Impact of multiplex PCR respiratory viral panel testing on antibiotic utilisation in children with acute febrile and respiratory illnesses Authors: Praveena Nediyara Bhaskaran, Merlin Moni, Dipu T T Sathyapalan, Soumya Jose, Ajai Krishna, Lima Rose, Asha Sudheer, Kiran Kishore, Veena Menon Published: 2025 ABSTRACT AbstractBackgroundThe utility of a PCR-based respiratory virus panel (RVP) in limiting antibiotic overuse in paediatric febrile and respiratory illnesses in low- and middle-income countries remains unclear.MethodsThis retrospective study included inpatients who underwent RVP testing within 48 hours of admission between January 2021 and December 2022. The primary outcome was the impact of the respiratory viral panel result on total duration of antibiotics. The secondary outcome was discontinuation of antibiotics following test results and duration of antibiotics in specific subgroups. A zero-inflated negative binomial (ZINB) regression identified predictors of outcome.ResultsOf the 210 cases, 152 (72.3%) tested positive for ≥1 virus. 42% of patients had their antibiotics discontinued after getting a positive result. Overall, we noted a minimal effect of a positive RVP result in reducing antibiotic utilisation. However, in influenza patients, we observed a significant decrease in median antibiotic days compared with those with negative RVP (3 days (IQR 0–5.5) vs 6 days (IQR 2.25–8); p=0.011). Patients with acute febrile illness with viral aetiology received fewer days of antibiotics compared with those with negative RVP (2.5 days (IQR 0–5) vs 5 days (IQR 2–7); p=0.035). ZINB regression analysis showed that pneumonia patients (OR=1.27; p=0.03) and those with paediatric intensive care unit stay (1.2; 0.02) had higher odds of prolonged antibiotic use despite RVP positivity.ConclusionIn our setting, RVP reduced antibiotic use in specific subgroups like influenza-positive patients and those with acute febrile illness and could be potentially used as an antibiotic stewardship tool in these settings.Keywords: Virology, Molecular Biology, Child Health WHAT IS ALREADY KNOWN ON THIS TOPIC WHAT THIS STUDY ADDS In a low- and middle-income setting, use of respiratory viral panel testing helped in decreasing antibiotic use among influenza-positive patients as well as those with acute febrile illnesses without a focus. In a low- and middle-income setting, use of respiratory viral panel testing helped in decreasing antibiotic use among influenza-positive patients as well as those with acute febrile illnesses without a focus. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY The implementation of narrower and more rapid diagnostic panels in non-severe respiratory infections could potentially enhance antibiotic stewardship. More extensive research with diverse populations is needed to thoroughly evaluate the effect of respiratory virus panel testing on antibiotic use and the overall cost of care. The implementation of narrower and more rapid diagnostic panels in non-severe respiratory infections could potentially enhance antibiotic stewardship. More extensive research with diverse populations is needed to thoroughly evaluate the effect of respiratory virus panel testing on antibiotic use and the overall cost of care. Introduction The Centre for Disease Dynamics, Economics and Policy, a leading global health research organisation, reports that approximately 58 000 children in India and 700 000 children globally die each year due to antibiotic-resistant infections.1 This alarming trend underscores the urgent need for effective solutions. In response, antibiotic stewardship programmes (ASPs) were established within healthcare settings to reduce the inappropriate use and misuse of antimicrobials. A robust diagnostic stewardship is essential to a hospital’s antimicrobial stewardship programme and involves using state-of-the-art diagnostics to ensure judicious use of antimicrobials.2 3 Acute respiratory infections (ARIs) are the most common cause of hospital admissions in paediatric inpatient units. ARI is one of the most common reasons for antibiotic prescriptions among hospitalised children.4 Differentiating viral from bacterial aetiology clinically can be challenging and this often leads to overuse of antibiotics. Traditionally, clinical features along with laboratory parameters, as well as findings on chest imaging, influence the clinicians’ decision regarding the initiation of antibiotics. Also, there are global differences in antibiotic prescribing practices, with low-income countries contributing significantly to overall antibiotic use.4,7 Newer molecular diagnostic tests, like the nucleic acid amplification tests done from the nasopharyngeal swabs, have high sensitivity and specificity in detecting respiratory pathogens. Syndromic respiratory viral panels (RVP) with multiplex PCR, which can detect multiple pathogens simultaneously, will help in identifying the common respiratory viruses within a short time.8 RVP is routinely done in emergency rooms and inpatient units in high-income countries. Studies