. Author manuscript; available in PMC: 2009 Jul 30.

Published in final edited form as: JAMA. 2008 May 14;299(18):2151–2163. doi: 10.1001/jama.299.18.2151

Table 3.

Sensitivity, Specificity, and PPV by Participant for Mammography Alone (M) vs. Combined Mammography + Ultrasound (M+US)

		N Participants	N Breast Cancers	Cumulative N Participants	Cumulative No. of True Positives	Sensitivity		Specificity	PPV: Participant Level	No. of Participants with either a Core or Surgical Biopsy^a	No. of Lesions with either a Core or Surgical Biopsy	No. of Participants with only a Cyst Aspiration	No. of Lesions with only a Cyst Aspiration	No. of Participants with an Atypical^b Biopsy	No. of Atypical Lesions that were Biopsied
		N Participants	N Breast Cancers	Cumulative N Participants	Cumulative No. of True Positives	All Cancers	Invasive Cancers	Specificity	PPV: Participant Level	No. of Participants with either a Core or Surgical Biopsy^a	No. of Lesions with either a Core or Surgical Biopsy	No. of Participants with only a Cyst Aspiration	No. of Lesions with only a Cyst Aspiration	No. of Participants with an Atypical^b Biopsy	No. of Atypical Lesions that were Biopsied
M	BI- RADS Assessment
	5	5	4	5	4	0.1000	0.0882	0.9996	0.8000	5	5	0	0	0	0
	4C	9	4	14	8	0.2000	0.2059	0.9977	0.5714	9	12	0	1	1	1
	4B	32	7	46	15	0.3750	0.3235	0.9881	0.3261	22	24	1	2	0	0
	4A	90	5	136	20^c	0.5000	0.4412	0.9553	0.1471	33	43	3	6	2	2
	3	177	1	313	21	0.5250	0.4706	0.8876	0.0671	16	19	12	16	0	0
	2^d	1421	14	1734	35	0.8750	0.8824	0.3458	0.0202	112	131	37	59	7	7
	1	903	5	2637	40	1.0000	1.0000	0.0000	0.0152	70	85	18	29	6	8
M+US
	5	6	5	6	5	0.1250	0.1176	0.9996	0.8333	6	8	0	1	0	0
	4C	26	12	32	17	0.4250	0.4706	0.9942	0.5313	26	35	0	0	2	2
	4B	68	10	100	27	0.6750	0.6765	0.9719	0.2700	58	73	5	9	4	4
	4A	206	4	306	31	0.7750	0.7353	0.8941	0.1013	112	128	42	74	6	6
	3	401	4	707	35^c	0.8750	0.8529	0.7412	0.0495	40	49	17	22	3	5
	2	98	0	805	35	0.8750	0.8529	0.7035	0.0435	3	3	1	1	0	0
	1^d	1832^e	5	2637	40	1.0000	1.0000	0.0000	0.0152	22	23	6	6	1	1

Includes 1 prophylactic mastectomy which showed DCIS in a patient with contralateral invasive ductal cancer; does not include two negative double prophylactic mastectomies nor 7 other single prophylactic mastectomies.

Atypical includes atypical ductal (ADH) or lobular hyperplasia (ALH), LCIS, atypical papilloma, and radial sclerosing lesion

One 4 mm invasive ductal cancer was considered BI-RADS 4a on M, missed on US, and classified as BI-RADS 3 on integrated M+US; this was diagnosed when the participant presented with palpable metastatic adenopathy 264 days after study entry.

The one participant with cancer due to melanoma presenting with axillary metastases was classified as BI-RADS 2 on M and BI-RADS 1 on integrated M+US.

Investigators were not asked to integrate findings on M+US if both exams recommended routine follow up: these cases were analyzed as “negative” on M+US. Of 1832 cases analyzed as BI-RADS 1 on M+US, 342 were BI-RADS 2 on mammography, 318 were BI-RADS 2 on US, 747 were BI-RADS 2 on both; 2 were BI-RADS 3 on initial mammography and downgraded to BI-RADS 1 after integration with US; the remaining 423 were BI-RADS 1 on both M and US.