A Prediction Model for Pediatric Radiographic Pneumonia

PEDIATRICS ◽  
2021 ◽  
Author(s):  
Sriram Ramgopal ◽  
Lilliam Ambroggio ◽  
Douglas Lorenz ◽  
Samir S. Shah ◽  
Richard M. Ruddy ◽  
...  
Keyword(s):  
Predictive Model ◽  
External Validation ◽  
Multivariable Analysis ◽  
Model Performance ◽  
Antibiotic Use ◽  
Community Acquired Pneumonia ◽  
Reduced Model ◽  
Risk Threshold ◽  
Breath Sounds ◽  
Predicted Risk

BACKGROUND: Chest radiographs (CXRs) are frequently used in the diagnosis of community-acquired pneumonia (CAP). We sought to construct a predictive model for radiographic CAP based on clinical features to decrease CXR use. METHODS: We performed a single-center prospective study of patients 3 months to 18 years of age with signs of lower respiratory infection who received a CXR for suspicion of CAP. We used penalized multivariable logistic regression to develop a full model and bootstrapped backward selection models to develop a parsimonious reduced model. We evaluated model performance at different thresholds of predicted risk. RESULTS: Radiographic CAP was identified in 253 (22.2%) of 1142 patients. In multivariable analysis, increasing age, prolonged fever duration, tachypnea, and focal decreased breath sounds were positively associated with CAP. Rhinorrhea and wheezing were negatively associated with CAP. The bootstrapped reduced model retained 3 variables: age, fever duration, and decreased breath sounds. The area under the receiver operating characteristic for the reduced model was 0.80 (95% confidence interval: 0.77–0.84). Of 229 children with a predicted risk of <4%, 13 (5.7%) had radiographic CAP (sensitivity of 94.9% at a 4% risk threshold). Conversely, of 229 children with a predicted risk of >39%, 140 (61.1%) had CAP (specificity of 90% at a 39% risk threshold). CONCLUSIONS: A predictive model including age, fever duration, and decreased breath sounds has excellent discrimination for radiographic CAP. After external validation, this model may facilitate decisions around CXR or antibiotic use in CAP.

scholarly journals A risk score for predicting hospitalization for community-acquired pneumonia in ITP using nationally representative data

2020 ◽  
Vol 4 (22) ◽  
pp. 5846-5857
Author(s):  
Ye-Jun Wu ◽  
Ming Hou ◽  
Hui-Xin Liu ◽  
Jun Peng ◽  
Liang-Ming Ma ◽  
...  
Keyword(s):  
Validation Cohort ◽  
External Validation ◽  
Multivariable Analysis ◽  
Area Under The Curve ◽  
Absolute Lymphocyte Count ◽  
Community Acquired Pneumonia ◽  
Charlson Comorbidity Index Score ◽  
Discrimination Power ◽  
Derivation Cohort ◽  
Medical Centers

Abstract Infection is one of the primary causes of death from immune thrombocytopenia (ITP), and the lungs are the most common site of infection. We identified the factors associated with hospitalization for community-acquired pneumonia (CAP) in nonsplenectomized adults with ITP and established the ACPA prediction model to predict the incidence of hospitalization for CAP. This was a retrospective study of nonsplenectomized adult patients with ITP from 10 large medical centers in China. The derivation cohort included 145 ITP inpatients with CAP and 1360 inpatients without CAP from 5 medical centers, and the validation cohort included the remaining 63 ITP inpatients with CAP and 526 inpatients without CAP from the other 5 centers. The 4-item ACPA model, which included age, Charlson Comorbidity Index score, initial platelet count, and initial absolute lymphocyte count, was established by multivariable analysis of the derivation cohort. Internal and external validation were conducted to assess the performance of the model. The ACPA model had an area under the curve of 0.853 (95% confidence interval [CI], 0.818-0.889) in the derivation cohort and 0.862 (95% CI, 0.807-0.916) in the validation cohort, which indicated the good discrimination power of the model. Calibration plots showed high agreement between the estimated and observed probabilities. Decision curve analysis indicated that ITP patients could benefit from the clinical application of the ACPA model. To summarize, the ACPA model was developed and validated to predict the occurrence of hospitalization for CAP, which might help identify ITP patients with a high risk of hospitalization for CAP.

