Development and validation of a clinical prediction rule to distinguish bacterial from viral pneumonia in children

Prevalence and predictors of radiographic pneumonia in children with wheeze: A systematic review and meta-analysis

BACKGROUND

Diagnostic uncertainty exists surrounding the identification of radiographic pneumonia in children with wheeze. It is important to determine the prevalence and clinical predictors of pneumonia in this population to limit chest radiography (CXR) and promote judicious antibiotic use.

OBJECTIVES

The objectives were to (1) estimate the prevalence of radiographic pneumonia in children with wheeze and (2) systematically review the diagnostic accuracy of clinical findings for the identification of radiographic pneumonia.

METHODS

Data sources were MEDLINE, PubMed Central, Cochrane Library, CINAHL, and Web of Science (January 1995 to September 2023). For study selection, two reviewers identified high-quality studies reporting on clinical characteristics associated with radiographic pneumonia in wheezing children (age 0-21 years). Using Covidence software, data regarding study characteristics, methodologic quality, and results were extracted. Data were pooled using random-effects meta-analysis.

RESULTS

A total of 8333 unique titles and abstracts were reviewed. Twelve studies, representing 7398 patients, were included. Fifteen percent of children with wheeze undergoing CXR had pneumonia. Findings associated with radiographic pneumonia included temperature ≥ 38.4°C (positive likelihood ratio [LR+] 2.1, 95% CI 1.2-3.6, specificity 85%), oxygen saturation < 92% (LR+ 3.6, 95% CI 1.4-8.9, specificity 89%), and grunting (LR+ 2.7, 95% CI 1.6-4.4, pooled specificity 91%). Factors associated with the absence of radiographic pneumonia included lack of fever (negative likelihood ratio [LR-] 0.67, 95% CI 0.52-0.85) and oxygen saturation ≥ 95% (LR- 0.64, 95% CI 0.42-0.98). Tachypnea and auscultatory findings were not associated with radiographic pneumonia.

DISCUSSION

Heterogeneity across studies limits generalizability. Additionally, all included studies overestimate the rate of radiographic pneumonia given the fact that all subjects had a CXR performed due to clinical suspicion of pneumonia.

CONCLUSIONS

Radiographic pneumonia occurs in 15% of wheezing children undergoing CXR for pneumonia. Auscultatory findings and tachypnea do not differentiate children with and without pneumonia, and the rate of radiographic pneumonia is very low in the absence of fever and hypoxemia.

Association of Radiology Findings with Etiology of Community Acquired Pneumonia among Children

OBJECTIVE

To evaluate the association between consolidation on chest radiograph and typical bacterial etiology of childhood community acquired pneumonia (CAP) in the Etiology of Pneumonia in the Community study.

STUDY DESIGN

Hospitalized children <18 years of age with CAP enrolled in the Etiology of Pneumonia in the Community study at 3 children's hospitals between January 2010 and June 2012 were included. Testing of blood and respiratory specimens used multiple modalities to identify typical and atypical bacterial, or viral infection. Study radiologists classified chest radiographs (consolidation, other infiltrates [interstitial and/or alveolar], pleural effusion) using modified World Health Organization pneumonia criteria. Infiltrate patterns were compared according to etiology of CAP.

RESULTS

Among 2212 children, there were 1302 (59%) with consolidation with or without other infiltrates, 910 (41%) with other infiltrates, and 296 (13%) with pleural effusion. In 1795 children, at least 1 pathogen was detected. Among these patients, consolidation (74%) was the most frequently observed pattern (74% in typical bacterial CAP, 58% in atypical bacterial CAP, and 54% in viral CAP). Positive and negative predictive values of consolidation for typical bacterial CAP were 12% (95% CI 10%-15%) and 96% (95% CI 95%-97%) respectively. In a multivariable model, typical bacterial CAP was associated with pleural effusion (OR 7.3, 95% CI 4.7-11.2) and white blood cell ≥15 000/mL (OR 3.2, 95% CI 2.2-4.9), and absence of wheeze (OR 0.5, 95% CI 0.3-0.8) or viral detection (OR 0.2, 95% CI 0.1-0.4).

CONCLUSIONS

Consolidation predicted typical bacterial CAP poorly, but its absence made typical bacterial CAP unlikely. Pleural effusion was the best predictor of typical bacterial infection, but too uncommon to aid etiology prediction.

Clinical algorithm reduces antibiotic use among children presenting with respiratory symptoms to hospital in central Vietnam

OBJECTIVE

To assess the safety and utility of a pragmatic clinical algorithm to guide rational antibiotic use in children presenting with respiratory infection.

METHODS

The effect of an algorithm to guide the management of young (< 5 years) children presenting with respiratory symptoms to the Da Nang Hospital for Women and Children, Vietnam, was evaluated in a before-after intervention analysis. The main outcome was reduction in antibiotic use, with monitoring of potential harm resulting from reduced antibiotic use. The intervention comprised a single training session of physicians in the use of an algorithm informed by local evidence; developed during a previous prospective observational study. The evaluation was performed one month after the training.

RESULTS

Of the 1290 children evaluated before the intervention, 102 (7.9%) were admitted to hospital and 556/1188 (46.8%) were sent home with antibiotics. Due to COVID-19, only 166 children were evaluated after the intervention of whom 14 (8.4%) were admitted to hospital and 54/152 (35.5%) were sent home with antibiotics. Antibiotic use was reduced (from 46.8% to 35.5%; p = 0.009) after clinician training, but adequate comparison was compromised. The reduction was most pronounced in children with wheeze or runny nose and no fever, or a normal chest radiograph, where antibiotic use declined from 46.7% to 28.8% (p < 0.0001). The frequency of repeat presentation to hospital was similar between the two study periods (141/1188; 11.9% before and 10/152; 6.6% after; p = 0.10). No child represented with serious disease after being sent home without antibiotics.

