A meta-analysis of mammography screening promotion

The purpose of this study was to identify factors that influence the effectiveness of interventions in increasing women's use of mammography screening programs. To this end, we conducted a systematic literature review of studies published between 1966 and 1997. In this review, we recorded data about the year and country in which studies were completed, the study design, the methods for measuring screening rates, various sample characteristics, the nature of the intervention, and the resulting screening rates. The PRECEDE model was used as a framework to make distinctions between the various interventions. To synthesize evidence about the baseline screening rates and the effect of interventions on the incidence of mammography screening, we fit random-effects logistic regression models. These models revealed that more recent studies (those conducted from 1990 to 1996) were associated with higher screening rates (odds ratio [OR], 2.1; 95% confidence interval [CI], 1.2-3.9). Conversely, those designed to target older women (minimum age, 50-65 years) and those set in clinics exhibited smaller screening rates (OR, 0.6, 95% CI, 0.3-1.0, and OR, 0.5; 95% CI, 0.3-0.8, respectively). The meta-analyses also suggested methodologic issues that must be considered before the relative strength of various interventions can be assessed rigorously.

Recommendations for a mammography quality assurance program

In 1989 the ACPSEM published a position paper entitled "A Quality Assurance Programme for Mass Screening in Mammography". This paper described test parameters and performance specifications for the equipment related aspects of a mammography quality assurance program. Advice on test equipment selection was also provided. In the intervening period of time there have been considerable advances in mammography technology creating a need to review a number of the paper's recommendations. There have also been considerable developments in the mammography quality assurance (QA) field, not the least of which includes the American College of Radiology Mammography Accreditation Program (ACR-MAP) and the similarly structured Royal Australian and New Zealand College of Radiologists' Mammography Accreditation Program (RANZCR-MAP). In light of these developments it was decided by the Radiology Interest Group to review the ACPSEM position on those aspects of mammography QA that fall within the medical physicist's area of expertise. This document represents the outcome of those deliberations.

An analysis of perceptual errors in reading mammograms using quasi-local spatial frequency spectra

In this pilot study the authors examined areas on a mammogram that attracted the visual attention of experienced mammographers and mammography fellows, as well as areas that were reported to contain a malignant lesion, and, based on their spatial frequency spectrum, they characterized these areas by the type of decision outcome that they yielded: true-positives (TP), false-positives (FP), true-negatives (TN), and false-negatives (FN). Five 2-view (craniocaudal and medial-lateral oblique) mammogram cases were examined by 8 experienced observers, and the eye position of the observers was tracked. The observers were asked to report the location and nature of any malignant lesions present in the case. The authors analyzed each area in which either the observer made a decision or in which the observer had prolonged (>1,000 ms) visual dwell using wavelet packets, and characterized these areas in terms of the energy contents of each spatial frequency band. It was shown that each decision outcome is characterized by a specific profile in the spatial frequency domain, and that these profiles are significantly different from one another. As a consequence of these differences, the profiles can be used to determine which type of decision a given observer will make when examining the area. Computer-assisted perception correctly predicted up to 64% of the TPs made by the observers, 77% of the FPs, and 70% of the TNs.

Factors affecting phantom scores at annual mammography facility inspections by the U.S. Food and Drug Administration

RATIONALE AND OBJECTIVES

The authors performed this study to evaluate the factors affecting phantom image score at the annual inspection of mammography facilities.

MATERIALS AND METHODS

In 1997, three U.S. Food and Drug Administration (FDA)-trained inspectors performed inspections of all mammography facilities in North Carolina. All federal and state inspection data were collected and evaluated by using linear regression analysis. Factors affecting the American College of Radiology phantom scores were assessed.

RESULTS

Phantom score was affected by inspector identity, view box luminance, and optical density. All of these factors had a statistically significant effect on mass score (P < .05). Inspector identity yielded a statistically significant effect on speck group score, fibril score, and total score. Luminance yielded a statistically significant effect on both speck group score and total score.

CONCLUSION

Phantom scoring should be automated to allow for more consistent interobserver scoring. In addition, radiology facilities can improve the likelihood of receiving a passing phantom score by reducing the ambient light and increasing the view box luminance in the location where the images are evaluated and the phantom is scored routinely. Radiologists should also consider increasing phantom and clinical image optical density to allow for improved phantom testing outcomes.

[Economic approach to breast cancer screening in Belgium]

In Belgium, the breast cancer screening by mammography involves a great number of women. It is largely non organized and regarded to the small number of data, it is impossible to determine if this screening is utile and if it has a good "cost-benefit" ratio. This work studies the economical aspects of breast cancer screening in the belgian healthcare system. From information found in other countries, we build four models corresponding to an organized and spontaneous screening. We studied the total processes from screening to diagnosis, including the quality assurance and the evaluation of effectiveness in the organized models. We then applied the reimbursements of the belgian health insurance in the models and compared the costs. It appears that a screening for breast cancer must be organized to give a best "cost-effectiveness" ratio. Pilot projects should be the best way to study the best organization modalities in Belgium.

Improving the time to diagnosis after an abnormal screening mammogram

INTRODUCTION

Five community-specific interventions to reduce the time to diagnosis after an abnormal breast screen have been evaluated.

METHODS

Subjects with abnormal screening mammograms in 1998 were assessed through five community pilot projects (N = 1137) and a control random sample assessed elsewhere in BC (N = 1053). The number, types, dates and physician costs of breast-related interventions after an abnormal screen were compared between pilots and control.

RESULTS

The median time to diagnosis for women without a biopsy was reduced from 23 days to 7 days (p = 0.001) in the pilot with facilitated referral to diagnosis. The median time to diagnosis for women with a biopsy was reduced from 57 days to 22-43 days in the pilots. Median physician costs per subject were lower (p = 0.02) in pilots that more frequently used core biopsy to obtain a diagnosis.

CONCLUSIONS

Process changes can improve the time to diagnosis after an abnormal breast screen, with similar or lower physician costs per subject. Facilitating the referral process had the greatest impact.

Accuracy of mammography and echography versus clinical palpation in the assessment of response to primary chemotherapy in breast cancer patients with operable disease

The response to primary chemotherapy is an important prognostic factor in patients with non metastatic breast cancer. In this study we compared the assessment of response performed by clinical palpation to that performed by echography and mammography in 141 out of 157 consecutive breast cancer patients (T2-4, N0-1, M0) submitted to primary chemotherapy. A low relationship was recorded between tumor size assessed clinically and that evaluated by either mammography: Spearman R = 0.38 or echography: R = 0.24, while a greater correlation was found between the tumor dimension obtained by the two imaging techniques (R = 0.62). According to the WHO criteria, the grade of response of breast cancer to primary chemotherapy, showed by mammography and echography, was less marked than the grade of response seen at clinical examination. Residual tumor size assessed clinically depicted a stronger correlation with pathological findings (R = 0.68) than the residual disease assessed by echography (R = 0.29) and mammography (R = 0.33). Post-chemotherapy histology evaluation revealed pathological complete response in three cases (2.1%). Two of these cases were judged as complete responders by clinical palpation but only one was recognized by mammography, and none by echography. Clinical response, but not the response obtained by the two imaging techniques, was a significant predictor for longer disease free survival (p = 0.04). To conclude, physical examination measurements remain the method of choice in evaluating preoperatively the disease response in trials of primary chemotherapy. Prediction of pathological outcome is not improved by echography and mammography.

Updated audit database for breast imaging/histopathology correlation

AIM

To develop, test and validate an audit database, evolved from a prior, peer-reviewed, audit tool for symptomatic multi-modality breast imaging services.

