Table 3.

Robust linear regression predictors of a supine advantage.

Characteristic	Intercept	Coefficient	Standard error	Range supine better	Unit
Pre-dosimetry
Plasticity (pendulousness)
Ratio breast depth prone/supine	− 0.056	− 0.272	0.222	< − 0.2, NAP	Ratio
Breast depth difference prone − supine	− 0.203	− 0.073	0.030	< − 2.8, NAP	cm
Breast depth prone	− 0.200	− 0.031	0.015	< − 6.5, NAP	cm
Breast/body
Breast volume/body weight ratio	− 0.358	− 0.026	0.016	< − 13.7, NAP	mL/kg
Tumor location
Lower inner quadrant vs. else	− 0.537	− 0.164	0.313	< − 3.3, NAP	Binary
Breast size
Right breast volume supine	− 0.362	− 0.320	0.183	< − 1.1, NAP	L
Left breast volume supine	− 0.406	− 0.191	0.157	< − 2.1, NAP	L
Breast depth supine	− 0.316	− 0.034	0.025	< − 9.3, NAP	cm
Inspiration breath-hold capability
Right Lung volume supine	− 0.776	0.154	0.182	> 5.0, NEN	L
Total Lung volume supine	− 0.778	0.086	0.098	> 9.1, NEN	L
Left Lung volume supine	− 0.745	0.166	0.194	> 4.5, NEN	L
Age	− 0.651	0.002	0.005	> 338, NEN	Years
Left anterior descending coronary-chest wall distance	− 0.565	0.021	0.087	> 26.9, NEN	cm
Body size
Weight	− 0.274	− 0.004	0.004	< − 63.8, NAP	kg
Body Mass Index	− 0.185	− 0.015	0.013	< − 12.4, NAP	kg/m2
Heart volume supine	− 0.999	0.878	0.598	> 1.1, NEN	L
Height	− 3.028	1.514	0.943	> 2.0	m
Other
Couch type Varian vs. Bionix	− 0.524	− 0.069	0.142	< − 7.6, NAP	Binary
Preference prone vs. else	− 0.586	0.155	0.160	> 3.8, NAP	Binary
Post-dosimetry
Penalty prone	− 1.619	0.563	0.096	> 2.9	PoDP
Penalty supine	1.290	− 0.763	0.054	< 1.7	PoDP

Range Supine Better computed from the intercept and coefficient to predict when the penalty score changes in favor of supine. NAP not attainable physically, e.g. length < 0. NEN not expected normally, e.g. age > 338 years.

PoDP percent of dose prescription.