Development and Temporal Validation of Machine Learning Models for Hyponatremia Risk in Community-Dwelling Older Adults: A Nationwide Claims-Based Study.
LightGBM model predicted hyponatremia risk in older adults with AUROC 0.746
Development and Temporal Validation of Machine Learning Models for Hyponatremia Risk in Community-Dwelling Older Adults: A Nationwide Claims-Based Study.
Hyponatremia is a clinically important electrolyte disorder in older adults, yet early identification is hindered by complex, non-linear interactions between comorbidities and polypharmacy.
This study aimed to develop and externally validate a machine learning (machine learning) prediction model for hyponatremia risk using nationwide claims data, focusing on medication patterns and clinical features.
A retrospective cohort study was conducted using the South Korean Health Insurance Review and Assessment Service-Aged Patient Sample (HIRA-APS).
correctly separates high risk from low risk about three times in four, on a 0.5-1.0 scale
The development and validation cohorts included 4810 and 648,586 patients, respectively.
The models had very high negative predictive values (>0.999) for ruling out low-risk individuals.
Tree-based ensemble matched linear models in discrimination but achieved better calibration.
These validated, interpretable machine learning models can serve as clinical decision support tools that rule out low-risk patients and prioritize monitoring for high-risk individuals.
Across sociodemographic subgroups, calibration was maintained after recalibration, whereas discrimination was lower in the oldest, most comorbid, frailest, highest-medication-burden, and lowest-socioeconomic groups-a gap to address before equitable deployment.