that have investigated their direct impact on patient care and antibiotic prescriptions have shown mixed results. However, they are yet to be widely adopted in regular practice in low- and middle-income countries, and their utility in this setting has not been studied widely. Since 2019, syndromic testing of respiratory viruses using a multiplex PCR assay has been available in routine clinical practice in our hospital. Given the inconsistencies in global study results on the clinical utility of RVP testing in decreasing the overuse of antibiotics, we decided to assess the impact of PCR testing for respiratory pathogens on inpatient antibiotic treatment duration among children. Methods Study design and population The study design was in accordance with the Strengthening the Reporting of Observational studies in Epidemiology guidelines. Patients and the public were not involved in the design, conduct, reporting or dissemination of this research. This retrospective observational study was conducted in a tertiary care paediatric teaching hospital in South India. The study population included paediatric inpatients ≤18 years of age admitted with either ARI or acute febrile illness between January 2021 and December 2022, for whom the viral panel testing was ordered within 48 hours of admission. Those with positive bacterial cultures, patients who died during the admission, those with underlying immunosuppressive conditions and chronic lung disease, and patients positive for Mycoplasma pneumoniae who warranted antibiotic treatment were excluded. Study measures and outcomes The clinical data were collected from the hospital’s Electronic Health Records and through patient chart review. These data included demographic characteristics such as age, hospital admission and discharge dates, day of RVP testing, primary discharge diagnosis and PICU stay during hospitalisation. Laboratory test values and imaging results were not included in the data. Based on the chart review, the primary discharge diagnosis was categorised into acute upper respiratory infection, acute lower respiratory infection/pneumonia, asthma exacerbation, bronchiolitis and acute febrile illness. The primary objective was the impact of a positive viral panel report on the total duration of antibiotics. The secondary objective was the discontinuation of antibiotics within 24 hours of identification of the virus in the test, duration of antibiotics among various subgroups, including different age groups, specific clinical diagnosis, type of virus isolated and those with and without paediatric intensive care unit (PICU) stay. Since the study outcome focused solely on antibiotic prescribing, information regarding the administration of antivirals during the study period was not collected. RVP testing The patients in the study had RVP testing using the CE/IVD-approved FTD Respiratory Pathogens 21 kit by Siemens Healthineers, Germany. This multiplexed kit detects the presence of influenza A virus, influenza A virus H1N1 swine-lineage, influenza B virus, human rhinovirus, human coronaviruses 229E, NL63, HKU1 and OC43, human parainfluenza viruses 1 through 4, human metapneumoviruses A and B, human bocavirus, M. pneumoniae, human respiratory syncytial viruses A and B, human parechovirus, enterovirus and human adenovirus in respiratory specimens. The total nucleic acid was extracted from nasopharyngeal swab specimens using the QIAamp viral RNA mini kit (Qiagen, Hilden, Germany). Multiplex real-time PCR was performed per the manufacturer’s instructions on the ABI 7500 (Thermofisher, USA) PCR instrument. The test results were available and reported within 24–48 hours. Statistical analysis Based on the mean and SD of duration of antibiotics among viral positive (5.5±3.27) and negative (7.1±2.55) patients observed from the pilot study conducted in 10 samples in each group, and with 95% confidence and 80% power, the minimum sample size came to 53 per group, totalling 106. Descriptive statistics used in the study analysis included medians and IQR. Comparison of demographic and clinical characteristics was performed using the Student’s t-test or Mann-Whitney U test for continuous variables and the χ2 test to compare categorical variables; a two-tailed p value <0.05 was set as significant. Regression analysis was performed to model the primary outcome (days of antibiotics), using predictor variables like RVP results, age group, clinical diagnosis and PICU admission during the study period. Since the outcomes were count-based variables with overdispersion and excessive zeroes (19%), we used a zero-inflated negative binomial (ZINB) regression to model the effects. ZINB regression modelled the factors influencing the days of antibiotics (count model), and a logit model identified factors leading to stopping antibiotics (zero-inflated component). All statistical analysis was conducted using R V.4.2.1 (R Foundation for Statistical Computing, Vienna, Austria). Patient and public involvement Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research. Results Study description A total of 3371 paediatric patients (≤18 years old) were admitted with respiratory symptoms during the study period. Molecular-based syndromic testing for respiratory viruses was performed on nasopharyngeal swabs obtained from 246 patients. 