scholarly journals Effectiveness of transfer learning for enhancing tumor classification with a convolutional neural network on frozen sections

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Young-Gon Kim ◽  
Sungchul Kim ◽  
Cristina Eunbee Cho ◽  
In Hye Song ◽  
Hee Jin Lee ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Transfer Learning ◽  
Frozen Section ◽  
Medical Center ◽  
External Validation ◽  
Model Performance ◽  
Classification Model ◽  
Training Dataset

AbstractFast and accurate confirmation of metastasis on the frozen tissue section of intraoperative sentinel lymph node biopsy is an essential tool for critical surgical decisions. However, accurate diagnosis by pathologists is difficult within the time limitations. Training a robust and accurate deep learning model is also difficult owing to the limited number of frozen datasets with high quality labels. To overcome these issues, we validated the effectiveness of transfer learning from CAMELYON16 to improve performance of the convolutional neural network (CNN)-based classification model on our frozen dataset (N = 297) from Asan Medical Center (AMC). Among the 297 whole slide images (WSIs), 157 and 40 WSIs were used to train deep learning models with different dataset ratios at 2, 4, 8, 20, 40, and 100%. The remaining, i.e., 100 WSIs, were used to validate model performance in terms of patch- and slide-level classification. An additional 228 WSIs from Seoul National University Bundang Hospital (SNUBH) were used as an external validation. Three initial weights, i.e., scratch-based (random initialization), ImageNet-based, and CAMELYON16-based models were used to validate their effectiveness in external validation. In the patch-level classification results on the AMC dataset, CAMELYON16-based models trained with a small dataset (up to 40%, i.e., 62 WSIs) showed a significantly higher area under the curve (AUC) of 0.929 than those of the scratch- and ImageNet-based models at 0.897 and 0.919, respectively, while CAMELYON16-based and ImageNet-based models trained with 100% of the training dataset showed comparable AUCs at 0.944 and 0.943, respectively. For the external validation, CAMELYON16-based models showed higher AUCs than those of the scratch- and ImageNet-based models. Model performance for slide feasibility of the transfer learning to enhance model performance was validated in the case of frozen section datasets with limited numbers.

#61: Antibiotic Use for Community-Acquired Pneumonia Among Hospitalized Children in Japan

2021 ◽  
Vol 10 (Supplement_1) ◽  
pp. S13-S13
Author(s):  
Chiaki Tao-Kidoguchi ◽  
Eiki Ogawa ◽  
Kensuke Shoji ◽  
Isao Miyairi
Keyword(s):  
Physician Behavior ◽  
Point Of Care ◽  
Descriptive Analysis ◽  
Antibiotic Use ◽  
Community Acquired Pneumonia ◽  
Clinical Criteria ◽  
Focus Of Infection ◽  
Tract Infections ◽  
Short Courses ◽  
Viral Etiologies