CONCLUSIONS

We observed a reduction in antibiotic use in young children with a respiratory infection after physician training in the use of a simple evidence-based management algorithm. However, the study was severely impacted by COVID-19 restrictions, requiring further evaluation to confirm the observed effect.

Diagnostic models predicting paediatric viral acute respiratory infections: a systematic review

OBJECTIVES

To systematically review and evaluate diagnostic models used to predict viral acute respiratory infections (ARIs) in children.

DESIGN

Systematic review.

DATA SOURCES

PubMed and Embase were searched from 1 January 1975 to 3 February 2022.

ELIGIBILITY CRITERIA

We included diagnostic models predicting viral ARIs in children (<18 years) who sought medical attention from a healthcare setting and were written in English. Prediction model studies specific to SARS-CoV-2, COVID-19 or multisystem inflammatory syndrome in children were excluded.

DATA EXTRACTION AND SYNTHESIS

Study screening, data extraction and quality assessment were performed by two independent reviewers. Study characteristics, including population, methods and results, were extracted and evaluated for bias and applicability using the Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies and PROBAST (Prediction model Risk Of Bias Assessment Tool).

RESULTS

Of 7049 unique studies screened, 196 underwent full text review and 18 were included. The most common outcome was viral-specific influenza (n=7; 58%). Internal validation was performed in 8 studies (44%), 10 studies (56%) reported discrimination measures, 4 studies (22%) reported calibration measures and none performed external validation. According to PROBAST, a high risk of bias was identified in the analytic aspects in all studies. However, the existing studies had minimal bias concerns related to the study populations, inclusion and modelling of predictors, and outcome ascertainment.

CONCLUSIONS

Diagnostic prediction can aid clinicians in aetiological diagnoses of viral ARIs. External validation should be performed on rigorously internally validated models with populations intended for model application.

PROSPERO REGISTRATION NUMBER

CRD42022308917.

Predicting the causative pathogen among children with pneumonia using a causal Bayesian network

BACKGROUND

Pneumonia remains a leading cause of hospitalization and death among young children worldwide, and the diagnostic challenge of differentiating bacterial from non-bacterial pneumonia is the main driver of antibiotic use for treating pneumonia in children. Causal Bayesian networks (BNs) serve as powerful tools for this problem as they provide clear maps of probabilistic relationships between variables and produce results in an explainable way by incorporating both domain expert knowledge and numerical data.

METHODS

We used domain expert knowledge and data in combination and iteratively, to construct, parameterise and validate a causal BN to predict causative pathogens for childhood pneumonia. Expert knowledge elicitation occurred through a series of group workshops, surveys and one-on-one meetings involving 6-8 experts from diverse domain areas. The model performance was evaluated based on both quantitative metrics and qualitative expert validation. Sensitivity analyses were conducted to investigate how the target output is influenced by varying key assumptions of a particularly high degree of uncertainty around data or domain expert knowledge.

RESULTS

Designed to apply to a cohort of children with X-ray confirmed pneumonia who presented to a tertiary paediatric hospital in Australia, the resulting BN offers explainable and quantitative predictions on a range of variables of interest, including the diagnosis of bacterial pneumonia, detection of respiratory pathogens in the nasopharynx, and the clinical phenotype of a pneumonia episode. Satisfactory numeric performance has been achieved including an area under the receiver operating characteristic curve of 0.8 in predicting clinically-confirmed bacterial pneumonia with sensitivity 88% and specificity 66% given certain input scenarios (i.e., information that is available and entered into the model) and trade-off preferences (i.e., relative weightings of the consequences of false positive versus false negative predictions). We specifically highlight that a desirable model output threshold for practical use is very dependent upon different input scenarios and trade-off preferences. Three commonly encountered scenarios were presented to demonstrate the potential usefulness of the BN outputs in various clinical pictures.

CONCLUSIONS

To our knowledge, this is the first causal model developed to help determine the causative pathogen for paediatric pneumonia. We have shown how the method works and how it would help decision making on the use of antibiotics, providing insight into how computational model predictions may be translated to actionable decisions in practice. We discussed key next steps including external validation, adaptation and implementation. Our model framework and the methodological approach can be adapted beyond our context to broad respiratory infections and geographical and healthcare settings.

The global burden of hospitalisation due to pneumonia caused by Staphylococcus aureus in the under-5 years children: A systematic review and meta-analysis

BACKGROUND

Pneumonia is a leading cause of childhood morbidity and mortality. This study aimed to estimate the global hospitalisation due to Staphylococcus aureus pneumonia in under-5 children.

METHODS

We conducted a systematic review and meta-analysis of primary studies following the PRISMA-P guidelines. We searched Medline, Embase, Global Health, CINAHL, Global Index Medicus, Scopus, China National Knowledge Infrastructure, Wanfang, and CQvip. We included studies reporting data on Staphylococcus aureus pneumonia, confirmed by detection of the pathogen in sterile-site samples in under-5 hospitalised children, published in English or Chinese language and conducted between 1st January 1990 and 4th November 2021 and between 1st January 1990 and 30th September 2020, respectively. We excluded those testing upper respiratory tract samples and not reporting data on samples with other bacteria or absence of bacteria. We screened papers against pre-specified criteria, extracted data and assessed the bacteriological quality, and combined epidemiological and microbiological quality of studies using two self-designed checklists. Pooled proportions of hospitalisation episodes for Staphylococcus aureus pneumonia amongst all-cause pneumonia and the 95% confidence intervals were calculated using the random-effects model. The review protocol was registered on PROSPERO (CRD42021236606).