36 patients were excluded from the analysis (figure 1). Among the 210 patients included in the study, approximately 65% were male, with a median age of 2.75 (IQR 1–7) years. Out of 210 patient samples tested, 152 (72%) tested positive for respiratory viruses (table 1). Among the RVP-positive cases, 23% tested positive for more than one virus. The clinical diagnoses of the study population included upper respiratory tract infections (n=23, 10.9%), pneumonia (n=76, 36.1%), bronchiolitis (n=37, 17.6%), asthma exacerbation (n=19, 9.04%) and acute febrile illness with no definite focus (n=55, 26.1%). The most frequently identified viruses were respiratory syncytial virus (RSV) (33%), rhinovirus (28%) and influenza (25%) (table 2). Flowchart showing patient selection. Table 1Study patient characteristics at the time of RVP testingVariablesTotal cohortRVP positive N (%)RVP negative N (%)P value210152 (100)58 (100)Age (median, IQR) in years3.89 (4.00)6.44 (5.25)Sex Female74 (35.3)54 (35.5)20 (34.5)1 Male136 (64.7)98 (64.5)38 (65.5)Age groups <2 (years)61 (40.1)11 (19)0.003714 2–5 (years)53 (34.9)17 (29.3) 5–12 (years)26 (17.1)18 (31) >12 (years)12 (7.9)12 (20.7)Diagnosis AFIs30 (19.7)25 (43.1)0.004843 Asthma16 (10.5)3 (5.2) Bronchiolitis31 (20.4)6 (10.3) Pneumonia55 (36.2)21 (36.2) UpRIs20 (13.2)3 (5.2)Viral aetiology Single infection117117 (77)0 (0) Coinfections3535 (23) PICU stay51 (33.6)13 (22.4)0.12AFI, acute febrile illness; PICU, paediatric intensive care unit; RVP, respiratory virus panel; UpRIs, upper respiratory illnesses. Study patient characteristics at the time of RVP testing AFI, acute febrile illness; PICU, paediatric intensive care unit; RVP, respiratory virus panel; UpRIs, upper respiratory illnesses. Table 2Frequently detected respiratory viruses in the study cohortS. No.Virus identifiedFrequency N (%)1Influenza A20 (13)2Influenza A H1N15 (3)3Influenza B13 (8.5)4Rhinovirus43 (28)5RSV50 (33)6Coronaviruses9 (6)7Parainfluenza virus22 (14.5)8Metapneumovirus2 (1.3)9Bocavirus10 (6.5)10Adenovirus17 (11.2)11Enterovirus7 (4.6)12Parechovirus1 (0.66) Frequently detected respiratory viruses in the study cohort Impact of RVP on antibiotic duration 82.2% of patients in the positive viral panel cohort and 79.3% in the negative cohort were initiated on antibiotics before testing. Of the 152 patients with a positive RVP test result, 64 (42%) had the antibiotics discontinued within 24 hours of the results. The median duration of antibiotic use among the viral panel-positive patients was 5 days (IQR 3 days–7 days) compared with 6 days (IQR 2.25 days–8 days) for those with a negative viral panel (p value=0.36) (figure 2A). However, a noticeable difference in the duration of antibiotics was observed in patients with a positive influenza test result. As indicated in figure 2B, the median duration of antibiotic treatment for influenza-positive patients was only 3 days (IQR 0 days–5.5 days) compared with those with a negative viral panel (p=0.011). A similar decreasing trend, though not significant, was also observed in patients with RSV infections. No significant difference in antibiotic use was noted between single and co-infection cases. RVP results/type of virus and antibiotic duration. (A) RVP results and antibiotic duration. (B) Type of virus isolated and antibiotic duration. RVP, respiratory virus panel. Next, we investigated whether the clinical diagnosis affected the days a patient received antibiotic treatment following the test results. As shown in figure 3A, for most clinical diagnoses, we did not observe any significant difference in the antibiotic treatment days; but in patients with acute febrile illness and a positive RVP test result, significantly fewer days of antibiotic treatment were noted (2.5 days (IQR 0–5) vs 5 days (IQR 2–7); p=0.035). The uneven distribution of bronchiolitis cases between the RVP positive and negative groups may explain the significantly longer duration of antibiotics in bronchiolitis cases with a positive RVP result. No difference in antibiotic duration was found when different age groups or those with and without PICU stay were compared (figure 3B,C). (A) Clinical diagnosis and antibiotic duration. (B) Age groups and antibiotic duration. (C) PICU stay and antibiotic duration. PICU, paediatric intensive care unit; RVP, respiratory virus panel. Multivariate analyses In the ZINB regression model (figure 4), examining predictors of antibiotic duration, several clinical and diagnostic variables were significantly associated with the number of antibiotic days (count component, figure 4A) and the likelihood of receiving no antibiotics (zero-inflated component, figure 4B). In the count model, which estimates predictors of longer antibiotic duration among patients receiving antibiotics, positive RVP results for influenza A