Table 3.

Unadjusted and adjusted analyses

	Vaginal SLT				Skeletal SLT
	Unadjusted		Adjusted		Unadjusted		Adjusted
Parameter	HR (95% CI)	P	HR (95% CI)	P	HR (95% CI)	P	HR (95% CI)	P
Treatment		.027		<.001		.023		.016
CRT vs RT	1.5 (1.1–2.3)		3.0 (1.7–5.2)		5.83 (1.28–26.7)		7.0 (1.4–34.1)
Dilator compliance		<.001		<.001		.709		NA
High	RG		RG
Moderate	3.0 (1.8–5.2)		3.6 (2.0–6.5)	<.001
Poor	5.0 (2.9–8.7)		8.5 (4.3–16.9)	<.001
Age		.001		.013		.049		.028
>50 vs ≤50	2.0 (1.3–2.9)		1.8 (1.1–3.0)		3.7 (1.01–13.7)		5.7 (1.2–27.0)

Abbreviations: CI = confidence interval; HR = hazard ratio; NA = not applicable; RG = reference group.

Unadjusted analysis used Cox proportional hazards model. Adjusted analysis used Cox proportional hazards model built with stepwise selection procedure. Covariate factors included histology, body mass index, race, cigarette smoking history, history of hypertension, history of diabetes, history of intestinal disorder, history of abdominopelvic surgery, pre-radiotherapy nodal dissection, FIGO stage, pelvic EBRT biologically equivalent dose in 2-Gy fractions (EQD2) (Gy3), paraortic nodal EBRT, brachytherapy dose rate, brachytherapy applicator type, pelvic EBRT/brachytherapy EQD2 (Gy3). Entry criterion was set at a P value <.20 on unadjusted log-rank analysis. All statistically significant P values on multivariate analysis were confirmed with Cox proportional hazards model with all covariates entered (data not shown).