Table 3.

Univariate analyses: Variables significantly associated with the interim or final classification after reviewing prior mammograms.

A. UNIVARIATE ANALYSES: VARIABLES SIGNIFICANTLY ASSOCIATED WITH THE INTERIM CLASSIFICATION (STEP 2) AFTER REVIEWING PRIORS WITH INDEX IMAGES PRESENT.
VARIABLE & CLASSES	WITHOUT SUSPICIOUS LESIONS	MINIMAL SIGNS	CLEARLY VISIBLE TUMOURS	TOTAL	PEARSON CHI-SQUARE
	97	88	25	210
Need of arbitration on prior imaging					0.002
No arbitration	93 (49%)	78 (41%)	18 (10%)	189
Arbitration	4 (19%)	10 (48%)	7 (33%)	21
Date of prior imaging					<0.001
Earliest tertile	32 (67%)	15 (31%)	1 (2%)	48
Intermediate tertile	40 (51%)	31 (39%)	8 (10%)	79
Latest tertile	25 (30%)	42 (51%)	16 (19%)	83
B. UNIVARIATE ANALYSES: VARIABLES SIGNIFICANTLY ASSOCIATED WITH THE FINAL CLASSIFICATION (STEP 3) AFTER REVIEWING PRIORS WITH INDEX IMAGES AND CLINICAL INFORMATION PRESENT.
VARIABLE & CLASSES	WITHOUT SUSPICIOUS LESIONS	MINIMAL SIGNS	CLEARLY VISIBLE TUMOURS	TOTAL	PEARSON CHI-SQUARE
	94	77	39	210
Need of arbitration on prior imaging					0.004
No arbitration	90 (48%)	69 (37%)	30 (16%)	189
Arbitration	4 (19%)	8 (38%)	9 (43%)	21
Date of prior imaging					<0.001
Earliest tertile	31 (65%)	15 (31%)	2 (4%)	48
Intermediate tertile	39 (49%)	27 (34%)	13 (17%)	79
Latest tertile	24 (29%)	35 (42%)	24 (29%)	83
Image quality at the tumour side					0.004
Good	76 (43%)	59 (38%)	21 (14%)	156
Not good	18 (33%)	18 (33%)	18 (33%)	54
Detector system used					0.036
Computed Radiography CR	29 (41%)	22 (31%)	20 (28%)	71
Direct Readout Digital Radiography DR	65 (47%)	55 (40%)	19 (14%)	139