and RSV were associated with significantly reduced antibiotic use, with ORs of 0.674 (CI 0.5 to 0.91; p=0.0096) and 0.749 (CI 0.57 to 0.98; p=0.038), respectively. Conversely, PICU stay and pneumonia diagnosis were associated with increased antibiotic duration, with ORs of 1.21 (CI 1.03 to 1.41; p=0.017) and 1.27 (CI 1.03 to 1.56; p=0.027), respectively. In the zero-inflated model, which estimates the likelihood of receiving no antibiotics, PICU admission and pneumonia diagnosis were both strongly associated with lower odds of being in the no-antibiotic group, with an OR of 0.139 (CI 0.04 to 0.54; p=0.004) and 0.142 (CI 0.04 to 0.49; p=0.002). Predictors of discontinuation of antibiotics. PICU, paediatric intensive care unit; RSV, respiratory syncytial virus. These findings suggest that children with positive RVP results for influenza A or RSV received shorter courses of antibiotics. In contrast, children with indicators of more severe illness, particularly those diagnosed with pneumonia or admitted to the PICU, were more likely to have prolonged durations of antibiotic treatment. Discussion Even though antibiotics were discontinued in 42% of patients after viral identification in the multiplex panel, the overall mean duration of antibiotics failed to show a statistically significant difference between the viral panel positive and negative patients. But importantly, in the subgroup of patients with a positive influenza result as well as acute febrile illness without respiratory symptoms, we noticed that there was a significant reduction in the duration of antibiotics. Childhood respiratory infections are predominantly viral. In our study population, 72.3% had one or more viruses isolated. In a multi-centre population-based surveillance study on community-acquired pneumonia requiring hospitalisation among US children <18 years, in which both viral and bacterial testing were done, virus detection was confirmed in 66%, bacterial detection in 8% and viral/bacterial co-detection in 7%.9 Another study found viral aetiology in 84.4% of 1353 hospitalised children with acute respiratory tract infection in southern Germany.10 Whether virus isolation from the nasopharyngeal swab has a causal attribution to the respiratory illness has been addressed in previous studies. In a systematic review of studies investigating the viral profile of under 5 children with and without Acute lower respiratory infections (ALRI), there was strong evidence for causal attribution of RSV, influenza virus, parainfluenza virus and metapneumovirus and less strong evidence for rhinovirus.11 Viral upper respiratory infections can be complicated by bacterial otitis media, and viral pneumonias can be complicated by bacterial coinfection or superinfection. The true burden of this coinfection can be hard to determine. In a single-centre retrospective study to determine bacterial coinfection among children admitted with respiratory infection, the authors found that only 3.5% of the children with respiratory virus isolation had a coinfection with bacteria, as determined by respiratory specimen and blood culture.12 But this rate might be higher in pneumonia. According to The Lancet review article on viral pneumonia, the authors found that evidence of probable mixed viral-bacterial infection was seen in up to 45% of cases of community-acquired pneumonia in children.13 There is limited data on coinfection rates from the low- and middle-income countries. This lack of data, along with the potential of bacterial coinfection, could be a reason for the continuation of the antibiotics even after the identification of viral aetiology. Many studies among children have investigated the utility of respiratory viral panel testing in determining antibiotic usage in different settings, with varying results. In a retrospective chart review of hospitalised children in a community hospital in the USA who underwent rapid diagnostic viral testing using direct fluorescence assay, positive patients had fewer days of both oral and intravenous antibiotics. Of note, pneumonia patients constituted only 12% of this cohort,14 whereas in our study, pneumonia patients constituted 36% in each group, which could have contributed to the continuation of antibiotics. In another retrospective single-centre study from a mid-sized children’s hospital in the USA, the authors found that, in uncomplicated respiratory tract infections, patients with positive viral panel reports had fewer days of antibiotics. Patients requiring ICU stay and those with complicated infections were excluded from the study.15 In our study, 34% of patients in the RVP+ve group required PICU stay. The PICU stay was one of the independent factors associated with an increase in antibiotic duration in the modelling. Studies looking at the utility of multiplex PCR panels specifically in PICU populations have shown limited utility in reducing antibiotic duration.16 17 We noticed a significant reduction in antibiotic duration in our patients who were influenza-positive compared with RVP-negative patients, even