Abstract Background Judicious use of antimicrobials is the cornerstone of action against antimicrobial resistance. Respiratory tract infections account for over 80% of pediatric antibiotic use in Japan. Antibiotics are generally used empirically for most hospitalized patients with pneumonia although it is becoming clearer that viral etiologies account for approximately 70% of these cases. Defining the characteristics of patients who are managed with a short course of antibiotics and subsequently do well, may lead to setting clinical criteria for early termination of antibiotics. Methods We performed a retrospective descriptive analysis. Medical charts of patients aged 3 months to 18 years, who were admitted with a diagnosis of pneumonia, bronchitis, bronchiolitis, or asthma to the Department of Interdisciplinary medicine at the National Center for Child Health and Development from March 2018 through February 2019 were reviewed. Those who had respiratory symptoms and were started on antibiotics within 48 hours of hospitalization were included. Those who had a focus of infection elsewhere or were immunocompromised were excluded. Results Of the 556 candidates, 80 patients met the criteria. The median age was 1.5 years which included 42.5% (34/80) with comorbidities. Underlying conditions included 9 with trisomy 21, and 8 with perinatal issues. Rapid antigen testing was performed and 7 patients with RSV, 5 patients with influenza, 1 patient with human metapneumovirus were identified. The average duration of antibiotic therapy was 7.2 days (range 2–14 days). There were no statistical differences in the characteristics of patients who received antibiotics for more or less than 5 days. The positivity of the rapid antigen test tended to be higher in those who received antibiotics for a shorter period (25% vs. 15%). There were no differences in the rate of readmission or complications between the two groups. Conclusion We were unable to identify a clear characteristic of patients who received short courses of antibiotics for pneumonia. The trend observed for those who had a point of care testing may suggest that the use of a multiplex PCR testing covering a greater number of pathogens would influence physician behavior in antibiotic use.

Evaluation of predictive model performance of an existing model in the presence of missing data

2021 ◽  
Author(s):  
Pin Li ◽  
Jeremy M. G. Taylor ◽  
Daniel E. Spratt ◽  
R. Jeffery Karnes ◽  
Matthew J. Schipper
Keyword(s):  
Missing Data ◽  
Predictive Model ◽  
Model Performance

scholarly journals Which Analysis Approach Is Adequate to Leverage Clinical Microdialysis Data? A Quantitative Comparison to Investigate Exposure and Reponse Exemplified by Levofloxacin

2021 ◽  
Author(s):  
David Busse ◽  
André Schaeftlein ◽  
Alexander Solms ◽  
Luis Ilia ◽  
Robin Michelet ◽  
...  
Keyword(s):  
Drug Exposure ◽  
Visual Predictive Check ◽  
Model Performance ◽  
Compartmental Analysis ◽  
Community Acquired Pneumonia ◽  
Target Site ◽  
Clinical Pharmacokinetics ◽  
Time Integral ◽  
Dosing Regimens ◽  
Therapeutic Decision Making

Abstract Purpose Systematic comparison of analysis methods of clinical microdialysis data for impact on target-site drug exposure and response. Methods 39 individuals received a 500 mg levofloxacin short-term infusion followed by 24-h dense sampling in plasma and microdialysate collection in interstitial space fluid (ISF). ISF concentrations were leveraged using non-compartmental (NCA) and compartmental analysis (CA) via (ii) relative recovery correction at midpoint of the collection interval (midpoint-NCA, midpoint-CA) and (ii) dialysate-based integrals of time (integral-CA). Exposure and adequacy of community-acquired pneumonia (CAP) therapy via pharmacokinetic/pharmacodynamic target-attainment (PTA) analysis were compared between approaches. Results Individual AUCISF estimates strongly varied for midpoint-NCA and midpoint-CA (≥52.3%CV) versus integral-CA (≤32.9%CV) owing to separation of variability in PK parameters (midpoint-CA = 46.5%–143%CVPK, integral-CA = 26.4%–72.6%CVPK) from recovery-related variability only in integral-CA (41.0%–50.3%CVrecovery). This also led to increased variability of AUCplasma for midpoint-CA (56.0%CV) versus midpoint-NCA and integral-CA (≤33.0%CV), and inaccuracy of predictive model performance of midpoint-CA in plasma (visual predictive check). PTA analysis translated into 33% of evaluated patient cases being at risk of incorrectly rejecting recommended dosing regimens at CAP-related epidemiological cut-off values. Conclusions Integral-CA proved most appropriate to characterise clinical pharmacokinetics- and microdialysis-related variability. Employing this knowledge will improve the understanding of drug target-site PK for therapeutic decision-making.