MATERIALS AND METHODS

The database is to cover all aspects of non-invasive breast imaging and recognize subspeciality inputs. When more than one imaging investigation is used for diagnosis, an overall breast imaging audit grade is to be introduced. Data derived from clinical studies has been input into the new database.

RESULTS

Results for mammography alone are virtually identical to those of the previous program. A slight increase in accuracy is suggested by using more investigations if the first investigation is not conclusive. More comprehensive reports can be issued.

CONCLUSION

The audit program can be used in the same context as the old audit. If mammography is the sole investigation, there is no change from the previous standards of a minimum mammography (ultrasound)/histopathology agreement of 70%, an equivocal rate of less than 15%, a false-positive rate of less than 7.5% and a false-negative rate of less than 6.5%. Although there is no statistical difference when more than one imaging investigation is used, there is some marginal improvement. It is suggested that initial audit standards for multi-imaging should be 75% for minimum agreement, a 10% maximum for an equivocal rate, a 5% maximum for false negative and an unchanged false positive rate of 7.5% maximum. These standards will be refined with experience. Due to the nature of the database, complex queries can be made including those about histopathological data. If widely used, the database will be a useful tool to audit the accuracy of symptomatic breast imaging services and types and frequencies of symptomatic breast disease, as seen in routine settings.

The accuracy and effectiveness of routine population screening with mammography, prostate-specific antigen, and prenatal ultrasound: a review of published scientific evidence

OBJECTIVE

To review published data regarding the accuracy and effectiveness of three screening tests: mammography, prostate-specific antigen (PSA), and prenatal ultrasound.

METHODS

Published evidence regarding the accuracy and effectiveness of the three tests was collected by computerized literature search and supplemented by manual review of relevant bibliographies.

RESULTS

Screening mammograms lower breast cancer mortality by about 20%. Most data come from women aged 50-64 years; women aged 40-49 years may also benefit, but the absolute risk reduction is lower. Up to 1,500 to 2,500 women must undergo screening to prevent one death from breast cancer. Mammograms miss approximately 12% to 37% of cancers, generate false-positive results, and cause anxiety while abnormal results are evaluated. PSA screening can detect 80% to 85% of prostate cancers but has a high false-positive rate. There is little direct evidence that early detection reduces morbidity or mortality. Indirect evidence includes a trend toward earlier stage tumors and steadily declining mortality rates in geographic areas where PSA screening has become common. Potential harms include the morbidity associated with evaluating abnormal results, and complications from treatment (e.g., impotence, incontinence). The overall balance of benefits and harms remains uncertain in the absence of better evidence. Prenatal ultrasound may reduce perinatal mortality, primarily through elective abortions for congenital anomalies, but does not appear to lower live birth rates. Although ultrasound has no proven effect on neonatal morbidity, it provides more accurate estimates of gestational age that prevent unnecessary inductions for post-term pregnancy. Screening detects multiple gestations, congenital anomalies, and intrauterine growth retardation, but direct health benefits from having this knowledge are unproved. Ultrasound has both positive and negative psychological effects on parents. The scans do not appear to harm childhood development.

CONCLUSIONS

Even for the most established screening tests, the appropriateness of routine testing depends on subjective value judgments about the quality of supporting evidence and about the trade-offs between benefits and harms. Individuals, clinicians, policy makers, and governments must weigh the evidence in light of these values and the constraints imposed by available resources.

The impact of alternative practices on the cost and quality of mammographic screening in the United States

The decentralized structure of health care in the Unites States hinders population-based analysis of breast cancer screening. Our objectives are to model mammography in the United States as a whole, to identify the variables that most profoundly affect cost and efficacy, and to develop a strategy to improve mammography screening from a population perspective. A spreadsheet model was used to represent the variables of mammography screening in the United States. The population-based national screening program in Sweden provides a framework for comparison. The outcome measures are the aggregate cost and the number of cancers detected by mammography. We used deterministic sensitivity analysis to calculate the impact of variation in practice. Aggregate costs of screening in the United States are in the range of $3-$5 billion dollars. The percentage of women screened, cost per mammogram, cancer to biopsy ratio, recall rate, and cost of recall have the most profound effect on the quality and cost of a national screening program. Variance of these high-impact variables, based on the U.S. population, modifies the aggregate cost of screening by over $2 billion. As mammography screening in the United States increases to include all women over age 40, high-impact variables should be optimized to decrease costs and improve breast cancer detection. Our model establishes which parameters are most important.

Effect of training with the american college of radiology breast imaging reporting and data system lexicon on mammographic interpretation skills in developing countries

RATIONALE AND OBJECTIVES

The authors evaluated the effect of training in the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS) lexicon on the interpretive skills of radiologists evaluating screening mammograms in Ukraine.

MATERIALS AND METHODS

As part of a program to improve breast cancer detection and treatment in Ukraine, a series of training sessions was given to a group of radiologists across Ukraine to improve their interpretive skills in screening mammography. The training sessions focused on the use of the lexicon and assessment categories developed by the ACR BI-RADS committee. Participants (n = 14) evaluated 30 test screening mammograms before and after the training sessions. The test sets were randomly selected from a larger collection of training sets containing normal, benign, and abnormal mammograms. False-positive, false-negative, true-positive, and true-negative evaluations were determined, and sensitivity, specificity, and positive predictive values were calculated for each participant before and after training.

RESULTS

The mean baseline sensitivity, specificity, and positive predictive values were 50%, 77%, and 43%, respectively. Each of these measures of interpretive skills improved significantly after training in the use of the lexicon, to 87%, 89%, and 78% (P < .0001, P < .01, and P < .0001, respectively).

CONCLUSION

As the use of mammography spreads throughout developing countries, it is essential to address training and educational needs, as well as equipment needs. The ACR BI-RADS lexicon provides a systematic and efficient method for training radiologists to interpret screening mammograms. Educating radiologists on the use of this lexicon proved an effective way to improve their interpretive skills in screening mammography.

Procedure for quantitatively assessing automatic exposure control in mammography: a study of the GE Senographe 600 TS

The correct operation of a mammography system's automatic exposure control system (AEC) is essential if mammograms are to be produced with a suitable film exposure. A methodology has been developed that allows the performance of the AEC to be assessed quantitatively using clinical films. By digitizing mammograms, the mean optical density (OD) in the main breast region and in a region of interest corresponding to the position of the AEC detector are evaluated for each film, together with the area of the main breast. Using these data it is possible to determine the relationships between the mean OD, breast size and AEC detector position. The performance of the AEC on a GE Senographe 600 TS system was investigated. The study found that there is a tendency to underexpose smaller breasts, i.e. with an area less than approximately 4000 mm(2). This is equivalent to a compressed tissue width of approximately 60---80 mm. The difference in mean OD between the mammograms of small and large breasts was up to 0.7 OD. Provided the sensitive area of the AEC detector is known, this method of assessing AEC performance can be used with any mammography system.

ROC curve analysis of lesion detectability on phantoms: comparison of digital spot mammography with conventional spot mammography

Although conventional screen--film mammography has excellent spatial resolution and is commonly used as a screening tool, certain inherent limitations prevent its further improvement. New digital mammography techniques, despite lower spatial resolution than screen--film mammography, may overcome these limitations. This study compared lesion detectability between charge coupled device-based digital spot mammography and conventional spot mammography. A total of 100 sets of images of specially designed breast phantoms was acquired, with variable background achieved by overlapping several layers of grapefruit fibre on a 4 cm thick lucite slab, using both modalities. 75 sets were "normal" images and 25 sets were images with simulated lesions. Four radiologists assessed the images according to a five-point confidence scale. The results were used to construct receiver operating characteristic curves. No statistical difference was observed between the two sets of curves for individual radiologists as well as pooled data. The lower spatial resolution of digital mammography was compensated for by its higher contrast sensitivity relative to conventional spot mammography.

The consequence of undertreating breast cancer in the elderly

BACKGROUND

Recent studies have noted that a large fraction of elderly patients do not receive conventional treatment for breast cancer. The consequences of undertreatment of the elderly have not been adequately assessed.

STUDY DESIGN

The senior author's database (PIT) was used to identify women undergoing potentially curative operations for breast cancer between 1978 and 1998. Risk factors, presentation, pathologic findings, treatment, and outcomes of 206 women aged over 70 years were compared with those of 920 younger patients. In addition, conventionally treated and "undertreated" elderly patients were identified, and their characteristics and outcomes were compared.

RESULTS

Older patients' cancers were more often visible on mammography, usually as a mass; younger patients' mammograms were less frequently positive, presenting more often with calcifications (p = 0.002). Cancers of the elderly were better differentiated (p < 0.001) and more likely to be estrogen- and progesterone-receptor positive (p < 0.001; p = 0.007). Patients over 70 had fewer mastectomies (19% versus 33%; p < 0.001) and were also less likely to undergo axillary node dissection (71% versus 81%, p = 0.006), postoperative radiation (69% versus 92%, p < 0.001), and chemotherapy (18% versus 48%, p < 0.001). Fifty-seven percent of older patients were treated with tamoxifen compared with 36% of younger patients (p < 0.001). Elderly patients' rates of local and distant recurrence were comparable to those of younger patients after both mastectomy and breast conservation. Ninety-eight patients (54%) over 70 were undertreated by conventional criteria. Undertreated elderly patients were significantly older (78 versus 76 years, p = 0.003), were diagnosed with excisional biopsy more often (69% versus 57%, p = 0.069) and with fine-needle aspiration less frequently (22% versus 38%, p = 0.069), and were more likely to have breast conservation (90% versus 73%, p = 0.004). Local and distant disease-free survival rates of both groups were comparable. Tamoxifen treatment significantly reduced the chance of developing distant metastasis in node-negative elderly patients with invasive tumors (p = 0.028). Omission of chemotherapy had no impact on disease control in the elderly. Axillary node status and estrogen-receptor status were significantly related to local disease-free survival, and axillary node status was significantly related to distant disease-free survival in multivariate analysis in the elderly.

CONCLUSIONS

Elderly breast cancer patients are frequently treated with breast conservation, omitting axillary dissection, radiation therapy, and chemotherapy. Despite undertreatment by conventional criteria, the rates of local recurrence and distant metastasis are not increased in comparison with conventionally treated elderly patients. Tamoxifen should be administered to elderly breast cancer patients with invasive tumors because it significantly improves distant control.

[Impact of menopause hormone replacement therapy on screening mammography reading]

PURPOSE

To assess difficulties in reading screening mammographies related to hormone replacement therapy (HRT) in the setting of a breast cancer screening program. Disagreement between two readers is used as the assessment criterion.

POPULATION AND METHODS

The study population consisted of all women participating in the ADEMAS breast cancer screening program in Bas-Rhin (France). Agreement between the two radiologists was considered when the same global conclusion (call-back or not of the women for further examination) or the same conclusion for each of the nine possible lesions was observed.

RESULTS

HRT is associated with a higher risk of disagreement between the two radiologists on global conclusion and on opacities. Conversely, disagreement risk is lowered for microcalcifications (whether benign or malignant) for HRT users.

CONCLUSION

HRT is associated with a higher risk of disagreement on global conclusion. This has an impact on whether the women should be called-back or not according to reading of breast cancer screening program mammographies.

[3D reconstruction of microcalcification clusters]

PURPOSE

A characterization of the geometrical or morphological aspect of microcalcifications is not sufficient to confirm a diagnosis of cancer. On the other hand, it seems that the shape of the cluster is a pertinent and helpful criterion for diagnosis. The aim of our study is to work out a method for 3D reconstructions of clusters of microcalcifications.

