HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2392050670v1 239/3/710    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lindfors, K. K. Articles by Hurlock, G. S. Search for Related Content PubMed PubMed Citation Articles by Lindfors, K. K. Articles by Hurlock, G. S.

Computer-aided Detection of Breast Cancer: A Cost-effectiveness Study¹

¹ From the Department of Radiology, University of California, Davis School of Medicine, Suite 3100, 4860 Y St, Sacramento, CA 95817. From the 2004 RSNA Annual Meeting. Received April 21, 2005; revision requested July 11; revision received June 23; final version accepted July 27. Address correspondence to K.K.L. (e-mail: kklindfors{at}ucdavis.edu).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Purpose: To analyze the cost-effectiveness of adding computer-aided detection (CAD) to a screening mammography program.

Materials and Methods: A Markov model was developed to compare three hypothetical groups of women aged 40–79 years. The first group was composed of women undergoing mammographic screening without CAD; the second, of women undergoing mammographic screening with CAD; and the third, of women undergoing observation without screening. Cost-effectiveness was expressed as the marginal cost per year of life saved (MCYLS). MCYLS was calculated for screening mammography with CAD compared with screening mammography alone and for screening mammography alone compared with observation. Sensitivity analyses were performed by varying the cost of CAD, the rates of cancer detection with CAD, and the stage distribution of breast cancers diagnosed with CAD.

Results: Adding CAD to a mammographic screening program resulted in a MCYLS of $19 058. The MCYLS of screening mammography alone compared with observation was $16 023. CAD increases the marginal effectiveness of screening by 29%; however, the marginal cost of screening is also increased by 34%. Varying the cost of CAD yields a linear increase in MCYLS from $8937 with CAD at $9 per case to $24 924 with CAD at $25 per case. The cost-effectiveness of CAD is dependent on the magnitude of the increase in cancer detection rates with CAD but is also affected by the stage distribution of cancers diagnosed with CAD.

Conclusion: The MCYLS is 19% greater for CAD added to screening versus screening mammography alone but is still within the accepted range for cost-effectiveness.

© RSNA, 2006

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Although screening mammography can substantially reduce breast cancer mortality, its sensitivity in demonstrating disease is only 65%–90% (1–3). The use of computer-aided detection (CAD) as an adjunct to screening mammography has been shown to increase breast cancer detection rates by up to 23% (4,5).

In 1998, the U.S. Food and Drug Administration approved the first commercially available CAD system for screening mammography. It is estimated that 17% of all Mammography Quality Standard Act–approved mammography sites are now using CAD and that there are about 1500 CAD units in use across the United States. Approximately 25%–30% of all screening mammograms are interpreted with CAD assistance (K. O'Shaughnessy, PhD, written communication, October 11, 2004), but to our knowledge no studies have been published on the cost-effectiveness of CAD. Thus, the purpose of our study was to analyze the cost-effectiveness of adding CAD to a screening mammography program.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Study Model
In this study, a Markov model was developed to compare three hypothetical populations of women. The first group was composed of women undergoing mammographic screening for breast cancer with CAD; the second, of women undergoing mammographic screening without CAD; and the third, of women undergoing observation without screening. The societal perspective was used for this analysis; all costs were counted regardless of who incurred them, and all benefits were included regardless of who received them. A commercial software product (Data 3.5 for Healthcare; TreeAge Software, Williamstown, Mass) was used to generate the model and to tabulate the costs and benefits accrued by each group. Each hypothetical group was followed up from age 40–79 years. In the screening groups, mammography (with or without CAD) was assumed to take place annually.

In the groups that underwent mammography, decision trees were constructed to encompass all possible outcomes of screening and subsequent diagnostic work-ups (Fig 1a). A similar tree was constructed for the group of patients who underwent observation without screening and included the development of a palpable breast mass with subsequent diagnostic examinations (Fig 1b).

View larger version (39K): [in this window] [in a new window] [Download PPT slide]   Figure 1a: Transition diagram of Markov decision tree model used for (a) groups undergoing screening mammography with or without CAD and (b) group undergoing observation without screening. Terminal states for each cycle (year) are shown in bold.