FINDINGS

Of 26,218 studies identified, thirty-five studies enroling 20,708 hospitalised pneumonia episodes were included. Out of the total hospitalised pneumonia cases in this population, the pooled proportion of Staphylococcal pneumonia cases was 3% (95% CI 2% to 4%; I²=96%). amongst 12 studies with higher microbiological quality, the pooled estimate was 6% (95% CI 2% to 10%; I²= 98%). Based on the recent global estimates of hospitalised pneumonia in this age group, the 3% and 6% estimates represent 738 thousand and 1.48 million hospitalisations in 2019, respectively. Based on the Grading of Recommendations Assessment, Development, and Evaluation (GRADE), the overall quality of evidence was considered to be moderate.

INTERPRETATION

Our findings are probably an underestimate because of the unknown and the likely limited sensitivity of current testing methods for Staphylococcal pneumonia diagnosis and widespread reported use of antibiotics before recruitment (in 46% of cases). Staphylococcus aureus is an important cause of hospitalisation for pneumonia in young children globally.

FUNDING

Bill and Melinda Gates Foundation (OPP 1,172,551) through a prime award to John Hopkins University.

A Tool to Distinguish Viral From Bacterial Pneumonia

BACKGROUND

Establishing the etiology of community-acquired pneumonia (CAP) in children at admission is challenging. Most of the admitted children with CAP receive antibiotics. We aimed to build and validate a diagnostic tool combining clinical, analytical and radiographic features to differentiate viral from bacterial CAP, and among bacterial CAP, typical from atypical bacteria.

METHODS

Design-observational, multi-center, prospective cohort study was conducted in 2 phases. Settings: 24 secondary and tertiary hospitals in Spain. Patients-A total of 495 consecutive hospitalized children between 1 month and 16 years of age with CAP were enrolled. Interventions-A score with 2 sequential steps was built (training set, 70% patients, and validation set 30%). Step 1 differentiates between viral and bacterial CAP and step 2 between typical and atypical bacterial CAP. Optimal cutoff points were selected to maximize specificity setting a high sensitivity (80%). Weights of each variable were calculated with a multivariable logistic regression. Main outcome measures-Viral or bacterial etiology.

RESULTS

In total, 262 (53%) children (median age: 2 years, 52.3% male) had an etiologic diagnosis. In step 1, bacterial CAPs were classified with a sensitivity = 97%, a specificity = 48%, and a ROC's area under the curve = 0.81. If a patient with CAP was classified as bacterial, he/she was assessed with step 2. Typical bacteria were classified with a sensitivity = 100%, a specificity = 64% and area under the curve = 0.90. We implemented the score into a mobile app named Pneumonia Etiology Predictor, freely available at usual app stores, that provides the probability of each etiology.

CONCLUSIONS

This 2-steps tool can facilitate the physician's decision to prescribe antibiotics without compromising patient safety.

Development of the Novel Pneumonia Risk Score to Predict Radiographic Pneumonia in Children

BACKGROUND

The diagnosis of pneumonia in children is challenging, given the wide overlap of many of the symptoms and physical examination findings with other common respiratory illnesses. We sought to derive and validate the novel Pneumonia Risk Score (PRS), a clinical tool utilizing signs and symptoms available to clinicians to determine a child's risk of radiographic pneumonia.

METHODS

We prospectively enrolled children 3 months to 18 years in whom a chest radiograph (CXR) was obtained in the emergency department to evaluate for pneumonia. Before CXR, we collected information regarding symptoms, physical examination findings, and the physician-estimated probability of radiographic pneumonia. Logistic regression was used to predict the presence of radiographic pneumonia, and the PRS was validated in a distinct cohort of children with suspected pneumonia.

RESULTS

Among 1181 children included in the study, 206 (17%) had radiographic pneumonia. The PRS included age in years, triage oxygen saturation, presence of fever, presence of rales, and presence of wheeze. The area under the curve (AUC) of the PRS was 0.71 (95% confidence interval [CI]: 0.68-0.75), while the AUC of clinician judgment was 0.61 (95% CI: 0.56-0.66) (P < 0.001). Among 2132 children included in the validation cohort, the PRS demonstrated an AUC of 0.69 (95% CI: 0.65-0.73).

CONCLUSIONS

In children with suspected pneumonia, the PRS is superior to clinician judgment in predicting the presence of radiographic pneumonia. Use of the PRS may help efforts to support the judicious use of antibiotics and chest radiography among children with suspected pneumonia.

Development of a risk score to guide targeted hepatitis C testing among human immunodeficiency virus patients in Cambodia

BACKGROUND

The World Health Organization recommends testing all human immunodeficiency virus (HIV) patients for hepatitis C virus (HCV). In resource-constrained contexts with low-to-intermediate HCV prevalence among HIV patients, as in Cambodia, targeted testing is, in the short-term, potentially more feasible and cost-effective.

AIM

To develop a clinical prediction score (CPS) to risk-stratify HIV patients for HCV coinfection (HCV RNA detected), and derive a decision rule to guide prioritization of HCV testing in settings where ‘testing all’ is not feasible or unaffordable in the short term.

METHODS

We used data of a cross-sectional HCV diagnostic study in the HIV cohort of Sihanouk Hospital Center of Hope in Phnom Penh. Key populations were very rare in this cohort. Score development relied on the Spiegelhalter and Knill-Jones method. Predictors with an adjusted likelihood ratio ≥ 1.5 or ≤ 0.67 were retained, transformed to natural logarithms, and rounded to integers as score items. CPS performance was evaluated by the area-under-the-ROC curve (AUROC) with 95% confidence intervals (CI), and diagnostic accuracy at the different cut-offs. For the decision rule, HCV coinfection probability ≥1% was agreed as test-threshold.

