ReviewComputer aids and human second reading as interventions in screening mammography: Two systematic reviews to compare effects on cancer detection and recall rate
Introduction
In many countries, including the UK, it is standard practice for each screening mammogram to be viewed independently by two readers who either confer on discordant cases or refer them for arbitration. It is sometimes argued that this ‘double reading’ is too expensive or too demanding of radiologists’ time.1 An alternative is to use computer programs that process digitised mammograms and alert readers to possible abnormalities. A systematic review identified six studies comparing computer aids (CAD) to double reading but concluded that they were methodologically flawed and the evidence was limited.2 This paper takes a different approach: two sets of studies are reviewed:
- •
studies comparing single reading with CAD to single reading without CAD;
- •
studies comparing double reading to single reading.
We assess the impact of both interventions on cancer detection and recall rate since an improvement in cancer detection rate at the cost of an increased recall rate may not present an enhancement of the screening test.
Section snippets
Types of studies
Prospective and retrospective studies where the intervention was incorporated into routine screening work and all cases selected only on the basis of the usual screening criteria were included.
Types of participants
All the studies of women in a screening age range (aged 40 and above) were considered.
Types of interventions
Only the studies using commercially available CAD systems were included. The studies of double reading in which the second reader was a trained film reader but not a radiologist were included.
Types of outcome measures
Only the studies reporting
Description of studies
Ten prospective studies comparing single reading with CAD to single reading without CAD were identified,6, 7, 8, 9, 10, 11, 12, 13, 14, 15 and 17 comparing single reading with double reading.14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30
Impact of CAD and double reading on cancer detection and recall rate
Matched CAD studies measure its impact more directly, comparing assessments on individual images before and after looking at prompts. Although all of these studies show an improvement in cancer detection rate, none shows a statistically significant improvement and their combined effect is not statistically significant. Since it is impossible to detect fewer cancers after taking a second look than that were detected initially, these studies are biased in favour of CAD. They will not detect if
Conclusion
There is evidence that double reading increases cancer detection rate and that double reading with arbitration does so whilst lowering recall rate. There is insufficient evidence to claim that CAD improves cancer detection rates, but it does increase recall rate. Comparing CAD and double reading with arbitration, there is no difference in cancer detection rate, but double reading with arbitration shows a significantly better recall rate. Therefore, the best current evidence shows grounds for
Conflict of interest statement
None declared.
Acknowledgements
Dr. Given-Wilson provided advice. Dr. Liston supplied additional data. The work was partly supported by the NHS Breast Screening Programme which had no input into the research or presentation of results.
References (60)
- et al.
Just how valuable is double reporting in screening mammography?
Clin Radiol
(1997) - et al.
Does the accuracy of single reading with CAD (computer-aided detection) compare with that of double reading?: a review of the literature
Clin Radiol
(2006) - et al.
Effectiveness and cost-effectiveness of double reading of mammograms in breast cancer screening: findings of a systematic review
Breast
(2001) - et al.
Double reading of mammography screening films – one radiologist or two?
Clin Radiol
(1993) - et al.
Impact of independent double reading of mammograms from the inception of a population-based breast cancer screening programme
Breast
(1995) - et al.
Can the NHS Breast Screening Programme afford not to double read screening mammograms?
Clin Radiol
(2003) - et al.
Scottish experience of double reading in the National Breast Screening Programme
Breast
(1998) - et al.
The efficacy of double reading mammograms in breast screening
Clin Radiol
(1994) - et al.
A pilot evaluation of the R2 image checker system and users’ response in the detection of interval breast cancers on previous screening films
Clin Radiol
(2001) - et al.
Can radiographers read screening mammograms?
Clin Radiol
(2003)
Just how valuable is double reporting in screening mammography?
Clin Radiol
Double reading of screening mammograms: the use of a third reader to arbitrate on disagreements
Breast
Cancers detectes par le deuxieme lecteur: analyse des donnees de la campagne de depistage du cancer du sein en Loire-Atlantique, 2003–2005 (nouveau cahier des charges)
J Radiol
Marginal modeling of binary cross-over data
Biometrics
metan – an alternative meta-analysis command
STATA Tech Bull Reprints
Screening mammograms: interpretation with computer-aided detection – prospective evaluation
Radiology
Screening mammography with computer-aided detection: prospective study of 12,860 patients in a community breast center
Radiology
Improved cancer detection using computer-aided detection with diagnostic and screening mammography: prospective study of 104 cancers
Am J Roentgenol
Prospective assessment of computer-aided detection in interpretation of screening mammography
Am J Roentgenol
Computer-aided detection with screening mammography in a university hospital setting
Radiology
Influence of computer-aided detection on performance of screening mammography
N Engl J Med
Changes in breast cancer detection and mammography recall rates after the introduction of a computer-aided detection system
J Natl Cancer Inst
Impact of computer-aided detection in a regional screening mammography program
Am J Roentgenol
Blinded comparison of computer-aided detection with human second reading in screening mammography
Am J Roentgenol
Comparison of computer-aided detection to double reading of screening mammograms: review of 231,221 mammograms
Am J Roentgenol
Mammography screening: an incremental cost effectiveness analysis of double versus single reading of mammograms
BMJ
Independent double reading of screening mammograms in The Netherlands: effect of arbitration following reader disagreements
Radiology
Comparison of radiographer/radiologist double film reading with single reading in breast cancer screening
J Med Screen
Medical radiologic technologist review: effects on a population-based breast cancer screening program
Radiology
Principes et premiers résultats de la campagne européenne de dépistage du cancer du sein dans le Bas-Rhin
Bull Acad Natl Med
Cited by (122)
Technical/Algorithm, Stakeholder, and Society (TASS) barriers to the application of artificial intelligence in medicine: A systematic review
2023, Journal of Biomedical InformaticsSecond reading in breast cancer screening program: State of knowledge and future
2022, Bulletin du CancerBone and Soft Tissue Tumors: Horizons in Radiomics and Artificial Intelligence
2022, Radiologic Clinics of North America