was more likely performed in patients perceived as more seriously ill, as PCT measurement may seem less necessary in milder cases with low suspicion of bacterial infection. High adherence to the PCT protocol might be essential for demonstrating reductions in antibiotic use, yet real-world clinical settings often pose challenges in achieving full adherence. Training is key to ensuring consistent and appropriate use of antibiotic stewardship interventions.38 Importantly, no active antimicrobial stewardship programme was in place during the study period, despite the availability of PCT testing and physicians’ awareness of the protocols. Previously, it has been shown that combining viral PCR results with low PCT values and real-time stewardship intervention can reduce antibiotic use.39 Exact times from admission to PCT testing and to first antibiotic prescription were not available due to the retrospective design. It is possible that in some cases, a singular gift of antibiotics was already given before PCT results became available, which may partly explain the lack of difference in antibiotic prescribing at admission between groups. In this retrospective study, we are unable to fully assess the impact of other tools used in antibiotic therapy management, such as CRP values. Microbiological yield was low in both groups, with positive blood cultures in ≤3.6% and positive sputum cultures in 10%–13% of influenza patients and 8.3%–25% of RSV patients (p=0.002). The higher sputum culture positivity in the RSV PCT group may have influenced antibiotic prescribing decisions. The single-centre setting may also limit the generalisability of the results to other hospitals. Future research should focus on prospective, multi-centre studies to better assess the effectiveness of PCT-guided protocols in diverse clinical settings. Additionally, combining PCT testing with other diagnostic tools or biomarkers may improve its utility in distinguishing pure viral from bacterial or combined viral-bacterial infections. To maximise its potential, PCT testing should be integrated into comprehensive, evidence-based stewardship programmes that include standardised protocols and clinician training. Conclusion This retrospective study showed that PCT testing influenced antibiotic prescribing based on PCT levels, but did not significantly reduce overall antibiotic use. While no higher rate of adverse outcomes was seen with PCT testing, in this setting, the present findings suggest that PCT is not an optimal tool to reduce antibiotic prescriptions in influenza and RSV infections. Supplementary material Footnotes Funding: This study received funding from Northwest Clinics (grant number FIO 2206). The funders had no role in the design and conduct of the study or analysis and interpretation of the data. Provenance and peer review: Not commissioned; externally peer-reviewed. Patient consent for publication: Not applicable. Ethics approval: This study involved human participants and was approved by the research ethics committee of Northwest Clinics (L020-115). As this was a retrospective study using anonymised clinical data, the requirement for informed consent was waived by the research committee. All data were handled confidentially, and patient privacy was protected throughout the research process (see statement above). Data availability free text: Data will be made available on request to researchers with a methodologically sound proposal beginning 3 months after publication until 5 years after publication. Proposals to gain access should be directed to lm.hessels@nwz.nl. Patient and public involvement: Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research. Data availability statement Data are available upon reasonable request. References Associated Data Supplementary Materials Data Availability Statement Data are available upon reasonable request.