View larger version (33K): [in this window] [in a new window] [Download PPT slide]   Figure 1b: Transition diagram of Markov decision tree model used for (a) groups undergoing screening mammography with or without CAD and (b) group undergoing observation without screening. Terminal states for each cycle (year) are shown in bold.

Reference Case
For cost-effectiveness analyses, a reference case is used to calculate the principal or baseline cost-effectiveness ratio. The values chosen for each variable in the reference case are those that are thought to most closely reflect actual practice. The reference case provides a baseline for comparison across different studies and for sensitivity analyses in the same study.

For this analysis, patient survival was based on the stage at diagnosis, which differed for women whose cancers were diagnosed at mammography, those that were diagnosed at mammography with CAD, or those that were diagnosed at observation (Table 1). For the observation group, the total percentage of stage II cancers was 39.0%; stage II cancers were not split into stage IIA and stage IIB in this group. Stage-specific survival rates were applied to each population (Table 2); in the observation group, the survival rate that was used for stage II cancers corresponded to the mean for stage IIA and stage IIB cancers combined, as reported in Table 2. Women who died of breast cancer were assumed to survive an average of 5 years with the disease. Women with negative mammographic findings could have immediate false-negative findings or could develop interval breast cancers. The distribution of the stage of breast cancer at diagnosis was assumed to be the same as in the observation group for both immediate false-negative results and interval cancers. At the end of each cycle (1 year), each hypothetical woman was determined to be in one of the following categories: alive without breast cancer, alive with breast cancer, dead from breast cancer, or dead from other causes.

Deaths from causes other than breast cancer were determined by using age-specific death rates (12) and by subtracting the calculated cancer death rates for that age group.

Model Validation
To validate the probabilities used for the outcomes in each arm of the model, the cancer incidence in each of the three groups (observation, mammography alone, and mammography with CAD) was calculated by using a Monte Carlo simulation of 32 000 women entering each of the three groups and cycling through the tree for 39 cycles (years). The percentage of women in each group who had breast cancer was calculated and compared both across groups and with published incidence figures.

The probabilities of death from breast cancer after 39 cycles (years) of the model were calculated for the observation, mammography alone, and mammography with CAD groups. The reductions in breast cancer mortality between groups were compared with published figures. The probability of survival by age, as predicted with the model, was compared with figures reported for the population of the United States in 2003.

The total percentage of women who died and the total percentage of women who were alive after 39 cycles were also calculated for each group. These percentages were added to ensure that 100% of the women were accounted for in the model.

Cost-effectiveness Analysis
Cost-effectiveness was expressed as the marginal cost per year of life saved (MCYLS) and was calculated by dividing the marginal cost (the difference in costs between comparison groups) by the marginal effectiveness (the difference in years of life accumulated between comparison groups).

For the reference case, the MCYLS was calculated for screening mammography with CAD compared with screening mammography alone; MCYLS was also calculated for screening mammography alone compared with observation alone.

The costs of breast imaging and interventional procedures were based on the 2003 average global Medicare reimbursements (Table 4). It was assumed that 75% of women who returned for diagnostic evaluation would undergo unilateral diagnostic mammography and that 25% would undergo ultrasonography (US); costs were prorated accordingly. With respect to core biopsies, it was assumed that 60% of biopsies would be performed by using US guidance with an automated gun and that 40% would be performed by using a stereotactic vacuum-assisted technique with placement of a clip; costs were prorated accordingly. The cost of treatment according to cancer stage was based on the results of current studies (Table 2).

Sensitivity Analyses
In the reference case, the breast cancer detection rate for mammography with CAD was 3.8 per 1000 women. Two different sensitivity analyses using cancer detection rates from 2.8 per 1000 women to 3.6 per 1000 women were performed. In the first of these analyses, it was hypothesized that the decrease in cancer detection rate for mammography with CAD compared with the reference case was caused by a decrease in the ability of CAD to demonstrate lesions, which therefore resulted in a decrease in the number of recalls compared with the reference case. It was also hypothesized that when a lesion was identified with CAD the lesion was still more likely to be a carcinoma than if it had been detected with screening alone.

