High-risk vs. low-risk groups | ||||
---|---|---|---|---|
Parameter | Percentages of LOO-CV loops | P (log-rank test) | Hazard ratio | 95% CI |
1-parameter best classifier (all 92 eligible patients who started with definitive chemoradiotherapy) | ||||
Total dose | 99 | 0.036 | 1.62 | 1.03–2.55 |
TLGmax(post) | 1 | |||
1-parameter TLGmax(post) classifier (all 92 eligible patients who started with definitive chemoradiotherapy) | 100 | 0.036 | 1.60 | 1.03–2.51 |
2-parameter best classifier (all 92 eligible patients) | ||||
Total dose | 100 | 0.0006 | 2.17 | 1.38–3.42 |
TLGmax(post) | 99 | |||
3-parameter best classifier (all 92 eligible patients) | ||||
Total dose | 100 | 0.0069 | 1.84 | 1.17–2.89 |
TLGmax(post) | 99 | |||
cT4 | 88 | |||
4-parameter best classifier (all 92 eligible patients) | ||||
Total dose | 100 | 0.017 | 1.73 | 1.10–2.71 |
TLGmax(post) | 99 | |||
cT4 | 95 | |||
cN2/N3 | 93 | |||
2-parameter best classifier (all 85 patients who received at least 60-Gy total dose) | ||||
Total dose | 100 | 0.0026 | 2.05 | 1.27–3.30 |
TLGmax(post) | 99 |
Percentages of LOO-CV (leave-one-out cross-validation) loops indicate consistency with which parameter is selected into best n parameter model across all LOO iteration loops.
Endpoint is overall survival