MATERIALS AND METHODS

After having tested different algorithms for extracting microcalcifications, we have developed a method of 3D reconstructions of the shape of the cluster from three radiographs. The technique, available using a standard mammography unit, consists of three principal steps

RESULTS

The technique has been tested with two simple tridimensional phantoms and has given results with sufficient precision for possible clinical use.

CONCLUSION

We plan now to apply the method in vivo with adaptation on a dedicated mammographic stereotactic unit with digital acquisition.

Audit of a one-stop breast clinic--revisited

BACKGROUND

A one-stop diagnostic service has been available for women with symptomatic breast disease at St Bartholomew's Hospital for 5 years and was originally audited in May 1993. In re-auditing the one-stop service our aim was to see if our practice had improved following the original audit and to look at the impact which these changes in practice had made to the service offered to the patient.

METHODS

A prospective audit of 4 consecutive clinics was undertaken in November 1997. A total of 300 patients (59 new and 241 follow up) were seen in clinic in this time. The primary outcome measure concerned the completeness of triple assessment in the 40 patients who required one stop investigations, including mammography, ultrasonography and fine-needle aspiration cytology. In addition, mean time to initial consultation and level of awareness of the one-stop facility and its attendant time delay were measured.

RESULTS

Of the 300 clinic attendees 40 (38 new, 2 follow-up) had one-stop investigations. As a result of the one-stop service being in practice, 36 patients (90%) had a definitive management decision made at their first outpatient visit. Of these 2 were symptomatic cancers, forming 5% of the workload. A total of 86% of the workload was benign. Four patients (10%) had equivocal results. The mean waiting time from designated appointment until surgical consultation was 36.7 minutes and was disappointingly unchanged from that of the previous audit. However this does not take into account the significant reduction in staffing levels which has occurred between the two periods of assessment.

CONCLUSIONS

The initial audit identified a significant problem with time constraints, necessitating that a large number of patients with carcinomas return at a later date for further investigations. Booking only new patients at the beginning of clinic has provided a solution. Disappointingly, our figures do not show a significant improvement in mean waiting time compared with the previous audit, despite allowing GPs greater access of referral. Encouragingly, we have been able to maintain a similar standard of provision of care despite lower staffing levels and to implement the changes suggested by the original audit (thereby closing the audit loop).

Baseline staging tests in primary breast cancer: a practice guideline

BACKGROUND

Breast cancer develops in over 7000 women each year in Ontario. These patients will all undergo some staging work-up at diagnosis. The Breast Cancer Disease Site Group of the Cancer Care Ontario Practice Guidelines Initiative reviewed the evidence and indications for routine bone scanning, liver ultrasonography and chest radiography in asymptomatic women who have undergone surgery for breast cancer.

METHODS

A systematic review of the published literature was combined with a consensus interpretation of the evidence in the context of conventional practice.

RESULTS

There were 11 studies of bone scanning reported between 1972 and 1980, involving a total of 1307 women; bone scans detected skeletal metastases in 6.8% of those with stage I breast cancer, 8.8% with stage II and 24.5% with stage III. A total of 5407 women participated in 9 studies of bone scanning reported between 1985 and 1995; in these studies, bone scans detected skeletal metastases in only 0.5% of women with stage I disease, 2.4% with stage II and 8.3% with stage III. Among 1625 women in 4 studies of liver ultrasonography reported between 1988 and 1993, hepatic metastases were detected in 0% of patients with stage I disease, 0.4% with stage II and 2.0% with stage III. Among 3884 patients in 2 studies of chest radiography published in 1988 and 1991, lung metastases were detected in 0.1% of those with stage I, 0.2% with stage II and 1.7% with stage III. False-positive rates ranged from 10% to 22% for bone scanning, 33% to 66% for liver ultrasonography and 0% to 23% for chest radiography. The false-negative rate for bone scanning was about 10%.

