Table 3. CAD (table summarises published studies on CAD in screening, including study design, quality features, and results).

Reference	Objectivea; comparator	Population (mean age in years)	Test evaluated 1 – In consecutive screenees 2 – Independent of comparator	No. of breast cancers (total sample size)	TPR or Sensitivity	FPR (based on recall rate)
Birdwell et al (2001) Burhenne et al (2000)	I only; M Views: 2 Reader: 1	Average-risk. CAD assessed on prior M in women later found to have breast cancer	1–No 2–No (yes for non-cancers)	542 (14 500)	M: 79% Incremental TPR of CAD=16%	FPR from before after study M: 8.3% CAD: 7.6%
Freer and Ulissey (2001)	I only; M Views: 2 Reader: 1	Average risk	1–Yes 2–No	49 (12 860)	M: 83.7% Incremental TPR of CAD=16.3%	M: 6.5% CAD: 7.7% Incremental FPR of CAD=1.2%
te Brake et al (1998)	I only; M Views: 1 Reader: 2	Women subsequently shown to have breast cancer	1–No 2–No	65 (207)	Incremental true positives reported (difficult to quantify)	Incremental false positives reported (difficult to quantify)

CAD=computer-aided diagnosis; M=mammography; ND=not described or unclear from paper; TPR=true-positive rate; FPR=false-positive rate.

^aObjective attainable in using the test: R=replacement for comparator, I=incremental to comparator; Views=number of mammography views per breast, Reader=number of mammography readers.