RESULTS

Among the 3045 enrolled HIV patients, 106 had an HCV coinfection. Of the 11 candidate predictors (from history-taking, laboratory testing), seven had an adjusted likelihood ratio ≥ 1.5 or ≤ 0.67: ≥ 50 years (+1 point), diabetes mellitus (+1), partner/household member with liver disease (+1), generalized pruritus (+1), platelets < 200 × 10⁹/L (+1), aspartate transaminase (AST) < 30 IU/L (-1), AST-to-platelet ratio index (APRI) ≥ 0.45 (+1), and APRI < 0.45 (-1). The AUROC was 0.84 (95%CI: 0.80-0.89), indicating good discrimination of HCV/HIV coinfection and HIV mono-infection. The CPS result ≥0 best fits the test-threshold (negative predictive value: 99.2%, 95%CI: 98.8-99.6). Applying this threshold, 30% (n = 926) would be tested. Sixteen coinfections (15%) would have been missed, none with advanced fibrosis.

CONCLUSION

The CPS performed well in the derivation cohort, and bears potential for other contexts of low-to-intermediate prevalence and little onward risk of transmission(i.e. cohorts without major risk factors as injecting drug use, men having sex with men), and where available resources do not allow to test all HIV patients as recommended by WHO. However, the score requires external validation in other patient cohorts before any wider use can be considered.

Hospital utilization rates for influenza and RSV: a novel approach and critical assessment

BACKGROUND

Influenza and respiratory syncytial virus (RSV) contribute significantly to the burden of acute lower respiratory infection (ALRI) inpatient care, but heterogeneous coding practices and availability of inpatient data make it difficult to estimate global hospital utilization for either disease based on coded diagnoses alone.

METHODS

This study estimates rates of influenza and RSV hospitalization by calculating the proportion of ALRI due to influenza and RSV and applying this proportion to inpatient admissions with ALRI coded as primary diagnosis. Proportions of ALRI attributed to influenza and RSV were extracted from a meta-analysis of 360 total sources describing inpatient hospital admissions which were input to a Bayesian mixed effects model over age with random effects over location. Results of this model were applied to inpatient admission datasets for 44 countries to produce rates of hospital utilization for influenza and RSV respectively, and rates were compared to raw coded admissions for each disease.

RESULTS

For most age groups, these methods estimated a higher national admission rate than the rate of directly coded influenza or RSV admissions in the same inpatient sources. In many inpatient sources, International Classification of Disease (ICD) coding detail was insufficient to estimate RSV burden directly. The influenza inpatient burden estimates in older adults appear to be substantially underestimated using this method on primary diagnoses alone. Application of the mixed effects model reduced heterogeneity between countries in influenza and RSV which was biased by coding practices and between-country variation.

CONCLUSIONS

This new method presents the opportunity of estimating hospital utilization rates for influenza and RSV using a wide range of clinical databases. Estimates generally seem promising for influenza and RSV associated hospitalization, but influenza estimates from primary diagnosis seem highly underestimated among older adults. Considerable heterogeneity remains between countries in ALRI coding (i.e., primary vs non-primary cause), and in the age profile of proportion positive for influenza and RSV across studies. While this analysis is interesting because of its wide data utilization and applicability in locations without laboratory-confirmed admission data, understanding the sources of variability and data quality will be essential in future applications of these methods.

Predictors of Unlikely Bacterial Pneumonia and Adverse Pneumonia Outcome in Children Admitted to a Hospital in Central Vietnam

BACKGROUND

Pneumonia is the leading cause of antibiotic use and hospitalization in Vietnam. There is a need for better prediction of unlikely bacterial pneumonia and adverse pneumonia outcome in order to guide hospital admission and improve rational antibiotic use.

METHODS

All children under 5 admitted with pneumonia (per clinician assessment) to the Da Nang Hospital for Women and Children were prospectively enrolled. Children were classified as having likely or unlikely bacterial pneumonia and followed for outcome assessment. A Bayesian model averaging approach was used to identify predictors of unlikely bacterial pneumonia and adverse pneumonia outcome, which guided the development of a pragmatic management algorithm.

RESULTS

Of 3817 patients assessed, 2199 (57.6%) met World Health Organization (WHO) pneumonia criteria. In total, 1594 (41.7%) children were classified as having unlikely and 129 (3.4%) as having likely bacterial pneumonia. The remainder (2399; 62.9%) were considered to have disease of uncertain etiology. Factors predictive of unlikely bacterial pneumonia were no fever, no consolidation on chest radiograph, and absolute neutrophil count <5 × 109/L at presentation, which had a negative predictive value (NPV) for likely bacterial pneumonia of 99.0%. Among those who met WHO pneumonia criteria, 8.6% (189/2199) experienced an adverse outcome. Not having any WHO danger sign or consolidation on chest radiograph had an NPV of 96.8% for adverse pneumonia outcome.

CONCLUSIONS

An algorithm that screens for predictors of likely bacterial pneumonia and adverse pneumonia outcome could reduce unnecessary antibiotic use and hospital admission, but its clinical utility requires validation in a prospective study.

Bacterial pneumonia and its associated factors in children from a developing country: A prospective cohort study

Introduction

Pneumonia in children is a common disease yet determining its aetiology remains elusive.

Objectives

To determine the a) aetiology, b) factors associated with bacterial pneumonia and c) association between co-infections (bacteria + virus) and severity of disease, in children admitted with severe pneumonia.

Methods

A prospective cohort study involving children aged 1-month to 5-years admitted with very severe pneumonia, as per the WHO definition, over 2 years. Induced sputum and blood obtained within 24 hrs of admission were examined via PCR, immunofluorescence and culture to detect 17 bacteria/viruses. A designated radiologist read the chest radiographs.

Results

Three hundred patients with a mean (SD) age of 14 (±15) months old were recruited. Significant pathogens were detected in 62% of patients (n = 186). Viruses alone were detected in 23.7% (n = 71) with rhinovirus (31%), human metapneumovirus (HMP) [22.5%] and respiratory syncytial virus (RSV) [16.9%] being the commonest. Bacteria alone was detected in 25% (n = 75) with Haemophilus influenzae (29.3%), Staphylococcus aureus (24%) and Streptococcus pneumoniae (22.7%) being the commonest. Co-infections were seen in 13.3% (n = 40) of patients. Male gender (AdjOR 1.84 [95% CI 1.10, 3.05]) and presence of crepitations (AdjOR 2.27 [95% CI 1.12, 4.60]) were associated with bacterial infection. C-reactive protein (CRP) [p = 0.007]) was significantly higher in patients with co-infections but duration of hospitalization (p = 0.77) and requirement for supplemental respiratory support (p = 0.26) were not associated with co-infection.