For this analysis, the probability of an interval cancer was increased as the cancer detection rate was decreased so that the total number of cancers per year was constant at 3.94 per 1000 women for the population. The recall rate for the CAD group was calculated for each cancer detection rate by using the formula: recall rate = CDR/PPV/PR, where CDR is the cancer detection rate, PPV is the positive predictive value of biopsy (0.38), and PR is the probability of identifying a BI-RADS 4 or 5 lesion with CAD (.133).

The second sensitivity analysis for cancer detection rates with CAD was based on the hypothesis that the decrease in cancer detection rate compared with the reference case was due to a decrease in the probability of identifying a BI-RADS 4 or 5 outcome after recall for abnormal screening mammographic results. In this analysis, the recall rate with CAD was constant at 0.077, but the probability of a BI-RADS category 4 or 5 outcome was varied with the cancer detection rate by using the formula: PR = CDR/RR/PPV, where RR is the recall rate (0.077). The interval cancer rates for this analysis were also increased as the cancer detection rates were decreased in order to keep the total number of cancers at 3.94 per 1000 women per year.

For both of the above-mentioned sensitivity analyses, the stage distributions of cancers that were diagnosed with each modality (mammography alone, mammography with CAD, and observation) and the number of immediate false-negative mammographic findings for each group were held constant, as in the reference case.

Because this cost-effectiveness analysis is heavily dependent on the stage distributions of breast cancers that are diagnosed by using the various methods, a sensitivity analysis with different stage distributions (Table 5) for the mammography alone versus mammography with CAD analysis was performed (T. E. Cupples, MD, written communication, October 5, 2004).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
Reference Case
The MCYLS for screening mammography alone was $16 023 and provides a basis for comparison with the MCYLS for screening mammography with CAD (Table 6). The addition of CAD to a mammographic screening program results in a MCYLS of $19 058. Adding CAD to a mammographic screening program increases the marginal effectiveness of screening by 29%; however, the marginal cost of the program is also increased by 34%.

The MCYLS increases linearly with the cost of CAD—that is, from a MCYLS of $8937 with CAD at $9 per case to a MCYLS of $24 924 with CAD at $25 per case (Fig 2).

View larger version (12K): [in this window] [in a new window] [Download PPT slide]   Figure 2: Linear relationship is demonstrated between MCYLS of screening mammography with CAD and the cost of CAD.

Model Validation
The total incidence of breast cancer for the 39 year period was 12.9% in the observation group, 13.1% in the mammography alone group, and 13.7% in the mammography with CAD group.

The probability of death from breast cancer after 39 cycles (years) was .0228 in the observation group and .0115 in the mammography alone group; thus the reduction in mortality with screening mammography alone was 50% in the model. The probability of death from breast cancer in the mammography with CAD group was .0084, which demonstrated a further reduction of 27% versus mammography alone (Table 9).

The percentages of women who were either dead or alive after 39 cycles (years) for the three comparison groups totaled 99.9% in the observation and mammography with CAD groups and 100% in the mammography alone group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 
In the hypothetical population of women in our study, the marginal cost of screening mammography alone for 39 cycles was $1318.43. If CAD is added to this screening program, the cost increases by $448.45 (34%) because of the cost of CAD itself and because of the increasing numbers of women who would be undergoing diagnostic imaging and biopsy. With this increase in cost, however, there is also an increased benefit in years of life saved owing to the increased cancer detection rate and in the downstaging of cancers diagnosed with CAD. Therefore, the cost-effectiveness ratio provides a more meaningful comparison. The MCYLS of screening mammography alone is $16 023. If CAD is added to a mammographic screening program, the MCYLS is $19 058, a 19% increase over the MCYLS for screening alone.

Because decisions about new health care programs involve consideration of both costs and benefits, these decisions are best made by performing a cost-effectiveness analysis. There is no firmly established monetary figure for determining that a medical procedure or intervention is cost-effective. In the United States, investigators involved in cost-effectiveness studies often suggest that $50 000–$100 000 per life year saved may represent the upper limit of appropriate expenditure. In our study, the MCYLS for screening added to observation ($16 023) and for CAD added to screening ($19 058) were both well below the $50 000–$100 000 limit. Therefore, these procedures would be acceptable medical interventions on the basis of cost-effectiveness criteria.

