Dataset Information

Development and Validation of a Predictive Model of the Risk of Pediatric Septic Shock Using Data Known at the Time of Hospital Arrival.

ABSTRACT:

Objective

To derive and validate a model of risk of septic shock among children with suspected sepsis, using data known in the electronic health record at hospital arrival.

Study design

This observational cohort study at 6 pediatric emergency department and urgent care sites used a training dataset (5 sites, April 1, 2013, to December 31, 2016), a temporal test set (5 sites, January 1, 2017 to June 30, 2018), and a geographic test set (a sixth site, April 1, 2013, to December 31, 2018). Patients 60 days to 18 years of age in whom clinicians suspected sepsis were included; patients with septic shock on arrival were excluded. The outcome, septic shock, was systolic hypotension with vasoactive medication or ?30 mL/kg of isotonic crystalloid within 24 hours of arrival. Elastic net regularization, a penalized regression technique, was used to develop a model in the training set.

Results

Of 2464 included visits, septic shock occurred in 282 (11.4%). The model had an area under the curve of 0.79 (0.76-0.83) in the training set, 0.75 (0.69-0.81) in the temporal test set, and 0.87 (0.73-1.00) in the geographic test set. With a threshold set to 90% sensitivity in the training set, the model yielded 82% (72%-90%) sensitivity and 48% (44%-52%) specificity in the temporal test set, and 90% (55%-100%) sensitivity and 32% (21%-46%) specificity in the geographic test set.

Conclusions

This model estimated the risk of septic shock in children at hospital arrival earlier than existing models. It leveraged the predictive value of routine electronic health record data through a modern predictive algorithm and has the potential to enhance clinical risk stratification in the critical moments before deterioration.

SUBMITTER: Scott HF

PROVIDER: S-EPMC6980682 | biostudies-literature | 2020 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Development and Validation of a Predictive Model of the Risk of Pediatric Septic Shock Using Data Known at the Time of Hospital Arrival.

Scott Halden F HF Colborn Kathryn L KL Sevick Carter J CJ Bajaj Lalit L Kissoon Niranjan N Deakyne Davies Sara J SJ Kempe Allison A

The Journal of pediatrics 20191113

<h4>Objective</h4>To derive and validate a model of risk of septic shock among children with suspected sepsis, using data known in the electronic health record at hospital arrival.<h4>Study design</h4>This observational cohort study at 6 pediatric emergency department and urgent care sites used a training dataset (5 sites, April 1, 2013, to December 31, 2016), a temporal test set (5 sites, January 1, 2017 to June 30, 2018), and a geographic test set (a sixth site, April 1, 2013, to December 31, ...[more]

PMID: 31733815

Similar Datasets

Project description:Rationale: We previously generated genome-wide expression data in children with septic shock, based on whole blood-derive RNA, having the potential to lead the field into novel areas of investigation. Objective: Herein we seek to validate our data through a bioinformatic approach centered on a validation patient cohort. Methods: Microarray- and bioinformatics-centered analyses involving our previous data as a training data set (n = 42) and a new, validation cohort (n = 30) as the test data set. Measurements and Main Results: Class prediction modeling using the training data set and the previously reported genome-wide expression signature of pediatric septic shock correctly identified 93 to 100% of septic shock patients in the test data set, depending on the class prediction algorithm and the gene selection method. Subjecting the test data set to an identical filtering strategy as that used for the training data set, demonstrated 72% concordance between the two gene lists. Subjecting the test data set to a purely statistical filtering strategy, with highly stringent correction for multiple comparisons, demonstrated less than 50% concordance with the previous gene filtering strategy. However, functional analysis of this statistics-based gene list demonstrated similar functional annotations and signaling pathways as that seen in the learning data set. In particular, we validated that pediatric septic shock is characterized by large scale repression of genes related to zinc homeostasis and lymphocyte function. Conclusions: These data demonstrate that the previously reported genome-wide expression signature of pediatric septic shock is applicable to a validation cohort of patients. Experiment Overall Design: Table 1: Clinical and demographic data for all subjects in test data set. Experiment Overall Design: Controls Septic Shock Experiment Overall Design: No. of individual subjects 15 30 Experiment Overall Design: Mean age (years) Â± S.D. 3.1 Â± 3.5 3.2 Â± 2.9 Experiment Overall Design: Mean PRISM Score Â± S.D. n/a 18.9 Â± 12.3 Experiment Overall Design: Gender (Male/Female) 8/7 16/14 Experiment Overall Design: Race (no.) A.A./Black (6) A.A./Black (2) Experiment Overall Design: Asian (4) White (26) White (5) Unreported (2)

Project description:Background: Septic shock heterogeneity has important implications for the conduct of clinical trials and individual patient management. We previously addressed this heterogeneity by indentifying 3 putative subclasses of children with septic shock based on a 100-gene expression signature corresponding to adaptive immunity and glucocorticoid receptor signaling. Herein we attempted to prospectively validate the existence of these gene expression-based subclasses in a validation cohort. Methods: Gene expression mosaics were generated from the 100 class-defining genes for 82 individual patients in the validation cohort. Patients were classified into 1 of 3 subclasses (“A”, “B”, or “C”) based on color and pattern similarity relative to reference mosaics generated from the original derivation cohort. Separate classifications were conducted by 21 individual clinicians and a computer-based algorithm. After subclassification the clinical database was mined for clinical phenotyping. Results: In the final consensus subclassification generated by clinicians, subclass A patients had a higher illness severity, as measured by illness severity scores and maximal organ failure, relative to subclasses B and C. The k coefficient across all possible inter-evaluator comparisons was 0.633. Similar observations were made based on the computer-generated subclassification. Patients in subclass A were also characterized by repression of a large number of genes having functional annotations related to zinc biology. Conclusions: We have validated the existence of subclasses of children with septic shock based on a biologically relevant, 100-gene expression signature. The subclasses can be indentified by clinicians without formal bioinformatics training, at a clinically relevant time point, and have clinically relevant phenotypic differences. Expression data from 82 children with septic shock and 21 normal controls were generated using whole blood-derived RNA samples representing the first 24 hours of admission to the pediatric intensive care unit. The controls were used for normalization. Subsequently, we used the expression data from 100 class defining genes to validate the existence of pediatric septic shock subclasses having phenotypic differences.

Dataset Information

Development and Validation of a Predictive Model of the Risk of Pediatric Septic Shock Using Data Known at the Time of Hospital Arrival.

Objective

Study design

Results

Conclusions

Publications

Development and Validation of a Predictive Model of the Risk of Pediatric Septic Shock Using Data Known at the Time of Hospital Arrival.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure