ARCHIVED - Organized Breast Cancer Screening Programs in Canada - Report on Program Performance in 2003 and 2004

2003 and 2004 Results

This report presents statistics for the 2003 and 2004 calendar years using data submitted up to August 2007. Further, the outcomes presented in this report are based on the 2007 report by the Evaluation Indicators Working Group (6). Some outcomes are based on a relatively small number of events but are included to accurately reflect the work of the Evaluation Indicators Working Group (for example: benign to malignant open surgical biopsy ratio). In these cases, sample size is presented. Data submission is staggered and may impact the completeness of cancer-related data for some programs. Unless otherwise noted, the summary statistics include data from all 10 provinces and apply to women aged 50 to 69. Importantly, the data from the Northwest Territories is only available for the 2004 reporting year and therefore tables requiring at least two years of data exclude this region.

Participation and Retention in Organized Screening Programs

Participation

Adequate participation in breast cancer screening is essential for reductions in mortality to occur in the target population. Based on extrapolation from randomized controlled trials, Canadian programs have established 70% as the target participation rate.

Participation rates include all 10 provinces and data from the Northwest Territories for screening exams in the year 2004 only. Overall, 1,345,382 Canadian women between 50 and 69, and 1,723,690 women of all ages, received a screening mammogram through a Canadian organized screening program in 2003 and 2004 (Table 2). Since the inception of the first Canadian organized screening program in British Columbia, over 6.5 million screening mammograms have been performed. Although these numbers appear high, the targeted program participation rate of 70% among women 50 to 69 years for biennial screening is far from being reached through organized programs. In 2003 and 2004, only 36.5% of the target population received a screening mammogram through an organized program. This represents a small improvement from 2001 and 2002 when only 33.9% of eligible women attended. The participation rate varies between programs from 10.4% in Alberta to 52.9% in New Brunswick (Figure 3).

Overall, many well-established programs have seen participation rates stabilize since 1997 and 1998; British Columbia (46-51%), Alberta (10-15%), and Saskatchewan (52-55%). While programs in Nova Scotia (27-41%) and Ontario (13-27%) have continued to see participation rates increase. Newer programs in Québec (12-48%), New Brunswick (36-53%), and Newfoundland and Labrador (18-26%) have also seen increases in participation rates.

Importantly, these rates do not include women who receive their breast cancer screening outside of an organized program. Results from population health surveys suggest that close to 62% of women between 50 and 69 years received a screening mammogram within the past two years (Figure 4). This figure is selfreported and may be slightly inflated as survey respondents tend to overestimate desirable behaviours, however, it is more closely aligned with the target of 70% set by the Evaluation Indicators Working Group Report.

Figure 3 – Participation in organized breast cancer screening programs, women aged 50-69, 2003 and 2004 screen years

Source: Statistics Canada data for 2003 and 2004 are used for denominator values.

Notes: Population estimates are averaged.
The national participation rate of 36.5% is indicated by the horizontal bar.

Retention

Optimal benefits from screening programs are achieved when regular participation in screening occurs. Two targets were set based on participation rates, sojourn time, screening interval studies and randomized controlled trials (7-9). The first, for women undergoing their initial screening mammogram, states that ≥75% of women should return within 30 months. The second states that ≥90% of women undergoing a subsequent screen should return within 30 months. Retention rate for women aged 50 to 69 excludes women who returned at age 70 years or older.

Overall, 74.3% of women aged 50 to 68 who received a screening mammogram between 2001 and 2002 were rescreened within 30 months during 2003 to 2004. Among women who received their first screening mammogram in the 2001 and 2002 calendar years, 64.9% returned for a subsequent mammogram within 30 months. Among women aged 50 to 68 who received a subsequent screening mammogram in the 2001 and 2002 calendar years, 76.8% returned for a subsequent mammogram within 30 months. (Table 6)

Figure 4 - Proportion of women aged 50-69 with a self-reported mammogram^{Footnote 2a} in the past two years by province, 2005 Canadian Community Health Survey

Source: Health Canada. 2005 Canadian Community Health Survey: share file.

Figure 5 – Cumulative probability of returning for a subsequent program screen by age group, 2000 and 2001 screen years

Note: Northwest Territories data are not included in this analysis.