Conclusions

Bacteria remain an important cause of very severe pneumonia in developing countries with one in four children admitted isolating bacteria alone. Male gender and presence of crepitations were significantly associated with bacterial aetiology. Co-infection was associated with a higher CRP but no other parameters of severe clinical illness.

Clinical Factors Associated With Chest Imaging Findings in Hospitalized Infants With Bronchiolitis

Despite recommendations against routine imaging, chest radiography (CXR) is frequently performed on infants hospitalized for bronchiolitis. We conducted a review of 811 infants hospitalized for bronchiolitis to identify clinical factors associated with imaging findings. CXR was performed on 553 (68%) infants either on presentation or during hospitalization; 466 readings (84%) were normal or consistent with viral illness. Clinical factors significantly associated with normal/viral imaging were normal temperature (odds ratio = 1.66; 95% CI = 1.03-2.67) and normal oxygen saturation (odds ratio = 1.77; 95% CI = 1.1-2.83) on presentation. Afebrile patients with normal oxygen saturations were nearly 3 times as likely to have a normal/viral CXR as patients with both fever and hypoxia. Our findings support the limited role of radiography in the evaluation of hospitalized infants with bronchiolitis, especially patients without fever or hypoxia.

Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration

The TRIPOD (Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis) Statement includes a 22-item checklist, which aims to improve the reporting of studies developing, validating, or updating a prediction model, whether for diagnostic or prognostic purposes. The TRIPOD Statement aims to improve the transparency of the reporting of a prediction model study regardless of the study methods used. This explanation and elaboration document describes the rationale; clarifies the meaning of each item; and discusses why transparent reporting is important, with a view to assessing risk of bias and clinical usefulness of the prediction model. Each checklist item of the TRIPOD Statement is explained in detail and accompanied by published examples of good reporting. The document also provides a valuable reference of issues to consider when designing, conducting, and analyzing prediction model studies. To aid the editorial process and help peer reviewers and, ultimately, readers and systematic reviewers of prediction model studies, it is recommended that authors include a completed checklist in their submission. The TRIPOD checklist can also be downloaded from www.tripod-statement.org.

Clinical risk factors for life-threatening lower respiratory tract infections in children: a retrospective study in an urban city in Malaysia

AIM

Lower respiratory tract infections (LRTIs) are an important cause of morbidity and mortality, especially in low income countries. The aim of this study was to determine risk factors of life-threatening LRTIs in hospitalised children in Malaysia.

METHODS

This retrospective study included children aged less than 18 years admitted for LRTIs over 13 months in a tertiary referral centre in Kuala Lumpur, Malaysia. Neonates, children with asthma and those with either no or a normal chest radiograph were excluded. Life-threatening infection was defined as that needing non-invasive ventilation or admission to the paediatric intensive care unit. Routine blood investigations and nasopharyngeal secretion results (bacterial and viral) were obtained. Chest radiographs were reviewed by a designated radiologist. Environmental data (rainfall, particulate matter ≤ 10 µm [PM10] and air pollution index [API]) was obtained from the respective government departments.

RESULTS

Three hundred and ninety-one episodes of LRTIs were included. Viruses were implicated in 48.5% of LRTIs, with respiratory syncytial virus (RSV) being detected in 44% of viral LRTIs. Forty-six (11.8%) children had life-threatening disease and the overall mortality rate was 1.3% (5 children). RSV was detected in 26% of children with life-threatening LRTIs. In multivariate logistic regression, chronic lung disease, presenting history of apnoea and signs of hypoxia, was associated with life threatening LRTIs. Increased LRTI admissions were associated with low rainfall but not PM10 nor API. Of those on follow-up, 39% had persistent respiratory symptoms.

CONCLUSION

One in nine children admitted with LRTI had a life-threatening LRTI. The aetiology was viral in almost half of admitted children. RSV was detected in a quarter of children with life-threatening LRTIs. Children who present with LRTIs and either have chronic lung disease, presenting history of apnoea or signs of hypoxia, should be observed carefully as the risk of deterioration to life-threatening illness is high.

Impact assessment of a decision rule for using antibiotics in pneumonia: a randomized trial

BACKGROUND

Although more than half of the cases of pneumonia in children can be due to virus, most of them receive antibiotic treatment. A previously published clinical prediction rule bacterial pneumonia score (BPS) allows the identification of children with pneumonia who do not require antibiotics, but its impact has not been evaluated. We assessed whether the use of the BPS for the initial management of patients with pneumonia results in decreasing the use of antibiotics than under standard management of this condition without increasing patients' risks.

METHODS

This was a randomized, parallel-group, observer-blind, controlled clinical trial comparing the use of antibiotics in children aged 3-60 months treated for pneumonia in an outpatient setting, according to two methods of initial management. Patients were assigned randomly to management according to the BPS (antibiotic indication with a BPS ≥ 4 points) or routine management (antibiotic indication based on the institutional guidelines). We calculated the proportion of the use of antibiotics in each group and evaluated each patient's clinical outcome.

RESULTS

We included 120 patients (60 BPS and 60 controls) with a mean age of 24.2 ± 14.1 months. The use of antibiotics was significantly lower in the BPS group (46.6% vs. 86.6; OR 0.13; 95% CI: 0.05-0.35; P < 0.001). We observed an unfavorable outcome in 10 patients (8.3%), 5 in each group (P = 1.0; OR: 1.0 95% CI: 0.2-3.6).

CONCLUSION

The use of antibiotics was significantly lower in the group managed according to the BPS compared to the conventionally treated group, without increasing the rate of treatment failure.

Discovery and validation of biomarkers to guide clinical management of pneumonia in African children

Lipocalin 2 distinguishes severe and bacterial pneumonia from nonsevere and nonbacterial pneumonia with a high level of precision. The clinical impact of this biomarker requires large-scale clinical evaluation.

Background. Pneumonia is the leading cause of death in children globally. Clinical algorithms remain suboptimal for distinguishing severe pneumonia from other causes of respiratory distress such as malaria or distinguishing bacterial pneumonia and pneumonia from others causes, such as viruses. Molecular tools could improve diagnosis and management.

Methods. We conducted a mass spectrometry–based proteomic study to identify and validate markers of severity in 390 Gambian children with pneumonia (n = 204) and age-, sex-, and neighborhood-matched controls (n = 186). Independent validation was conducted in 293 Kenyan children with respiratory distress (238 with pneumonia, 41 with Plasmodium falciparum malaria, and 14 with both). Predictive value was estimated by the area under the receiver operating characteristic curve (AUC).

Results. Lipocalin 2 (Lpc-2) was the best protein biomarker of severe pneumonia (AUC, 0.71 [95% confidence interval, .64–.79]) and highly predictive of bacteremia (78% [64%–92%]), pneumococcal bacteremia (84% [71%–98%]), and “probable bacterial etiology” (91% [84%–98%]). These results were validated in Kenyan children with severe malaria and respiratory distress who also met the World Health Organization definition of pneumonia. The combination of Lpc-2 and haptoglobin distinguished bacterial versus malaria origin of respiratory distress with high sensitivity and specificity in Gambian children (AUC, 99% [95% confidence interval, 99%–100%]) and Kenyan children (82% [74%–91%]).

Conclusions. Lpc-2 and haptoglobin can help discriminate the etiology of clinically defined pneumonia and could be used to improve clinical management. These biomarkers should be further evaluated in prospective clinical studies.

Diagnostic features of community-acquired pneumonia in children: what's new?

AIM

To critically summarise the available data on diagnosis of CAP in children, focusing on the newest findings and on the need for new studies.

METHODS

Eighty studies on the diagnosis of paediatric community-acquired pneumonia were scrutinised.

RESULTS

We found no significant associations between the signs or symptoms and aetiology of pneumonia and concluded that chest radiographs remain controversial and real-time polymerase chain reaction appears more sensitive than blood cultures.

CONCLUSION

Antibiotic overuse could make it difficult to differentiate viral and bacterial causes. Molecular methods provide promising tools for diagnosing infection by atypical bacteria, but are expensive and should be used selectively.

Field testing the utility of procalcitonin and the acute infantile observation score in febrile infants 6 to 36 months old presenting to the pediatric emergency department with no obvious focus of infection

BACKGROUND

Serious bacterial illness (SBI) presents a diagnostic challenge in febrile infants. History, clinical signs, and laboratory information combined with experiential knowledge affects decisions to admit and treat.

AIM

To assess the utility of serum procalcitonin and the Acute Infantile Observation Score (AIOS) performed at emergency department presentation in predicting (a) confirmed serious bacterial illness and (b) illness severity.

METHODS

Sensitivity, specificity, and likelihood ratios were calculated for C-reactive protein, white cell count, serum procalcitonin, and AIOS.

RESULTS

Forty-six infants were recruited. Seven had a diagnosis of SBI, 28 were moderately ill with length of stay >24 hours, and 12 were severely ill with length of stay >96 hours. The positive likelihood ratios for confirmed SBI were C-reactive protein = 5.3, procalcitonin = 0.43, white cell count = 1.9, and AIOS = 1.5.

CONCLUSION

Procalcitonin and the AIOS do not modify the diagnostic uncertainty of the ED physician assessing the febrile infant with respect to admission or antibiotic treatment above standard laboratory investigations.

A 4 year prospective study to determine risk factors for severe community acquired pneumonia in children in southern China

BACKGROUND

Pneumonia is the major cause of death under 5 years. With high CAP numbers in China and growing access to PICUs, factors associated with severe CAP need to be determined to optimize care.

OBJECTIVE

To prospectively determine PICU CAP admission features and outcomes.

METHODS

A 4 year prospective study of CAP aged 1 month to <14 years admitted to PICU, Children's Hospital Affiliated to Soochow University, China. All were managed in a standard manner. Clinical, laboratory, and imaging findings were collected systematically. All received antibiotics.

RESULTS

Eight hundred ten (7%) of 10,836 CAP hospital admissions needed PICU. Seven hundred seven (87%) were enrolled. PICU CAP children were young (76% ≤ 12 months) and 33% had co-morbid conditions; 21% congenital heart disease.21% required mechanical ventilation. The average length of PICU stay was 5 days (range, 3-27). The case fatality rate was 5.8%. Viruses were detected in 38%, RSV 24%; bacteria in 23%, Streptococcus pneumoniae 7%, Haemophilus influenza b 4%, Mycoplasma 11%. On single factor analysis, PICU admission respiratory rate >70/min, grunting/groaning, head nodding, cyanosis, and anemia were associated with respiratory failure and with fatality. On multivariate analysis only presence of congenital heart disease, Trisomy 21 and immunodeficiency correlated with fatality; not microbe nor PICU findings.

CONCLUSIONS

Young age and underlying congenital heart disease were associated factors for PICU support in CAP in China. Early referral if altered sensorium, high respiratory rate, head nodding, grunting and anemia, and universal access to conjugated vaccines may decrease morbidity and mortality.