Our analysis shows that the cost-effectiveness of CAD is heavily dependent on the cost of CAD itself. In the reference case, the cost of CAD was predicated on the basis of CAD use with film-screen mammography, a labor-intensive system in which all images must be digitized. With the advent of digital mammography, many operating systems will integrate CAD, which will automatically be applied to the soft-copy display. This may reduce costs and increase the cost-effectiveness of CAD.

There are conflicting reports in the literature regarding the change in breast cancer detection rates with the use of CAD. When the same images that are interpreted without CAD and are then interpreted with CAD, the incremental increases in detection with CAD range from 9.0%–19.5% (4,20). When historical changes in breast cancer detection rates are compared before and then after the introduction of CAD, results are also mixed. Gur et al (18) reported no statistically significant increase in cancer detection rates after the introduction of CAD, yet Cupples (21) reported an increase of 13%. These differences in results suggest that CAD may be more useful and more cost-effective in certain practice settings. As shown in our analysis, the cancer detection rate with CAD has a substantial effect on cost-effectiveness. The magnitude of change in cancer detection rates can have a profound effect on the incremental cost-effectiveness of adding CAD to a screening mammography program, as we have shown.

CAD has also been shown to change the stage distribution for cancers detected with its use (4). In the study by Freer and Ulissey (4), the detection of stage 0 and stage I cancers increased by 4.4% when CAD was used. Detecting breast cancer at an earlier stage is a primary goal of CAD and would certainly have a substantial effect on effectiveness, even without a change in the actual cancer detection rate (as shown by the sensitivity analysis on stage distributions presented in our study). When stage 0 and stage I cancers were increased by 12.8% with CAD while the cancer detection rate was held constant, the marginal effectiveness was 2.6 times greater than the effectiveness in the reference case. The cost-effectiveness of CAD was increased by a factor of 2.4.

There are several inherent limitations in the model we used. Because there is no published information on the actual reduction in breast cancer mortality that can be expected when CAD is added to a screening program, a surrogate measure of mortality reduction was used to estimate benefit in this analysis. Published figures of stage distributions for cancers diagnosed with or without CAD, together with published breast cancer mortality rates according to stage, were used to simulate the reduction in mortality from breast cancer that could be expected when CAD is used. The calculated additional reduction in breast cancer mortality when CAD was added to a mammographic screening program was 27% in the reference case.

For each year of screening, the reference case in this analysis used constant cancer detection rates for mammographic screening alone and for mammographic screening with CAD. In the study by Freer and Ulissey (4), the cancer detection rate with CAD was 19.5% higher compared with mammography alone; however, the study presents data for only one screening with CAD. There have been no reports in the literature about cancer detection rates with CAD in subsequent years, but it is likely that this rate will diminish after the 1st year. The magnitude of this diminution is unknown but, if substantial, would reduce the effectiveness of CAD and thereby also reduce its cost-effectiveness.

It is also possible that the incremental benefit of CAD may diminish with time as interpreting radiologists become trained in the use of CAD. Lesions that would not have been identified prior to the institution of CAD might be more readily identified by the radiologist after he or she gains familiarity with the abnormalities marked by CAD. This would increase the effectiveness of screening alone and reduce the benefit of adding CAD to a screening program. The converse has also been suggested—that is, that radiologists would become less effective and more dependent on CAD and might overlook lesions not marked by CAD.

Another limitation of our study is that interval cancers were assumed to have the same stage distribution as cancers in the observation group, but this is likely an overestimate because some of these cancers would be discovered at the next screening. If the mortality rates for interval cancers are overestimated, the cost-effectiveness of CAD may be somewhat underestimated.

