Dataset Information

Development and performance of electronic acute kidney injury triggers to identify pediatric patients at risk for nephrotoxic medication-associated harm.

ABSTRACT:

Background

Nephrotoxic medication-associated acute kidney injury (NTMx-AKI) is a costly clinical phenomenon and more common than previously recognized. Prior efforts to use technology to identify AKI have focused on detection after renal injury has occurred.

Objectives

Describe an approach and provide a technical framework for the creation of risk-stratifying AKI triggers and the development of an application to manage the AKI trigger data. Report the performance characteristics of those triggers and the refinement process and on the challenges of implementation.

Methods

Initial manual trigger screening guided design of an automated electronic trigger report. A web-based application was designed to alleviate inefficiency and serve as a user interface and central workspace for the project. Performance of the NTMx exposure trigger reports from September 2011 to September 2013 were evaluated using sensitivity (SN), specificity (SP), positive and negative predictive values (PPV, NPV).

Results

Automated reports were created to replace manual screening for NTMx-AKI. The initial performance of the NTMx exposure triggers for SN, SP, PPV, and NPV all were ?0.78, and increased over the study, with all four measures reaching ?0.95 consistently. A web-based application was implemented that simplifies data entry and couriering from the reports, expedites results viewing, and interfaces with an automated data visualization tool. Sociotechnical challenges were logged and reported.

Conclusion

We have built a risk-stratifying system based on electronic triggers that detects patients at-risk for NTMx-AKI before injury occurs. The performance of the NTMx-exposed reports has neared 100% through iterative optimization. The complexity of the trigger logic and clinical workflows surrounding NTMx-AKI led to a challenging implementation, but one that has been successful from technical, clinical, and quality improvement standpoints. This report summarizes the construction of a trigger-based application, the performance of the triggers, and the challenges uncovered during the design, build, and implementation of the system.

SUBMITTER: Kirkendall ES

PROVIDER: S-EPMC4081739 | biostudies-literature | 2014

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Development and performance of electronic acute kidney injury triggers to identify pediatric patients at risk for nephrotoxic medication-associated harm.

Kirkendall E S ES Spires W L WL Mottes T A TA Schaffzin J K JK Barclay C C Goldstein S L SL

Applied clinical informatics 20140402 2

<h4>Background</h4>Nephrotoxic medication-associated acute kidney injury (NTMx-AKI) is a costly clinical phenomenon and more common than previously recognized. Prior efforts to use technology to identify AKI have focused on detection after renal injury has occurred.<h4>Objectives</h4>Describe an approach and provide a technical framework for the creation of risk-stratifying AKI triggers and the development of an application to manage the AKI trigger data. Report the performance characteristics o ...[more]

PMID: 25024752

Similar Datasets

Project description:Electronic health records (EHRs) have given rise to large and complex databases of medical information that have the potential to become powerful tools for clinical research. However, differences in coding systems and the detail and accuracy of the information within EHRs can vary across institutions. This makes it challenging to identify subpopulations of patients and limits the widespread use of multi-institutional databases. In this study, we leveraged machine learning to identify patterns in medication usage among hospitalized pediatric patients receiving renal replacement therapy and created a predictive model that successfully differentiated between intermittent (iHD) and continuous renal replacement therapy (CRRT) hemodialysis patients. We trained six machine learning algorithms (logistical regression, Naïve Bayes, k-nearest neighbor, support vector machine, random forest, and gradient boosted trees) using patient records from a multi-center database (n = 533) and prescribed medication ingredients (n = 228) as features to discriminate between the two hemodialysis types. Predictive skill was assessed using a 5-fold cross-validation, and the algorithms showed a range of performance from 0.7 balanced accuracy (logistical regression) to 0.86 (random forest). The two best performing models were further tested using an independent single-center dataset and achieved 84-87% balanced accuracy. This model overcomes issues inherent within large databases and will allow us to utilize and combine historical records, significantly increasing population size and diversity within both iHD and CRRT populations for future clinical studies. Our work demonstrates the utility of using medications alone to accurately differentiate subpopulations of patients in large datasets, allowing codes to be transferred between different coding systems. This framework has the potential to be used to distinguish other subpopulations of patients where discriminatory ICD codes are not available, permitting more detailed insights and new lines of research.

Dataset Information

Development and performance of electronic acute kidney injury triggers to identify pediatric patients at risk for nephrotoxic medication-associated harm.

Background

Objectives

Methods

Results

Conclusion

Publications

Development and performance of electronic acute kidney injury triggers to identify pediatric patients at risk for nephrotoxic medication-associated harm.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets