(HealthDay News) — The cardiothoracic ratio (CTR) has real-world prognostic value for chronic kidney disease (CKD), according to a study published online in Communications Medicine.
Che-Yi Chou, from Asia University Hospital in Taiwan, and colleagues developed a machine learning-based algorithm to calculate the baseline and serial CTRs, which were then used to classify patients into trajectory groups. The analysis included 3117 patients with CKD (aged 18 to 89 years).
The researchers identified 3 CTR trajectory groups (low, 30.1%; medium, 48.1%; high, 21.8%). The adjusted hazard ratios for end-stage renal disease, cardiovascular mortality, and all-cause mortality in patients with baseline CTRs ≥0.57 were 1.35, 2.89, and 1.50, respectively, versus CTRs <0.47. Greater effect size was also seen for high versus low CTR trajectories, particularly for cardiovascular mortality. A model with CTR as a continuous variable yields significantly higher C-statistics of 0.719 compared with a reference model (0.698) for cardiovascular mortality and 0.697 (versus 0.693) for all-cause mortality.
“Our findings support the real-world prognostic value of the CTR, as calculated by a machine learning annotation tool, in CKD,” the authors write. “Our research presents a methodological foundation for using machine learning to improve cardioprotection among patients with CKD.”