The “CROMa” Project: A Care Pathway for Clinical Management of Patients with Bisphosphonate Exposure

Table 5

Logistic regression model (forward stepwise procedure) for BRONJ presence, stratified for gender.

(a) Male

Variable	OR	Robust (SE)		95% CI
Behavioral habits	.92	.15	0.62	1.07–1.39
Therapy intervals	3.14	1.01	<.01	1.68–5.89

Number of observations= 61; log likelihood= −13.27; = 24.50; value < .01.

(b) Female

Variable	OR	Robust (SE)		95% CI
Behavioral habits	1.22	.08	<.01	1.07–1.39
Oncology bone disease	17.90	14.03	<.01	3.85–83.25
Therapy regimen	2.85	1.41	.03	1.08–7.50
Therapy intervals	2.37	.73	<.01	1.29–4.32

Number of observations= 403; log likelihood= −60.03; = 33.10; value < .01.