The measurement of effectiveness used in this study was years of life saved. Some authorities have suggested that benefit should be expressed in quality-adjusted life years instead (22), but there are limited data and no generally accepted standards on how participation in a mammographic screening program affects quality of life. There may be both negative (eg, false-positive mammographic and biopsy results) and positive (eg, the knowledge that no cancer is present or the establishment of improved health habits) effects of participation in mammographic screening. Furthermore, the cost per year of life saved and the cost per quality-adjusted life year are highly related to one another; both types of analyses will generally lead to similar decisions regarding resource allocation (23).

The 50% reduction in breast cancer mortality in this computerized model is comparable with the reduction in mortality reported in previous studies of service screening in Sweden (24,25). The total incidence of breast cancer from patients 40–79 years of age, as predicted in this model, is similar to the cumulative lifetime risk of in situ and invasive breast cancer (15.96%) reported by the National Cancer Institute (26). The probability of survival according to age, as predicted by the model, is similar to data reported for the population of the United States (19).

In conclusion, our analysis shows that the MCYLS for CAD added to screening mammography was $19 058, which is well within the range that is generally considered to be cost-effective. CAD not only adds to the cost of screening for breast cancer but also increases its effectiveness.

	ADVANCES IN KNOWLEDGE

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 

• Adding CAD to mammographic screening increases marginal effectiveness by 29% but increases marginal cost by 34%.
  
• CAD added to screening increases the marginal cost per year life saved by 19% compared with screening alone.
  
• The marginal cost per year of life saved for CAD added to screening is $19 058, which is in the range of an acceptable cost-effectiveness for a medical intervention.
  
• The marginal cost per year of life saved for CAD increases exponentially as the cancer detection rate decreases.
  
• The cost of CAD and the stage distribution of cancers diagnosed with CAD have a substantial effect on the cost-effectiveness of CAD.
  

	FOOTNOTES

 

Abbreviations: BI-RADS = Breast Imaging Reporting and Data System • CAD = computer-aided detection • MCYLS = marginal cost per year of life saved

² Current address: Department of Radiology, Brigham and Women's Hospital, Boston, Mass.

Authors stated no financial relationship to disclose.

Author contributions: Guarantor of integrity of entire study, K.K.L.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, all authors; experimental studies, all authors; statistical analysis, all authors; and manuscript editing, K.K.L., M.C.M., C.J.R.

	References

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCES IN KNOWLEDGE References

 

