How does age affect baseline screening mammography performance measures? A decision model

Table 2 Breast Cancer Surveillance Consortium (BCSC) recent studies of screening mammography accuracy.

Parameter	Value^†	Range^‡	Ages	Period^§	Previous exams	Years
Sensitivity*	71.3%		< 40	1	No	1996–01
	53%			2
	82.1%		40–49	1	No
	61%			2
	92.1%		50–59	1	No
	68%			2
	89.3%		60–69	1	No
	66%			2
	68.2%		< 40	1	Yes
	70.7%		40–49	1	Yes
	78.1%		50–59	1	Yes
	79.7%		60–69	1	Yes
	88.6%	85.8–91.4	40–89	1	No	1996–00
	76.8%	75.7–77.9	40–89	1	Yes
	77.8%	77.6–78.0	40–89	1	Both
	65.3%		40–89	2	None past 5 yrs	1996–99
	53.8%		40–89	2	Yes
Specificity^\|\|	85.9%	85.6–86.1	40–89	1	No	1996–00
	92.5%	0.0	40–89	1	Yes
	92.1%	91.9–92.3	40–89	1	Both
	90.5%		40–89	2	None past 5 yrs	1996–99
	93.3%		40–89	2	Yes

* Positive exam is Breast Imaging Reporting and Data System (BI-RADS)[27] category 0, 4, 5, and category 3 if there is immediate diagnostic evaluation.[26]
^† The first four age-dependent one-year baseline sensitivity values are from reference [25]. The matching two-year values also used in the model are derived (see text). Matching one-year previous mammography sensitivity values are provided for comparison from reference [25].
^‡ Ranges are 95% confidence intervals available only for one-year accuracy values ages 40–89 from reference [26].
^§ Time frame for calculating sensitivity and specificity. Two-year accuracy values for ages 40–89 are from reference [24].
^|| The one-year specificity values are from reference [26]. The baseline specificity value is used in the model for all age groups.

ISSN: 1472-6947