Derivation of a prediction model for emergency department acute kidney injury

Phillips, Aled Owain; Foxwell, David; Pradhan, Sara; Zouwail, Soha; Rainer, Timothy

doi:10.1016/j.ajem.2020.12.017

Cited by 2 publications

(12 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The study flowchart according to the PRISMA statement is shown in Figure 1. The review included 150 studies (encompassing 14.4 million patients) with requisite data to enable sROC curves in 103 studies. The median duration of study data collection was 4 years (range, 3 months to 17 years).…”

Section: Resultsmentioning

confidence: 99%

“…Most studies reported prospective data collection (124 [82.7%]) . Based on prespecified clinical scenario subgroups, 29 studies assessed AKI associated with nephrotoxic agents, including 26 studies examining radio-contrast medium associated AKI and 3 studies that reported AKI associated with other nephrotoxins; 64 studies assessed AKI prediction in the postoperative setting; 21 studies assessed AKI prediction in intensive care units (ICU); 35 studies assessed AKI in general hospitalizations…”

Section: Resultsmentioning

confidence: 99%

“…Based on prespecified clinical scenario subgroups, 29 studies assessed AKI associated with nephrotoxic agents, including 26 studies examining radio-contrast medium associated AKI and 3 studies that reported AKI associated with other nephrotoxins; 64 studies assessed AKI prediction in the postoperative setting; 21 studies assessed AKI prediction in intensive care units (ICU); 35 studies assessed AKI in general hospitalizations; and 1 assessed AKI in the outpatient setting . The most common region where the studies had been conducted was the Asia-Pacific region, accounting for 40.0% of the included studies (60 studies), followed by North America (37.3% [56 studies])…”

Section: Resultsmentioning

confidence: 99%

“…The AKI definitions varied across the studies, with most being the KDIGO criteria (76 [50.7%]), followed by the AKIN criteria (17 [11.3%]) and the RIFLE criteria (9 [6.0%]) . Nearly a quarter of the studies (35 [23.3%]) defined AKI using criteria specific to their particular studies .…”

Section: Resultsmentioning

confidence: 99%

“…Nearly a quarter of the studies (35 [23.3%]) defined AKI using criteria specific to their particular studies . A large majority of the studies defined AKI using the least severe stage of the disease (stage 1, 118 of 146 [80.8%]) . The median (IQR) values for AKI event rate and sample size for model development were 13.4% (7.14%-27.9%) and 1598 (478-16 668), respectively.…”

Section: Resultsmentioning

confidence: 99%

See 4 more Smart Citations

Characterization of Risk Prediction Models for Acute Kidney Injury

et al. 2023

View full text Add to dashboard Cite

ImportanceDespite the expansion of published prediction models for acute kidney injury (AKI), there is little evidence of uptake of these models beyond their local derivation nor data on their association with patient outcomes.ObjectiveTo systematically review published AKI prediction models across all clinical subsettings.Data SourcesMEDLINE via PubMed (January 1946 to April 2021) and Embase (January 1947 to April 2021) were searched using medical subject headings and text words related to AKI and prediction models.Study SelectionAll studies that developed a prediction model for AKI, defined as a statistical model with at least 2 predictive variables to estimate future occurrence of AKI, were eligible for inclusion. There was no limitation on study populations or methodological designs.Data Extraction and SynthesisTwo authors independently searched the literature, screened the studies, and extracted and analyzed the data following the Preferred Reporting Items for Systematic Review and Meta-analyses guideline. The data were pooled using a random-effects model, with subgroups defined by 4 clinical settings. Between-study heterogeneity was explored using multiple methods, and funnel plot analysis was used to identify publication bias.Main Outcomes and MeasuresC statistic was used to measure the discrimination of prediction models.ResultsOf the 6955 studies initially identified through literature searching, 150 studies, with 14.4 million participants, met the inclusion criteria. The study characteristics differed widely in design, population, AKI definition, and model performance assessments. The overall pooled C statistic was 0.80 (95% CI, 0.79-0.81), with pooled C statistics in different clinical subsettings ranging from 0.78 (95% CI, 0.75-0.80) to 0.82 (95% CI, 0.78-0.86). Between-study heterogeneity was high overall and in the different clinical settings (eg, contrast medium–associated AKI: I2 = 99.9%; P &lt; .001), and multiple methods did not identify any clear sources. A high proportion of models had a high risk of bias (126 [84.4%]) according to the Prediction Model Risk Of Bias Assessment Tool.Conclusions and RelevanceIn this study, the discrimination of the published AKI prediction models was good, reflected by high C statistics; however, the wide variation in the clinical settings, populations, and predictive variables likely drives the highly heterogenous findings that limit clinical utility. Standardized procedures for development and validation of prediction models are urgently needed.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

See 3 more Smart Citations

Characterization of Risk Prediction Models for Acute Kidney Injury

et al. 2023

View full text Add to dashboard Cite

show abstract

Machine-learning-based analysis of biomedical time-series data: the monitoring and prediction of disease progression

Zhang

2024

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

This study examines the spatial and temporal patterns of influenza and malaria incidence using an ARMA-BP combination model. The approach employs the dynamic series method to identify epidemic patterns of these diseases while assessing serial autocorrelation coefficients, performing randomness tests, and establishing a forecasting model. Additionally, it evaluates the impact of seasonal and meteorological factors on the epidemiology of influenza and malaria to ascertain the model’s efficacy in predicting incidence rates and trends. The findings indicate that the peak period for influenza incidence typically occurs during the transition from winter to spring, specifically between weeks 2 and 14. The correlation coefficients between temperature variables and malaria incidence generally ranged from 0.7 to 0.9. The ARMA-BP model demonstrated robust short-term predictive capabilities for influenza, showing a high degree of concordance in predictions for 2021 and 2022, though it performed less satisfactorily for 2023. For malaria, the predicted and actual incidence trends were largely consistent, with prediction errors consistently below 0.01. Consequently, this underscores the need for enhanced data collection on factors influencing disease dynamics. This research provides valuable decision-making support, scientific insights, and theoretical guidance for enhancing disease monitoring and prediction strategies.

show abstract

Derivation of a prediction model for emergency department acute kidney injury

Cited by 2 publications

References 12 publications

Characterization of Risk Prediction Models for Acute Kidney Injury

Characterization of Risk Prediction Models for Acute Kidney Injury

Machine-learning-based analysis of biomedical time-series data: the monitoring and prediction of disease progression

Contact Info

Product

Resources

About