Content uploaded by Ye Shen
Author content
All content in this area was uploaded by Ye Shen on May 14, 2018
Content may be subject to copyright.
The impact of false positive breast cancer screening mammograms
on screening retention: A retrospective population cohort study in
Alberta, Canada
Ye Shen, MPH,
1
Marcy Winget, PhD,
2
Yan Yuan, PhD
1
ABSTRACT
OBJECTIVES: The impact of false positives on breast cancer screening retention is inconsistent across international studies. We investigate factors associated
with screening retention, including false positive screening results, invasiveness of diagnostic procedures, and geographic variation in Alberta, Canada.
METHODS: A total of 213 867 women aged 50–67 years who had an index screen mammogram between July 2006 and June 2008 were evaluated at
30 months post index screen to determine the screening retention rate. The association of screening retention with invasiveness of the diagnostic procedure,
time to diagnostic resolution, and region of residence were investigated using multivariable log binomial regression, adjusting for women’s age.
RESULTS: Women with false positive screening results were less likely to return for their next recommended screening than those with a true negative result
(62.0% vs. 68.7%). Compared to women with normal screening results, the adjusted risk ratios of fail-to-rescreen for women with imaging-only follow-up,
needle sampling, and open biopsy were 1.08 (95% CI: 1.05–1.12), 1.72 (95% CI: 1.44–2.07) and 2.29 (95% CI: 2.09–2.50) respectively. Screening
retention rates were slightly higher for rural residents than urban residents. Time to diagnostic resolution was not associated with screening retention.
Screening retention peaked at one year from the index date of the previous screening.
CONCLUSION: Higher awareness of the strong negative impact that biopsies in the case of a false positive screening have on screening retention is needed.
Such awareness can inform intervention strategies to mitigate the impact and improve screening retention rate.
KEY WORDS: Breast cancer; screening; retention; false positive
La traduction du résumé se trouve à la fin de l’article. Can J Public Health 2017;108(5-6):e539–e545
doi: 10.17269/CJPH.108.6154
Biennial or triennial breast cancer screening mammography,
as an effective public health strategy to reduce breast
cancer mortality, has been recommended to women aged
50–74 in the United States, Canada and most European
countries.
1–3
The survival benefits of breast cancer screening are
deemed to outweigh the harms from over-diagnosis, over-
treatment, false positive screening results, and benign biopsies.
4,5
The latter two have been shown to be associated with depression
and long-term anxiety in women and possibly reduce the
likelihood of future screening.
6,7
False positives account for about
9% of all screening mammograms and make up approximately
93% of abnormal calls in Canada and the US.
8,9
It is therefore
important to understand the magnitude of the impact of false
positives on screening retention in order to mitigate it.
The impact of false positives on screening retention rates,
however, is conflicting across studies. In a systematic review,
false positive screening mammograms were not associated with
retention rate in European countries, but were associated with an
increased retention rate in the US.
10
A recent study in the United
Kingdom showed that while the retention rate was not affected by
false positives, it was reduced in women who underwent biopsies.
6
Two studies in Canada, both published over a decade ago, found
that false positive screening results reduced the likelihood of
screening retention.
11,12
There are important differences in the
organization and delivery of screening programs and in the
characteristics of the populations screened across different
countries. Furthermore, practice has changed with respect to
follow-up procedures in the past decade; for instance, core biopsy
is now used broadly.
13
Efforts have also been made to improve
organized screening performance, which could affect retention
rates as well as the impact of false positives on them. It is therefore
unknown whether and which of the findings from previous studies
are applicable today.
The primary objective of this study was to investigate the impact
of false positives on breast cancer screening retention and to
determine whether invasiveness of the follow-up procedure and
time to diagnostic resolution have independent effects on
screening retention. A secondary objective was to investigate the
extent of geographical variation in retention rates. Implications of
these associations are discussed and recommendations are made to
improve screening programs and to benefit screen-eligible women.
Author Affiliations
1. School of Public Health, University of Alberta, Edmonton, AB
2. School of Medicine, Stanford University, Stanford, CA, USA
Correspondence: Yan Yuan, PhD, School of Public Health, University of Alberta,
3-299 Edmonton Clinic Health Academy, 11405 –87 Avenue, Edmonton, AB T6G 1C9,
Tel: 780-248-5853, E-mail: yan.yuan@ualberta.ca
Conflict of Interest: None to declare.
QUANTITATIVE RESEARCH
© 2017 Canadian Public Health Association or its licensor. CANADIAN JOURNAL OF PUBLIC HEALTH •VOL. 108, NO. 5-6, 2017 e539
METHODS
Overview of breast cancer screening program in Alberta
The population-based Alberta Breast Cancer Screening Program
(ABCSP) was established in 2004 through the collaboration of two
organizations: Screen Test (ST), which utilizes radiologists
employed by Alberta Health Services, and the Alberta Society
of Radiologists (ASR), a non-profit professional organization
representing 92% of the fee-for-service radiologists and radiology
residents in Alberta.
14
The ABCSP ensures an organized approach
for screen-eligible women to access screening mammography. Prior
to the launch of the provincial-wide ABCSP, ST represented the
much smaller organized “screening program” available in the
province, while screening by ASR radiologists represented
“opportunistic screening”. Breast cancer screening and diagnostic
procedures performed in Alberta, including procedure type, date,
results and follow-up recommendations, were captured by
complementary ASR and ST databases. Breast cancer screening
and diagnostic procedures include imaging (screening and
diagnostic mammography, ultrasound, MRI) and biopsies
(aspiration, stereotactic core, closed, surgical/open). Breast
Imaging Reporting and Data System (BI-RADS) scores
15
are
captured for imaging procedures. BI-RADS classifies lesions into
seven categories: 0 for incomplete and further imaging is required,
1 for negative findings, 2 for benign findings, 3 for probably
benign, 4 for suspicious abnormality, 5 for a mammographic
appearance, and 6 for known malignancy.
ST provides mammography services in clinics in two
metropolitan cities (Edmonton and Calgary); mobile units
visit rural and remote communities throughout the province
once a year.
16
Additionally, ASR-member radiologists provide
mammography services in community radiology clinics
throughout the province.
Study design and data linkage
The Canadian province of Alberta has a single-payer publicly
funded health care system under which standard medical care,
including breast cancer screening services, are free. The Alberta
clinical guideline recommended breast cancer screening at least
every two years for women between 50 and 69 years of age during
the study period.
17
In order to satisfy the age eligibility at screening
retention, 67 years of age was chosen as the upper limit for
inclusion in this study.
All women aged 50–67 years who had at least one screening
mammogram between July 1, 2006 and June 30, 2008 were
identified from the combined ASR and ST database. Women with
screen-detected breast cancer or who developed breast cancer prior
to their scheduled subsequent screening mammogram were
excluded. Breast cancers were identified from the Alberta Cancer
Registry (the third edition of International Classification of Disease
for Oncology (ICD-O-3) code C50 behaviors 2 and 3).
18
This study was approved by the ethics board at the University of
Alberta. Databases were linked using the unique provincial health
care identification number that was anonymized for data analysis.
Quality assurance and cross checks were performed to ensure
accuracy and completeness.
Index screening
Awoman’s first screening mammogram during the study period
was referred to as her index screen. The index screen test was
classified into either normal screening result group (BI-RADS
score 1 or 2) or abnormal group. At least two of the following three
criteria were required for the index screen to be classified into the
abnormal group: 1) from test result: a BI-RADS* score 0, 3, 4 or 5
for the index screen; 2) from radiologist’s recommendation: an
immediate, 3-month or 6-month follow-up recommendation; and
3) from follow-up test: at least one breast-related diagnostic
procedure within 30 days of index mammogram. Data for which
criterion 1) and criterion 2) were not consistent were assumed to
have a data entry error and the test was classified according to
criterion 3). For example, a screening mammogram record of a
BI-RADS* score 5 with a recommendation for a follow-up in two
years would not occur in practice and is evidence of a data error.
To determine whether the BI-RADS* score or the recommendation
was incorrect, we used criterion 3) which reflects what actually
occurred. If a follow-up breast-related diagnosis procedure was not
identified within 30 days of the screening mammogram, the
screening mammogram was deemed normal. Since women
diagnosed with breast cancer during the study period were
excluded from the study, the abnormal group only consists of
false positives.
Breast cancer diagnostic follow-up procedures conducted in
response to a false positive were categorized as follows (in order
of decreasing invasiveness): open biopsy, needle sampling
(fine needle aspiration, core needle biopsy, and closed biopsy),
imaging-only follow-up (typically diagnostic mammography
and/or ultrasound), and no follow-up procedure. There
were two possible explanations for the “no follow-up
procedure”: 1) The follow-up test data were missing: this could
occur if the breast-related diagnostic tests were performed by
radiologists who are outside ASR (approximately 8% of
radiologists in Alberta), resulting in test data not being captured
in the ASR database; and 2) those women did not comply with
the recommendation.
Diagnostic resolution
The most invasive procedure within six months of an abnormal
index screen was deemed the diagnostic resolution procedure,
based on an adaptation of a previously validated algorithm.
19
The
corresponding procedure date was used to calculate time to
diagnostic resolution and served as the index date for the
screening retention period for women with false positives. For
women with no follow-up procedures, the index date for the
screening retention period was six months after the initial screen,
to account for possible missing follow-up test dates. For women in
the normal screening group, the date of the index screen was used
as the index date of the screening retention period.
Screening retention
Screening retention was defined as the receipt of a subsequent
screening mammogram between 9 and 30 months from the index
date (see Figure 1), as 30-month is consistent with the definition
for calculating retention rate in Canada.
20
FALSE POSITIVE IMPACTS SCREENING RETENTION
e540 REVUE CANADIENNE DE SANTÉ PUBLIQUE •VOL. 108, NO. 5-6
Statistical analysis
Screening retention rate was tabulated by the index screen result
and diagnostic follow-up procedure category, region of residence,
and time to diagnostic resolution. Region of residence was
categorized into the following three groups based on the
Regional Health Authorities (RHA) that existed at the beginning
of the study period: the RHAs that included Edmonton and Calgary
are classified as the metropolitan region; central and southern
Alberta are classified as small cities/rural region; northern Alberta is
classified as the remote region, where access to health care is most
limited (Figure 2). A histogram showing time from index screen to
rescreen in the study population is shown in Figure 3. A
multivariable log-binomial regression model was used to estimate
the risk ratios of fail-to-rescreen associated with invasiveness of
diagnostic procedure, region of residence, and time to diagnostic
resolution, adjusted for women’s age. SAS
®
9.4 (SAS Institute, Cary,
NC) was used for data management and analyses.
RESULTS
A total of 213 867 women were eligible and included in the study.
The index screen results for 20 105 (9.4%) was a false positive: the
most invasive follow-up procedures performed were imaging
studies for 16 695 (83.0%), needle biopsy for 243 (1.2%) and
open biopsy for 1499 (7.5%) of those with a false positive. The
benign biopsy rate is 8.1 per 1000 screen.
Retention rate
Table 1 shows the retention rates for normal and abnormal index
screen and each follow-up procedure type, stratified by region, time
to diagnostic resolution, and age group. The retention rates were
62.2% and 68.7% for the false positive and normal index screen
groups respectively. The unadjusted risk ratio was 0.9 (95% CI:
0.90–0.92). As the invasiveness of procedure increased, the
retention rate decreased. The retention rates were 64.0%, 57.6%,
and 39.8% for imaging-only follow-up, needle sampling and open
biopsy respectively. For women with no follow-up procedure after
an abnormal result, the retention rate was 65.0%.
The retention rate varied considerably across regions by
invasiveness of procedure. For women who resided in
metropolitan regions, the retention rate for those with an index
false positive result decreased with increasing invasiveness of
procedure (63.9% for imaging only follow-up, 53.1% for needle
sample and 37.2% for open biopsy). This trend did not exist for
residents outside the metropolitan areas, however, retention rate
was lowest for residents of the small cities/rural region who
received an open biopsy. Women with false positives had lower
retention rate than those with normal results across all regions. The
retention rate increased with increasing rurality (67.9% for
metropolitan region, 68.7% for small cities/rural region and
70.5% for remote region). A longer time to diagnostic resolution
was also associated with a lower retention rate for those who
received open biopsy: 48% same day, 41% within 1 month, and
36% within 6 months, but not for those who received imaging
only (63%, 65% and 65% respectively).