In general, younger women (40 to 49 years) were more likely to return for subsequent screening within 30 months compared to older women (70+ years) regardless of whether it was an initial (63.9% and 53.5% respectively) or subsequent screen (86.7% and 68.4% respectively) (Table 7). Most women returned for subsequent screening between 21 and 27 months after their 2001 to 2002 screen, however, women between age 40 and 49 were more likely than older women to return between 12 and 15 months (Figure 5).

Results of Screening

The goal of organized screening programs is to identify disease in asymptomatic women and at the same time minimize the number of healthy women who receive abnormal screening results. Both the abnormal call rate and the positive predictive value offer insight into the process of accurately identifying asymptomatic women with breast cancer.

Figure 6 – Age distribution of program screens by province, 2003 and 2004 screen years

Abnormal Call Rates

The abnormal call rate refers to the percentage of all women screened who are referred for further testing because of abnormalities found during the screening mammogram and is one way to measure of the quality of a screening program. The Canadian target is <10% for women undergoing their first screen and <5% of women undergoing their subsequent timely screens.

Among women 50 to 69 years, the abnormal call rate for women receiving their first screening mammogram is 12.1% and for a subsequent screening mammogram is 6.5% (Table 6). These rates have been relatively stable since 2001 and 2002. Radiologist inexperience and/or low reading volumes can contribute to unnecessarily high abnormal call rates, as can delays in rescreening. For all age groups, the abnormal call rate rises after a screening interval of 30 months indicating the importance of regular screening intervals (Figure 7).

Positive Predictive Value

The positive predictive value (PPV) is determined by the proportion of women with an abnormal call who go on to be diagnosed with invasive or in situ cancer. A high PPV reflects the minimization of unnecessary follow-up procedures. The Canadian target is ≥5% for first screens and ≥6% for subsequent timely screens.

Among women aged 50 to 69 years, and based on detection by mammography alone, the PPV meets these targets (5.0% and 7.3% for initial and subsequent screening mammograms respectively) and has been relatively stable since 2001. It is worth noting that PPV is sensitive to the age distribution of the screened population, which is among the reasons why the Canadian targets are only intended for women age 50 to 69. The PPV improves dramatically with age for it is as low as 2.3% for women between 40 and 49 years undergoing their initial screening mammogram and as high as 13.8% in the 70+ age group (Table 6-8).

Figure 7 - Abnormal call rate^{Footnote 3a} by age group, 2003 and 2004 screen years

Notes: The median time for women to return for screening is as follows: Rescreen (>9 months, ≤18 months) by 12.5 months; Rescreen (>18 months, ≤30 months) by 24.4 months; Rescreen (>30 months) by 35.7 months.

Diagnostic Interventions

As suggested by the PPV, most women who receive abnormal screening results do not actually have breast cancer, however, additional assessment is required to determine the defi nitive diagnosis. The provision of timely, well coordinated, and minimized follow-up assessment has been shown to reduce fear and anxiety associated with abnormal results (10). Women who receive abnormal results require additional radiological or surgical assessment including diagnostic mammography, ultrasonography, core or open biopsy, and/or fine needle aspiration.

Analysis of diagnostic test type (Figure 8) includes all 10 provinces and data from the Northwest Territories for screening exams in the year 2004 only. In 2003 and 2004, approximately three quarters of women who received an abnormal screen were followed-up with additional breast imaging only. A further 13% received breast imaging combined with core biopsy or fine needle aspiration; an increase from 9.3% in 2001 and 2002 (Figure 8). Lastly, there was a shift from the use of the more invasive open biopsy to the less invasive core biopsy from 2001 and 2002 to 2003 and 2004. Core biopsy increased from 9.6% (9,187 women) to 12.3% (13,648 women) and open biopsy decreased from 7.2% (6,874 women) to 5.6% (6,188 women) (Table 4).

Diagnostic Interval

The diagnostic interval is the duration of time from the abnormal screen to its resolution. Excessively long diagnostic intervals can have negative psychological impact and potentially worsen prognosis (10). The Canadian target is ≥90% of abnormal screens will be resolved with 5 weeks if no tissue biopsy is required and ≤90% within 7 weeks if tissue biopsy is required.