RECOMMENDATIONS

The following recommendations apply to women with newly diagnosed breast cancer who have undergone surgical resection and who have no symptoms, physical signs or biochemical evidence of metastases. Routine bone scanning, liver ultrasonography and chest radiography are not indicated before surgery. In women with intraductal and pathological stage I tumours, routine bone scanning, liver ultrasonography and chest radiography are not indicated as part of baseline staging. In women who have pathological stage II tumours, a postoperative bone scan is recommended as part of baseline staging. Routine liver ultrasonography and chest radiography are not indicated in this group but could be considered for patients with 4 or more positive lymph nodes. In women with pathological stage III tumours, bone scanning, liver ultrasonography and chest radiography are recommended postoperatively as part of baseline staging. In women for whom treatment options are restricted to tamoxifen or hormone therapy, or for whom no further treatment is indicated because of age or other factors, routine bone scanning, liver ultrasonography and chest radiography are not indicated as part of baseline staging.

Radiological review of interval cancers in an Australian mammographic screening programme. Radiology Quality Assurance Group of BreastScreen Victoria

OBJECTIVES

To determine the proportion and features of invasive interval cancers that could be considered detectable at the time of the previous screen and the proportion of cases that could be classified as true intervals, false negatives, minimal signs, or radiographically occult lesions.

SETTING

BreastScreen Victoria, the Victorian component of the BreastScreen Australia mammography screening programme.

METHODS

Two separate review methodologies were adopted. Firstly a blinded review of interval, screen detected, and normal cases was undertaken, followed by a confirmation exercise to determine the proportion of invasive interval cancers that could be considered detectable at the time of the previous screen. Secondly, an unblinded review was performed to classify interval cases as true interval, false negative, minimal signs, or radiographically occult.

RESULTS

From the blinded review, it was estimated that 38% of interval cases may be considered "potentially detectable" at the time of screening. Comparison of the characteristics of interval and screen detected cases indicates that interval cases are more likely to be smaller, equivocal, ill defined masses. In the unblinded exercise, 41% of interval cases were classified as false negatives and a further 16% as minimal signs, 33% true intervals, and 10% radiographically occult. Of the interval cancers considered potentially detectable at screening, 97% were classified as false negatives in the unblinded review.

CONCLUSIONS

This study highlights the importance of adopting staged review methods with both blinded and unblinded components. The blinded review and confirmation exercise allows the determination of the proportion of interval cases that could be considered potentially detectable at screening. The unblinded review provides an active important opportunity for professional development and review and a mechanism to link into the blinded review through further classification of interval cases.

Radiological review of incidence breast cancers

OBJECTIVES

To examine the rate of incidence cancers detectable on review of previous screening mammograms using two reviewing methods. To compare the results with a previous study of interval cancers using the same reviewing methods.

SETTING

Almost 50000 women are regularly invited for service screening at Stockholm Söder Hospital. From 1989 to 1993, 119 women were identified with breast cancer detected at screening and the previous round attendance (incidence cancer).

METHODS

Screening mammograms, obtained before detection of the incidence cancers, were reviewed first mixed with other screening images (ratio 1:8) and then non-mixed. Reviewers from the screening unit responsible for the mammograms as well as reviewers from other units interpreted all images by both single and double reading.

RESULTS

The proportion detected on retrospective review varied between 5% and 50% depending on the review method used and the number of reviewers included to classify a case as truly identified. Generally more cancers were detected when non-mixed samples of mammograms were reviewed than when mixed samples were reviewed (mean increase 23%) and when interpreted by double reading compared with single reading (mean increase 14%).

CONCLUSIONS

In an experimental retrospective set up, fewer incidence cancers were identified in mixed than in non-mixed review. Generally more incidence cancers were identified on review (22%) than previously reported for interval cancers (14%), probably reflecting differences in tumour biology and growth. How many women with potentially visible incidence cancers would have benefited from earlier tumour detection still needs to be evaluated.