with the turnaround time (TAT) of 24–48 hours. Some studies from low- and middle-income countries have shown decreased antibiotic prescription rates when rapid influenza antigen tests were positive compared with negative children.18 19 A shorter turnaround time of the test has also been found to be associated with a reduction in antibiotic use.20 Similarly, in acute febrile illnesses without any respiratory symptoms, identification of virus prompted treating physicians to discontinue antibiotics and hence a significant reduction in antibiotic duration was found. These findings can be used in antibiotic stewardship in low-income countries with a high baseline antibiotic prescription rate. In a review article on the utility of RVP as an antimicrobial stewardship tool among predominantly adult patients, the authors found that even though RVP alone had limited value, when used with procalcitonin, it could reduce antibiotic use and duration.21 Active antimicrobial stewardship along with RVP also found to have an impact on the reduction of antibiotic usage.22 In our patient cohort, several factors may have influenced the continuation of antibiotics despite the identification of a viral pathogen. The antibiotic initiation rate itself was very high, which reflects the prescription behaviour among the physicians. Another contributing reason may have been delays in the communication of RVP test results, particularly over weekends when results were typically reviewed 24–48 hours after sample collection. This was partly due to the hospital laboratory’s use of batch testing protocols implemented for cost-effectiveness. Lack of promptness in informing the treating physician of viral positivity may have significantly influenced the clinical decision to continue antibiotic use even when not warranted. Such delays may have limited the ability of clinicians to incorporate the viral positivity results into early antibiotic de-escalation. The study found that 36% of the RVP+ve patients had pneumonia, while 34% required PICU stay. We found that a diagnosis of pneumonia and PICU stay were independently associated with prolonged antibiotic therapy, regardless of RVP positivity. Finally, we have not taken into consideration the level of inflammatory markers like white cell count, C reactive protein or procalcitonin. High levels of inflammatory markers or findings on imaging could have influenced the clinician’s decision to initiate and continue antibiotics. Nevertheless, this study has a few limitations; first, as a retrospective study conducted at a single tertiary care centre, there is a risk of selection bias, which may limit the generalisability of the findings to other populations or settings. Second, we did not consider the role of chest radiographs or biomarkers like CRP while evaluating the diagnostic utility of RVP testing on antibiotic use, ie, specifically, this study interprets the effect of using RVP testing alone as a diagnostic tool in clinical decision-making regarding antibiotic utilisation. Most studies that have reported a positive impact of RVPs on antibiotic utilisation have employed RVP testing in conjunction with other investigations. We did not evaluate how TAT affected the antibiotic duration. Finally, we also did not assess the utilisation of antivirals as a response to RVP test results. Most studies on the utility of RVP are conducted in high-income countries. The cost of viral panel testing can be a barrier to routine use in resource-limited settings. More research is needed to evaluate the cost-effectiveness of respiratory viral panel testing regarding its ability to reduce antibiotic use and hospital stays. To overcome these cost barriers, stewardship efforts should develop protocols that ensure timely results and promote judicious use of antibiotics in high-acuity settings like PICU when viral aetiology is confirmed. Conclusion Molecular-based syndromic testing for respiratory viruses in children in low-middle income settings reduced the antibiotic duration in certain subgroups of patients. The implementation of narrower and more rapid diagnostic panels in non-severe respiratory infections and acute febrile illnesses could potentially benefit institutional ASP. More extensive research with diverse populations among rural and urban settings in the country is needed to thoroughly evaluate the effect of RVP testing on antibiotic utilisation and overall cost of care. Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Patient consent for publication: Not applicable. Provenance and peer review: Not commissioned; externally peer reviewed. Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Ethics approval: The study was approved by the Ethics Committee of Amrita School of Medicine (ID: ECASM-AIMS-2024-416), and individual patient consent was waived. Data availability statement Data are available upon reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information. References Data Availability Statement Data are available upon reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information.