Predictive Modeling
Conditions
Keywords
Lab Result Communication, Risk Stratification
Brief summary
This study aims to improve how lab results are communicated to older adults by refining a predictive model that uses electronic health record (EHR) data. The model was originally developed to estimate the risk of chronic kidney disease (CKD) progression. Researchers will use existing health data to test and improve the accuracy of the model and explore how it might be adapted for use in other health conditions. The study does not involve direct interaction with patients and is conducted entirely using de-identified data in a secure environment.
Interventions
This study analyzes retrospective electronic health record (EHR) data from older adults to refine and validate a predictive model for other conditions in future studies.
Sponsors
University of California, Los Angeles
Study design
Observational model
COHORT
Time perspective
RETROSPECTIVE
Eligibility
Sex/Gender
ALL
Age
65 Years to No maximum
Healthy volunteers
No
Inclusion criteria
include, but are not limited to: * being over the age of 65; having at least 5 years of clinical follow up; and having a serum creatinine lab test conducted
Exclusion criteria
* Patients younger than 65 years old * Patients with less than 5 years of clinical follow-up * Patients from health systems outside of the UC Health network.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Performance of the Risk Prediction Model | Up to 5 years of retrospective follow up | Evaluate the predictive performance of a machine learning-based risk model using retrospective Electronic Health Records (EHR) data. The model estimates the likelihood of disease progression in older adults. The model should be designed to be adaptable to various clinical conditions. Metrics include Area Under the Receiver Operating Characteristic Curve (AUC-ROC), sensitivity, and specificity. |
Countries
United States
Contacts
CONTACTKatelyn Nguyen
Outcome results
None listed