Women aged 50–59 and 60–67 have similar rescreen rates by
invasiveness of procedure (Table 1).
Time from index to rescreen
The provincial screening guideline recommended women aged
50–69 receive screening for breast cancer at least every two years
during the study period. The screening retention peaked close to
one year (12 months) from the index date (Figure 3). A second peak
of screening retention occurred close to two years from the index
date, but the number of women at the second peak was much
lower than the number at the first peak. Among women who
rescreened within 30 months, approximately 50% had their
rescreening mammograms within 15 months of their index screen.
Log-binomial regression analysis of fail-to-rescreen
Figure 4 illustrates the adjusted risk ratio estimates of factors
associated with fail-to-rescreen. Age is adjusted in the model using
a cubic spline. Compared with women who had a normal index
screen result, the adjusted risk ratios of fail-to-rescreen were
1.08 (95% CI: 1.05–1.12) in women who had imaging-only
Figure 1. Diagram of outcome definition
FALSE POSITIVE IMPACTS SCREENING RETENTION
CANADIAN JOURNAL OF PUBLIC HEALTH •VOL. 108, NO. 5-6, 2017 e541
follow-up, 1.72 (95% CI: 1.44–2.07) in women who had needle
sampling and 2.29 (95% CI: 2.09–2.50) among women who had
open biopsy.
Screening retention rates varied to a smaller extent across regions
after adjusting for other factors. In small cities/rural and remote
regions, the risk ratios are 0.99 (95% CI: 0.98–1.00; p=0.193) and
0.96 (95% CI: 0.94–1.00; p=0.026) respectively, compared to
the metropolitan region. Time to diagnostic resolution was not
significantly associated with the fail-to-rescreen in the
multivariable regression analysis.
DISCUSSION
The factor most strongly related to screening retention after a false
positive screening result was the procedure used for diagnostic
resolution: screening retention decreased with increasing
invasiveness of the diagnostic resolution procedure. Women who
had an open biopsy were 2.3 times less likely to be rescreened
within 30 months after their diagnostic resolution compared to
those who had a normal index screen. The negative effect of false
positives on screening retention is consistent with the findings
from two-decade-old Canadian studies
11,21
and a more recent study
conducted in Spain, where breast cancer screening is also free and
is recommended biennially.
22
Retention rates were higher in the
Spanish study than in ours: retention rates for women with false
positive vs. normal results were 78.3% vs. 81.9%, compared to ours
which were 68.1% vs. 68.7% respectively. Both studies found
significantly lower retention rates, 66.5% (Spain) and 45.8%
(Alberta) for those who underwent invasive procedures, including
aspiration, closed biopsy and/or open biopsy.
22
Invasive follow-up
tests have been shown to create psychological distress in the
context of false positive breast cancer screens,
23
which can last for
up to three years;
7,24
it is likely that psychological distress plays an
important role in screening retention.
We did not find a statistical association between time to
diagnostic resolution and retention rate in our study. This is
consistent with a study conducted in The Netherlands.
25
Women
residing in remote regions were more likely to rescreen compared
to the women residing in metropolitan regions. This may reflect
that the availability of care is valued and acted upon in remote
regions where the health care resources are limited.
Fifty percent of the women in our study were rescreened within
15 months in spite of provincial guidelines at the time for biennial
screening. This is of concern as more frequent screening results in
higher cumulative false positive findings,
26
which in turn increases
the risk of invasive procedures and, based on our study and others,
6
lowers screening retention. In addition, modeling studies show
that biennial screening does not lead to higher prevalence of late-
stage breast cancer than annual screening.
27
More frequent
screening mammography also increases women’s exposure to
X-ray and cost to the publicly funded health system. Combined,
these facts suggest that annual breast cancer screening is
unnecessary, and may even be harmful, for average-risk women.