Clinical prediction rules and pediatric infectious diseases

Clinical prediction rules (CPR) are tools including appropriately weighted clinical aspects (history, physical examination and/or complementary tests) showing the odds for a specific diagnosis or prognosis. Their development includes a complex and strict process to achieve the scientific strength which supports use in clinical settings. Although CPR may be developed for almost any clinical situation, they are particularly useful in complex decision making, high-risk situations and for health cost reduction. Most CPR in pediatrics are devoted to infectious diseases, but only a few of them are used in daily practice. Reluctance in using them may be related to the most pediatrician's expectation of 100% sensitivity, when only a few CPR have sensitivity >90%. It is important to take into account that even a less-than-perfect CPR may be more sensitive than the physician's clinical judgment alone.

A literature review and survey of childhood pneumonia etiology studies: 2000-2010

The Pneumonia Etiology Research for Child Health (PERCH) project is the largest multicountry etiology study of childhood pneumonia since the Board on Science and Technology in International Development studies of the 1980s. However, it is not the only recent or ongoing pneumonia etiology study, and even with seven sites, it cannot capture all epidemiologic settings in the developing world. Funding providers, researchers and policymakers rely on the best available evidence to strategically plan programs, new research directions and interventions. We aimed to describe the current landscape of recent pneumonia etiology studies in children under 5 years of age in the developed and developing world, as ascertained by a literature review of relevant studies with data since the year 2000 and a survey of researchers in the field of childhood pneumonia. We collected information on the study population, study design, case definitions, laboratory samples and methods and identified pathogens. A literature review identified 88 studies with child pneumonia etiology results. As of June 2010, our survey of researchers identified an additional 65 ongoing and recently completed child pneumonia etiology studies. This demonstrates the broad existing context into which the PERCH study must be placed. However, the landscape analysis also reveals a multiplicity of case definitions, levels of clinician involvement, facility types, specimen collection, and laboratory techniques. It reinforces the need for the standardization of methods and analyses for present and future pneumonia etiology studies in order to optimize their cumulative potential to accurately describe the microbial causes of childhood pneumonia.

Clinical prediction rules for children: a systematic review

CONTEXT

The degree to which clinical prediction rules (CPRs) for children meet published standards is unclear.

OBJECTIVE

To systematically review the quality, performance, and validation of published CPRs for children, compare them with adult CPRs, and suggest pediatric-specific changes to CPR methodology.

METHODS

Medline was searched from 1950 to 2011. Studies were selected if they included the development of a CPR involving children younger than 18 years. Two investigators assessed study quality, rule performance, and rule validation as methodologic standards.

RESULTS

Of 7298 titles and abstracts assessed, 137 eligible studies were identified. They describe the development of 101 CPRs addressing 36 pediatric conditions. Quality standards met in fewer than half of the studies were blind assessment of predictors (47%), reproducibility of predictors (18%), blind assessment of outcomes (42%), adequate follow-up of outcomes (36%), adequate power (43%), adequate reporting of results (49%), and 95% confidence intervals reported (36%). For rule performance, 48% had a sensitivity greater than 0.95, and 43% had a negative likelihood ratio less than 0.1. For rule validation, 76% had no validation, 17% had narrow validation, 8% had broad validation, and none had impact analysis performed. Compared with CPRs for adult health conditions, quality and rule validation seem to be lower.

CONCLUSIONS

Many CPRs have been derived for children, but few have been validated. Relative to adult CPRs, several quality indicators demonstrated weaknesses. Existing performance standards may prove elusive for CPRs that involve children. CPRs for children that are more assistive and less directive and include patients' values and preferences in decision-making may be helpful.

How frequent is varicella-associated pneumonia in children?

Varicella is a frequent though mild infection in children, but it can cause important morbidity in adults. The most frequent complication in adults is varicella pneumonia. However, lower airway complications associated to varicella have been scarcely studied in children. We retrospectively reviewed the clinical records of the children hospitalized for varicella-associated pneumonia in the three public hospitals on the Island of Mallorca. We discovered that 17/213 (8%) children hospitalized for varicella were diagnosed of pneumonia. The rate of hospitalization for varicella complication due to pneumonia was 4.3 cases per 10,000 varicella infections. Only one patient was diagnosed of varicella pneumonia, which accounts for 0.3 cases per 10,000 varicella infections. Nine of 17 (53%) cases were classified as bacterial pneumonia. Statistical differences (p < 0.05) in the median time from rash (5.9 vs. 2.4 days) and fever (4.1 vs. 2.2 days) to admission were observed between bacterial and viral pneumonia. However, outcome differences measured by the median length of stay, need for oxygen, and admission to the intensive care unit were not observed. Varicella pneumonia is a very rare complication of varicella in children. Most of the lower airway infections in the course of varicella are associated to other co-infections.

Validation and development of a clinical prediction rule in clinically suspected community-acquired pneumonia

OBJECTIVES

To develop a mathematical model to predict the probability of having community-acquired pneumonia and to evaluate an already developed prediction rule that has not been validated in a clinical scenario.

METHODS

Children who presented with fever and had presumptive clinical diagnosis of pneumonia were evaluated in 4 institutions of different complexity during 1 year. The variables assessed were sex, age, respiratory rate, days with fever, maximum body temperature, presence of tachypnea, cough, chest pain, intercostal retraction, nasal flaring, abdominal pain, vomiting, grunting, rales, decreased breath sounds, wheezing, fatigue, loss of appetite, loss of sleep, and season of the year. The chest radiographs were photographed and then interpreted by 2 pediatric radiologists.

RESULTS

A total of 257 children were evaluated: 179 (69%) had clinical and radiological diagnosis of community-acquired pneumonia, and 78 (30%) had no radiological confirmation. A total of 96 photographs were recorded, and in 64 of the cases, there was agreement in the diagnosis between the evaluating pediatrician and the radiologists (kappa index = 0.68).With the calculated probabilities, it was possible to build a receiving operating characteristic curve and, based on the estimated coefficients we calculated, a value associated to the probability of having pneumonia.