1. Kerlikowske K, Carney PA, Geller B, et al. Performance of screening mammography among women with and without a first-degree relative with breast cancer. Ann Intern Med 2000;133:855–863.[Abstract/Free Full Text]
2. Yankaskas BC, Cleveland RJ, Schell MJ, Kozar R. Association of recall rates with sensitivity and positive predictive values of screening mammography. AJR Am J Roentgenol 2001;177:543–549.[Abstract/Free Full Text]
3. Sickles EA. Quality assurance: how to audit your own mammography practice. Radiol Clin North Am 1992;30:265–275.[Medline]
4. Freer TW, Ulissey MJ. Screening mammography with computer-aided detection: prospective study of 12,860 patients in a community breast center. Radiology 2001;220:781–786.[Abstract/Free Full Text]
5. Brem RF, Baum J, Lechner M, et al. Improvement in sensitivity of screening mammography with computer-aided detection: a multi-institutional trial. AJR Am J Roentgenol 2003;181:687–693.[Abstract/Free Full Text]
6. Andersson I, Aspegren K, Janzon L, et al. Mammographic screening and mortality from breast cancer: the Malmö mammographic screening trial. BMJ 1988;297:943–948.[Abstract/Free Full Text]
7. American Cancer Society. Breast cancer facts and figures. Atlanta, Ga: American Cancer Society, 1999–2000; 7.
8. Legorreta AP, Brooks RJ, Leibowitz AN, Solin LJ. Cost of breast cancer treatment: a 4-year longitudinal study. Arch Intern Med 1996;156:2197–2201.[Abstract/Free Full Text]
9. Warren Burhenne LJ, Wood SA, D'Orsi CJ, et al. Potential contribution of computer-aided detection to the sensitivity of screening mammography. Radiology 2000;215:554–562. [Published correction appears in Radiology 2000;216:306.][Abstract/Free Full Text]
10. Baker LH. Breast cancer detection demonstration project: 5-year summary report. CA Cancer J Clin 1982;32:194–225.[Abstract/Free Full Text]
11. Seltzer MH. Breast complaints, biopsies, and cancer correlated with age in 10,000 consecutive new surgical referrals. Breast J 2004;10:111–117.[CrossRef][Medline]
12. U.S. Department of Commerce, Economics and Statistics Administration and Bureau of the Census. Vital statistics: statistical abstract of the United States. 111th ed. Washington, DC: U.S. Government Printing Office, 1991; 84.
13. U.S. Department of Health and Human Services. 2003 Physicians fee schedule, APC codes. Federal Register 2002;67:66717–66913.[Medline]
14. U.S. Department of Health and Human Services. Medicare program: revision to payment policies under the physician fee schedule for calendar year 2003. Federal Register 2002;67:79965–80184.[Medline]
15. U.S. Department of Health and Human Services. 2003 Physicians fee schedule. Federal Register 2003;68:9567–9580.[Medline]
16. Linver MN. Practice operations: coding and billing—2003 update. The Society of Breast Imaging 6th Postgraduate Course Syllabus, 2003; 45–54.
17. McDonough KL. Breast cancer stage cost analysis in a managed care population. Am J Manag Care 1999;5(suppl 6):S377–S382.[Medline]
18. Gur D, Sumkin JH, Rockette HE, et al. Changes in breast cancer detection and mammography recall rates after the introduction of a computer-aided detection system. J Natl Cancer Inst 2004;96:185–190.[Abstract/Free Full Text]
19. U.S. Census Bureau. Vital statistics: statistical abstract of the United States. Washington, DC: U.S. US National Center for Health Statistics, 2003;84.
20. Helvie MA, Hadjiiski L, Makariou E, et al. Sensitivity of noncommercial computer-aided detection system for mammographic breast cancer detection: pilot clinical trial. Radiology 2004;231:208–214.[Abstract/Free Full Text]
21. Cupples TE. Impact of computer-aided detection (CAD) in a regional screening mammography program (abstr). Radiology 2001;221(P):520.
22. Gold MR, Siegel JE, Russell LB, et al. Cost-effectiveness in health and medicine. New York, NY: Oxford University Press, 1996.
23. Tengs TO. Cost-effectiveness versus cost-utility analysis of interventions for cancer: does adjusting for health-related quality of life really matter? Value Health 2004;7(1):70–78.[CrossRef][Medline]
24. Duffy SW, Tabár L, Chen HH, et al. The impact of organized mammography service screening on breast carcinoma mortality in seven Swedish counties: a collaborative evaluation. Cancer 2002;95:458–469.[CrossRef][Medline]
25. Tabár L, Vitak B, Chen HH, Yen MF, Duffy SW, Smith RA. Beyond randomized controlled trials: organized mammographic screening substantially reduces breast carcinoma mortality. Cancer 2001;91:1724–1731.[CrossRef][Medline]
26. National Cancer Institute. SEER cancer statistics review, 1975–2001. National Cancer Institute Web site. http://srab.cancer.gov/devcan/. Accessed August 15, 2004.

This article has been cited by other articles:

				D. Regge, C. Hassan, P. J. Pickhardt, A. Laghi, A. Zullo, D. H. Kim, F. Iafrate, and S. Morini Impact of Computer-aided Detection on the Cost-effectiveness of CT Colonography Radiology, February 1, 2009; 250(2): 488 - 497. [Abstract] [Full Text] [PDF]

				F. J. Gilbert, S. M. Astley, M. G.C. Gillan, O. F. Agbaje, M. G. Wallis, J. James, C. R.M. Boggis, S. W. Duffy, and the CADET II Group Single Reading with Computer-Aided Detection for Screening Mammography N. Engl. J. Med., October 16, 2008; 359(16): 1675 - 1684. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2392050670v1 239/3/710    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lindfors, K. K. Articles by Hurlock, G. S. Search for Related Content PubMed PubMed Citation Articles by Lindfors, K. K. Articles by Hurlock, G. S.