scholarly journals External validation of a predictive model of survival after cytoreductive nephrectomy for metastatic renal cell carcinoma

2018 ◽  
Vol 36 (12) ◽  
pp. 1973-1980 ◽  
Author(s):  
Lorenzo Marconi ◽  
Roderick de Bruijn ◽  
Erik van Werkhoven ◽  
Christian Beisland ◽  
Kate Fife ◽  
...  
Keyword(s):  
Renal Cell Carcinoma ◽  
Cell Carcinoma ◽  
Predictive Model ◽  
Metastatic Renal Cell Carcinoma ◽  
Renal Cell ◽  
External Validation ◽  
Cytoreductive Nephrectomy

External validation of a predictive model of adverse events following spine surgery

2021 ◽  
Author(s):  
Parastou Fatemi ◽  
Yi Zhang ◽  
Summer S. Han ◽  
Natasha Purington ◽  
Corinna C. Zygourakis ◽  
...  
Keyword(s):  
Adverse Events ◽  
Predictive Model ◽  
Spine Surgery ◽  
External Validation

scholarly journals Antibiotic Prescribing for Adults Hospitalized in the Etiology of Pneumonia in the Community Study

2017 ◽  
Vol 4 (2) ◽  
Author(s):  
Sara Tomczyk ◽  
Seema Jain ◽  
Anna M Bramley ◽  
Wesley H Self ◽  
Evan J Anderson ◽  
...  
Keyword(s):  
Antibiotic Use ◽  
Community Acquired Pneumonia ◽  
Community Study ◽  
Antibiotic Prescribing ◽  
Beta Lactam ◽  
First Line ◽  
Factors Associated ◽  
Admission Ward ◽  
Healthcare Associated ◽  
To Receive

Abstract Background Community-acquired pneumonia (CAP) 2007 guidelines from the Infectious Diseases Society of America (IDSA)/American Thoracic Society (ATS) recommend a respiratory fluoroquinolone or beta-lactam plus macrolide as first-line antibiotics for adults hospitalized with CAP. Few studies have assessed guideline-concordant antibiotic use for patients hospitalized with CAP after the 2007 IDSA/ATS guidelines. We examine antibiotics prescribed and associated factors in adults hospitalized with CAP. Methods From January 2010 to June 2012, adults hospitalized with clinical and radiographic CAP were enrolled in a prospective Etiology of Pneumonia in the Community study across 5 US hospitals. Patients were interviewed using a standardized questionnaire, and medical charts were reviewed. Antibiotics prescribed were classified according to defined nonrecommended CAP antibiotics. We assessed factors associated with nonrecommended CAP antibiotics using logistic regression. Results Among enrollees, 1843 of 1874 (98%) ward and 440 of 446 (99%) ICU patients received ≥1 antibiotic ≤24 hours after admission. Ward patients were prescribed a respiratory fluoroquinolone alone (n = 613; 33%), or beta-lactam plus macrolide (n = 365; 19%), beta-lactam alone (n = 240; 13%), among other antibiotics, including vancomycin (n = 235; 13%) or piperacillin/tazobactam (n = 157; 8%) ≤24 hours after admission. Ward patients with known risk for healthcare-associated pneumonia (HCAP), recent outpatient antibiotic use, and in-hospital antibiotic use <6 hours after admission were significantly more likely to receive nonrecommended CAP antibiotics. Conclusions Although more than half of ward patients received antibiotics concordant with IDSA/ATS guidelines, a number received nonrecommended CAP antibiotics, including vancomycin and piperacillin/tazobactam; risk factors for HCAP, recent outpatient antibiotic, and rapid inpatient antibiotic use contributed to this. This hypothesis-generating descriptive epidemiology analysis could help inform antibiotic stewardship efforts, reinforces the need to harmonize guidelines for CAP and HCAP, and highlights the need for improved diagnostics to better equip clinicians.

Sign in / Sign up

Export Citation Format

Share Document