Currently, the Canadian guidelines recommend breast cancer
screening for women aged 50–74 every 2–3 years, although these
guidelines have not been embraced in all provinces.
1
Although invasive procedures appear to reduce breast cancer
screening retention rates, sometimes invasive tests are necessary to
Figure 2. Region division. Metropolitan region: Regions of
Edmonton and Calgary, i.e., R6 and R3. Small cities/
rural region: R1, R2, R4, R5 and R7. Remote region:
R8 and R9
Figure 3. Proportion of rescreening mammograms performed
each month of the total conducted between 9 and
30 months of the index screen
FALSE POSITIVE IMPACTS SCREENING RETENTION
e542 REVUE CANADIENNE DE SANTÉ PUBLIQUE •VOL. 108, NO. 5-6
determine whether a tumour is present. Identifying and
implementing factors that positively contribute to screening
behaviours is therefore important as they may overcome the
negative impact of invasive follow-up procedures. Tailored
invitation letters as well as motivational telephone calls have
been shown to positively impact breast cancer screening
behaviour.
28,29
The combination of a tailored letter and
motivational phone calls may improve screening retention
among those with false positive screening mammograms,
particularly those who have invasive follow-up tests.
Our study is the first population-based study to assess breast
cancer screening retention and factors related to it in Alberta, and
the most recent one in more than a decade in Canada. It provides
an updated and detailed picture of the screening retention in
Alberta, which is useful to screening programs in Canada and
elsewhere. A strength of the study was our use of population-based
data from the screening program, however, there are a few notable
limitations to our analyses, largely based on the data available for
the dataset. First, we were not able to determine and adjust for
whether the index screen in our study was the initial screen for a
particular woman. One study reported that initial screening had
higher false positive results (12% vs. 6%) and lower retention rate
(70% vs. 81%).
8
A recent report from the Canadian Partnership
Against Cancer (CPAC) also found that women with initial
screening had lower retention rates compared to those with
subsequent screening. Screening participation rates in Alberta
have been relatively stable, however –between 55% and 60% in
the last 10 years
13,30,31
–so we expect that our adjustment for age
in the multivariable regression analysis mitigates the confounding
effect of the index screen status.
The second limitation is that we did not have access to detailed
demographic data; some factors have been found to be associated
with screening rates.
23,32
The third limitation is that about 8% of
the biopsy data are estimated to be missing from the ASR database.
This could lead to a slight underestimation of the odds of fail-to-
rescreen for women who had benign biopsies, which means the
reported effect size is likely to be conservative. In contrast, the fourth
limitation –limiting the definition of rescreen to 30 months –
may overestimate the effect size. It is possible that a higher
proportion of women who had false positive results at their index
screen were rescreened more than 30 months after their index
screen than those with a true negative. Thirty months, however,
has been used consistently in studies
8,13
to define rescreen rates, so
our analysis is comparable to previous reported rates. Furthermore,
it is unlikely that even if the rescreen definition were extended to
include screens within 36 months, the more than 10% difference
in retention rates for those women who had needle sampling or
open biopsy would be eliminated. Last, the follow-up time for
105 women who had an abnormal index screen was slightly less
than 30 months (the median follow-up time was 28 months) and
they were categorized as “fail-to-rescreen”; as this group only
accounted for 0.5% of all women with an abnormal index screen,
the expected bias introduced is negligible.