Nationally, 74.3% of women not requiring a tissue biopsy received resolution within five weeks and 46.3% of women requiring tissue biopsy received resolution within seven weeks. The proportion of women who did not require tissue biopsy and received resolution within fi ve weeks has been showing gradual annual improvement, however, the proportion of women requiring tissue biopsy who received resolution within seven weeks has been relatively stable over time (Table 6-8).

Benign to Malignant Open Surgical Biopsy Ratio^{Footnote 3}

High quality pre-surgical assessment reduces the number of women requiring invasive follow-up who ultimately have a normal or benign result. The ratio of benign results compared to malignant results after an open surgical biopsy is one way to assess the quality of the pre-surgical assessment. This includes women who went directly to an open surgical biopsy as their primary assessment and those who underwent an inconclusive core biopsy prior to a definitive diagnosis by open surgical biopsy. The target for this indicator is ≤1:1; for every malignant cancer detected there should be one or fewer benign results.

In 2003 and 2004, the ratio was 1.8:1, meaning that close to two benign results for every malignant cancer case were detected by open surgical biopsy. The ratio among women undergoing their initial mammogram was 2.6:1 while women undergoing a subsequent mammogram had a ratio of 1.6:1 (Table 6). This result varies by the age of the women. Women 70 years and older undergoing a subsequent mammogram had the best ratio (0.9:1) while women between 40 and 49 years undergoing their first mammogram had the poorest ratio (5.6:1) (Table 7).

Figure 8 – Combinations of diagnostic procedures after an abnormal screen, women aged 50-69, 2003 and 2004 screen years

Northwest Territories data are for 2004 only.

The benign to malignant open surgical biopsy ratio reflects the experience of very few women because there has been a substantial shift to core biopsy, and away from open surgical biopsy, as a means to achieve definitive diagnosis. As a result, the indicator is nearing the end of its usefulness as the declining numbers of procedures result in ratios that are difficult to interpret. The number of women undergoing this procedure is included in Table 6 to illustrate this point.

Benign Open Surgical Biopsy Rate

The rate of open surgical biopsy can provide an indication of the quality of pre-surgical assessment, however, no target as yet has been set for this indicator.

In 2003 and 2004, the benign open surgical biopsy rate was 4.5 and 2.6 per 1,000 screens (initial and subsequent screens respectively). The biopsy rate is lower among older women (70+ years) undergoing their first screening mammogram compared to younger women, however, rates among women undergoing subsequent screening mammograms shows little variation by age group. Since 2000, the rate has decreased for both initial and subsequent screening mammograms suggesting a shift away from the use of open surgical biopsy (Table 6-8).

Benign to Malignant Core Biopsy Ratio

The ratio of benign to malignant core biopsies, can provide an indication of the quality of pre-surgical assessment, however, no target as yet has been set for this indicator.

In 2003 and 2004, the benign to malignant core biopsy ratio was 2.8:1 for initial screens and 1.5:1 for subsequent screens, and is lowest in older women (70+ years). For women undergoing their first screen the ratio has decreased to 2.8:1 from 3.3:1 since 2000, however, for women undergoing subsequent screens the value has been relatively stable between 1.4 and 1.6:1 (Table 6-8).

Benign Core Biopsy Rate

The rate of benign core biopsy can provide an indication of the quality of presurgical assessment, however, no target has been set for this indicator.

In 2003 and 2004, the benign core biopsy rate was 11.6 and 4.7 per 1,000 screens (initial and subsequent screens respectively). The biopsy rate is lowest among older women (70+ years) undergoing subsequent screens. Since 2000, the rate has increased for both initial and subsequent screening mammograms suggesting a shift toward the use of core biopsy (Table 6-8).

Cancer Detection

In total, 6,900 cancers (invasive, in situ and unclassified types combined) were detected among women aged 50 to 69 during 2003 and 2004 by organized screening programs (Table 6). Other breast cancers among Canadian women were detected by opportunistic screening (outside of an organized program) or when a woman became symptomatic of disease.

