A Test With Inherent Limitations

December 31, 2006

Mammography is auseful tool. However,it has limitations. Until recently,it had been viewedas the ultimate diagnostictest, capable of detecting allbreast cancers in their earliest,treatable stages. Thissimply is not true. Evenwhen performed by themost capable institutionsand radiologists, mammographyhas a sensitivity of80% to 85% for the detectionof breast cancer. This hasbeen established by numerousstudies.1,2

Mammography is auseful tool. However,it has limitations. Until recently,it had been viewedas the ultimate diagnostictest, capable of detecting allbreast cancers in their earliest,treatable stages. Thissimply is not true. Evenwhen performed by themost capable institutionsand radiologists, mammographyhas a sensitivity of80% to 85% for the detectionof breast cancer. This hasbeen established by numerousstudies.1,2Further evidence ofthe imprecision of mammographyis the fact thatthere is no shortage of newexperimental methods toaid in the detection ofbreast cancer. Researchershave tried to improve onthe sensitivity of mammographyby the use of lasers,heat detection, radiotracers,and other technologies.3-6The sensitivity of 80%to 85% is not a reflection ofthe radiologist's skill;rather, it is a limitation ofthe test itself. One studyfound that this sensitivityrange remained constanteven when different groupsof radiologists interpretedthe images.7Both the governmentand radiologic agencieskeep a watchful eye on thework of radiologists whoread mammograms. Mammographyis one of themost strictly regulatedareas of radiology--andperhaps all of medicine.Stringent standards dictatewhich type of film to use,how often to clean thedarkroom, how much continuingeducation readingphysicians must have, thevolume of films one mustread to become a certifiedreader, and more. Lawsuitsagainst institutions and individualphysicians whoperform mammographyprovide additional pressureto do high-qualitywork.Mammography simplyhas inherent limitations.Keep in mind, however, thatthere are very few diagnostictests in medicine thathave a sensitivity of 100%.

References:

REFERENCES:


1.

Harvey SC, Geller B, OppenheimerRG, et al. Increase in cancer detectionand recall rates with independent doubleinterpretation of screening mammography.

Am J Roentgenol

. 2003;180:1461-1467.

2.

Warren Burhenne LJ, Wood SA,D’Orsi CJ, et al. Potential contributionof computer-aided detection to thesensitivity of screening mammography.

Radiology

. 2000;215:554-562.

3.

Sampalis FS, Denis R, Picard D, etal. International prospective evaluationof scintimammography with(99m) Technetium sestamibi.

Am JSurg

. 2003;185:544-549.

4.

Riefke B, Licha K, Semmler W.Contrast media for optical mammography[in German].

Radiologe

. 1997;37:749-755.

5.

Davies MJ. Fluorescent cancer detection.

Trends Biotechnol

. 2001;19:162.

6.

Boehm T, Hochmuth A, Malich A,et al. Contrast-enhanced near-infraredlaser mammography with a prototypebreast scanner: feasibility study withtissue phantoms and preliminary resultsof imaging experimental tumors.

Invest Radiol

. 2001;36:573-581.

7

. Anderson ED, Muir BB, Walsh JS,Kirkpatrick AE. The efficacy of doublereading mammograms in breastscreening.

Clin Radiol

. 1994;49:248-251.