Advanced gastrointestinal stromal tumor: reliable classification of imatinib plasma trough concentration via machine learning

Table 3 Forward stepwise binary logistic regression analysis

Variable	R	SE	OR (95% CI)	p
RBC (*10¹²/L)	-1.148	0.174	0.317 (0.226–0.446)	< 0.001
NEU%	0.026	0.008	1.027 (1.010–1.043)	0.001
PLT	0.003	0.001	1.003 (1.000–1.006)	0.029
Gender (“Female” for reference)	-0.691	0.207	0.501 (0.334–0.752)	< 0.001
Metastatic site (“Liver” for reference)	0.407	0.181	1.502 (1.054–2.140)	0.024
Daily IM dose (“ ≤ 200mg/d” for reference)				0.007
300mg/d	1.158	0.596	3.185 (0.991–10.235)
400mg/d	1.779	0.546	5.923 (2.030–17.281)
500mg/d	1.556	0.631	4.740 (1.375–16.341)
≥ 600mg/d	2.014	0.732	7.490 (1.784–31.445)

Through forward stepwise binary logistic regression analysis, variables such as HB, LYM%, and age at diagnosis are excluded
Abbreviations: NEU% Percentage of neutrophils, HB Hemoglobin, RBC Red blood cell count, LYM% Percentage of lymphocytes, PLT Platelet count, R Regression coefficient, SE Standard error, OR Odds ratio

ISSN: 1471-2407