Among women ≥40 years, 79.7% (6,851 women) were diagnosed with invasive and 20.3% (1,747 women) with in situ cancers. This includes data from women diagnosed in the Northwest Territories during 2004. The proportion of cancers considered invasive increased with age; 70.8% of women aged 40 to 49 were diagnosed with invasive cancers compared to 85.3% of women 70 or more years (Table 5).

Invasive Cancer Detection Rate

The targets for invasive cancer detection rates established in Canada are >5 per 1,000 first screens and >3 per 1,000 subsequent timely screens.

In Canada, women undergoing their first screen had an invasive cancer detection rate of 4.7 cases per 1,000 screens. Women undergoing subsequent screens had an invasive cancer detection rate of 3.7 cases per 1,000 screens^{Footnote 4} (Table 6). As anticipated, the invasive cancer detection rate increased in older women and when subsequent screening was not timely (Figure 9).

In Situ Cancer Detection Rate

Ductal carcinoma in situ (DCIS) is a form of cancer detected through mammography screening, however, there is limited evidence supporting the transition of all forms of DCIS to invasive cancer. Because of this, no target has been set for in situ cancer detection rates in Canada. Despite this, it is important to monitor rates of detection until appropriate targets can be set. In Canada, women undergoing their first screen had a DCIS detection rate of 1.3 cases per 1,000 screens. Women undergoing subsequent screens had a DCIS detection rate of 1.0 case per 1,000 screens^{Footnote 4} (Table 6).

Invasive Tumour Size and Node Negative Rate

Cancer detected at earlier stages has more treatment options, less recurrence, and improved survival. Research in Canada has shown that 97.9% of women with stage I breast cancer survive at least five years while only 27.9% of women diagnosed in stage IV survive for five years (11). Early stage cancer has smaller tumours and no lymph node involvement. Based on size of tumour, the Canadian target is for greater than 25% of invasive tumours to be ≤10mm and greater than 50% of invasive tumours to be ≤15mm. The second target is for >70% of women with invasive cancer to have no lymph node involvement.

Analysis of cancer stage (Table 5) includes all 10 provinces and data from 2004 screening exams in the Northwest Territories. Among women aged ≥40 years, diagnosed with breast cancer in 2003 and 2004, 46.8% of tumours were classified as stage I and fewer than three percent were classified as stage III/IV (Table 5). The proportion of women with DCIS (stage 0) detected decreased with age while the proportion with stage I invasive cancer increased with age. Stage II - IV invasive cancer stayed relatively stable across age groups.

Accordingly, the proportion of invasive tumours less than 10 mm was 36.4% and almost three quarters of women had negative lymph nodes at diagnosis (Table 6). A larger proportion of older women had smaller tumours at diagnosis and negative lymph nodes at diagnosis compared to younger women (Table 7).

Figure 9 – Cancer detection (Invasive and In situ) rate per 1,000 screens by age group, 2003 and 2004 screen years

Notes: The shaded area indicates the rate of invasive cancers detected, while the non-shaded area indicates the rate of DCIS cancers detected.
Northwest Territories data are not included in this analysis.

The median time for women to return for screening is as follows: Rescreen (>9 months, ≤18 months) by 12.5 months; Rescreen (>18 months, ≤30 months) by 24.4 months; Rescreen (>30 months) by 35.7 months.

Post-screen Cancers

Post-screen cancers are those cancers that develop after a normal screening mammogram but before the next screen and serve as an indicator of the sensitivity of the screening program. Post-screen cancers can include cancers that occur after the recommended 12 or 24 months in women who do not return for their regular annual or biennial screen respectively (non-compliant cancers) or women who become symptomatic before their next regular screen (interval cancers). These cancers do not include cases referred for diagnostic follow-up with a benign result. Post-screen cancers were calculated based on all women screened from 2000 to 2001 who developed an interval cancer during 2000 to 2003. Non-compliant cancers were not included in this calculation. In order to ensure consistency between provinces this report also considers interval cancers to include those detected by a screening mammogram that have taken longer than six months to diagnosis. The target is for less than 6 women per 10,000 person years be diagnosed with a post screen cancer within 12 months of screening and less than 12 women per 10,000 person years within 24 months.

Nationally, the post-screen cancer detection rate was 5.4 per 10,000 person years within 12 months and 7.9 per 10,000 person years within 24 months (Table 6).