CONCLUSIONS

We developed a model including 5 variables of high level of sensitivity for the diagnosis of pneumonia. To use it, it would be useful to apply the appropriate software. In addition, we validated a clinical prediction rule of 4 variables that proved to have 93.8% sensitivity to diagnose pneumonia in children with a fever and localized rales, or decreased breath sounds, or tachypnea, or any combination of these 4 variables.

Viral pneumonia: epidemiological, clinical, pathophysiological and therapeutic aspects

In humans, the most common types of infection are respiratory tract infections, among which viral infections predominate. Viruses can also infect the low respiratory tract, causing bronchiolitis, bronchitis and pneumonia. The objective of this review article was to show epidemiological, pathophysiological, clinical and therapeutic aspects of viral community-acquired pneumonia. These types of pneumonia are commonly caused by influenza A and B; parainfluenza 1, 2 and 3; respiratory syncytial virus; or adenovirus. We also address the types of pneumonia caused by hantaviruses, metapneumoviruses and rhinoviruses.

Two-transcript gene expression classifiers in the diagnosis and prognosis of human diseases

Background

Identification of molecular classifiers from genome-wide gene expression analysis is an important practice for the investigation of biological systems in the post-genomic era - and one with great potential for near-term clinical impact. The 'Top-Scoring Pair' (TSP) classification method identifies pairs of genes whose relative expression correlates strongly with phenotype. In this study, we sought to assess the effectiveness of the TSP approach in the identification of diagnostic classifiers for a number of human diseases including bacterial and viral infection, cardiomyopathy, diabetes, Crohn's disease, and transformed ulcerative colitis. We examined transcriptional profiles from both solid tissues and blood-borne leukocytes.

Results

The algorithm identified multiple predictive gene pairs for each phenotype, with cross-validation accuracy ranging from 70 to nearly 100 percent, and high sensitivity and specificity observed in most classification tasks. Performance compared favourably with that of pre-existing transcription-based classifiers, and in some cases was comparable to the accuracy of current clinical diagnostic procedures. Several diseases of solid tissues could be reliably diagnosed through classifiers based on the blood-borne leukocyte transcriptome. The TSP classifier thus represents a simple yet robust method to differentiate between diverse phenotypic states based on gene expression profiles.

Conclusion

Two-transcript classifiers have the potential to reliably classify diverse human diseases, through analysis of both local diseased tissue and the immunological response assayed through blood-borne leukocytes. The experimental simplicity of this method results in measurements that can be easily translated to clinical practice.

Diagnostic utility of the soluble triggering receptor expressed on myeloid cells-1 in bronchoalveolar lavage fluid from patients with bilateral lung infiltrates

Background

Differential diagnosis of patients with bilateral lung infiltrates remains a difficult problem for intensive care clinicians. Here we evaluate the diagnostic role of soluble triggering receptor expressed on myeloid cells-1 (sTREM-1) in bronchoalveolar lavage (BAL) specimens from patients with bilateral lung infiltrates.

Methods

We conducted a prospective observational study on 80 patients with bilateral lung infiltrates with clinical suspicion of infectious pneumonia. Patients were categorized into three groups: bacterial or fungal infection, intracellular or viral infection, and noninfectious inflammatory disease. sTREM-1 concentrations were measured, and BAL fluid and Clinical Pulmonary Infection Score (CPIS) were analyzed.

Results

The sTREM-1 concentration was significantly increased in patients with bacterial or fungal pneumonia (n = 29, 521.2 ± 94.7 pg/ml), compared with that in patients with viral pneumonia, atypical pneumonia or tuberculosis (n = 14, 92.9 ± 20.0 pg/ml) or noninfectious inflammatory disease (n = 37, 92.8 ± 10.7 pg/ml). The concentration of sTREM-1 in BAL fluid, but not CPIS, was an independent predictor of bacterial or fungal pneumonia, and a cutoff value of more than 184 pg/ml yielded a diagnostic sensitivity of 86% and a specificity of 90%.

Conclusion

The sTREM-1 level in BAL fluid from patients with bilateral lung infiltrates is a potential marker for the differential diagnosis of pneumonia due to extracellular bacteria.

Community-acquired pneumonia: a review and recent advances

In the beginning of this 21st century, community-acquired pneumonias (CAP) are still responsible for a significant number of deaths among young children in many developing countries. Public health initiatives such as those proposed by the World Health Organization (WHO) for the management of CAP by means of identifying highly predictable signs and symptoms have had great positive impact in some communities. Still, this approach induces an overdiagnosis and overtreatment of CAP in children below the age of 5 years due to the misclassification of pneumonia in children with fast breathing associated with viral bronchiolitis. Even among children of developed countries, CAP is an important public health problem and many aspects of current diagnostic and management measures are discussed here. In this article, we review the epidemiology and basic concepts of CAP and update current information on clinical evaluation and management of the disease.

Pneumonien

Pneumonien sind wichtige Verursacher kindlicher Morbidität und in den Entwicklungsländern eine führende Todesursache. Die ätiologische Diagnostik bakterieller Infektionen ist unbefriedigend, die therapeutischen Möglichkeiten bei viralen Pneumonien sehr beschränkt. Klare Möglichkeiten zur Differenzierung der einzelnen Pneumonien fehlen. Für die Behandlung ambulant erworbener Pneumonien orientiert man sich am besten an den zu erwartenden Erregern, die ein gewisse Altersabhängigkeit aufweisen. Amoxicillin und Makrolide stellen die besten, weil zielorientiertesten, Antibiotika für ambulante Pneumonien dar. Angesichts noch relativ günstiger Empfindlichkeitsverhältnisse in mitteleuropäischen Ländern ist nicht primär mit Therapieversagern zu rechnen. Besondere Probleme ergeben sich bei neonatalen Pneumonien sowie bei Pneumonien als Folge von Immunsuppression. Allergisch bedingte Pneumonien müssen gesondert untersucht werden.