In order to maximize the benefits of breast cancer screening in
populations and individuals, greater efforts are needed to
minimize both the risk of false positives as well as their burden
on women, which affect likelihood of rescreening. Suggested
tactics include screening average-risk women no more than
Table 1. Frequency of procedure and retention rate (% screened within 30 months of index date) by invasiveness of procedure, stratified by region and time to diagnostic resolution
Normal index
screen False positive index screening Total
Imaging-only
follow-up Needle
sampling Open biopsy No follow-up procedure
within 6 months Total n(retention rate)
n(retention
rate)
n(retention
rate)
n(retention
rate)
n(retention
rate)
n(retention rate) n(retention
rate)
Overall 193 762 (68.7) 16 695 (64.0) 243 (57.6) 1499 (39.8) 1668 (65.0) 20 105 (62.2) 213 867 (68.1)
Region*
Metropolitan 145 860 (68.5) 13 838 (63.9) 162 (53.1) 1201 (37.2) 1247 (65.0) 16 448 (61.9) 162 308 (67.9)
Small city/rural 42 119 (69.1) 2584 (65.1) 62 (66.1) 254 (46.9) 202 (57.4) 3102 (63.1) 45221 (68.7)
Remote 5783 (70.9) 273 (61.5) 19 (68.4) 44 (68.2) 219 (71.7) 555 (66.3) 6338 (70.5)
Time to diagnostic resolution
Same day 193 762 (68.7) 8959 (63.2) 47 (63.8) 110 (48.2) NA 9116 (63.0) 202 878 (68.4)
Within 1 month 0 6825 (65.0) 40 (35.0) 778 (41.4) NA 7643 (62.4) 7643 (51.9)
Within 6 months 0 911 (65.2) 156 (61.5) 611 (36.2) NA 1678 (54.3) 1678 (46.6)
No follow-up procedure within 6 months 0 NA NA NA 1668 (65.0) 1668 (65.0) 1668 (65.0)
Age (years)
50–59 131 006 (68.6) 11 444 (63.6) 176 (59.1) 980 (40.4) 1114 (64.9) 13714 (62.0) 144 720 (68.0)
60–67 62 756 (69.0) 5251 (64.9) 67 (53.7) 519 (38.5) 554 (65.2) 6391 (62.7) 69147 (68.4)
Note: Women with breast cancer during study period were excluded.
* Region: region division can be found in Figure 2.
NA =not applicable.
FALSE POSITIVE IMPACTS SCREENING RETENTION
CANADIAN JOURNAL OF PUBLIC HEALTH •VOL. 108, NO. 5-6, 2017 e543
biennially, minimizing invasive testing and providing targeted
communication with those with a previous false positive screen.
Further improvements in technology are also needed to decrease
false positive rates and reduce the need for invasive follow-up
tests.
REFERENCES
1. Canadian Task Force on Preventive Health Care, Tonelli M, Connor Gorber S,
Joffres M, Dickinson J, Singh H, et al. Recommendations on screening for
breast cancer in average-risk women aged 40–74 years. CMAJ 2011;
183(17):1991–2001. PMID: 22106103. doi: 10.1503/cmaj.110334.
2. US Preventive Services Task Force. Screening for breast cancer: U.S. Preventive
Services Task Force recommendation statement. Ann Intern Med 2009;
151(10):716–26, W-236. PMID: 19920272. doi: 10.7326/0003-4819-151-10-
200911170-00008.
3. Altobelli E, Lattanzi A. Breast cancer in European Union: An update of
screening programmes as of March 2014 (review). Int J Oncol 2014;
45(5):1785–92. PMID: 25174328. doi: 10.3892/ijo.2014.2632.
4. Duffy SW, Tabar L, Olsen AH, Vitak B, Allgood PC, Chen TH, et al. Absolute
numbers of lives saved and overdiagnosis in breast cancer screening, from a
randomized trial and from the Breast Screening Programme in England. J Med
Screen 2010;17(1):25–30. PMID: 20356942. doi: 10.1258/jms.2009.009094.
5. Nyström L, Andersson I, Bjurstam N, Frisell J, Nordenskjöld B, Rutqvist LE.
Long-term effects of mammography screening: Updated overview of the
Swedish randomised trials. Lancet 2002;359(9310):909–19. PMID: 11918907.
doi: 10.1016/S0140-6736(02)08020-0.
6. Maxwell AJ, Beattie C, Lavelle J, Lyburn I, Sinnatamby R, Garnett S, et al. The
effect of false positive breast screening examinations on subsequent
attendance: Retrospective cohort study. J Med Screen 2013;20(2):91–98.
PMID: 24009091. doi: 10.1177/0969141313499147.
7. Brodersen J, Siersma VD. Long-term psychosocial consequences of
false-positive screening mammography. Ann Fam Med 2013;11(2):106–15.
PMID: 23508596. doi: 10.1370/afm.1466.
8. Canadian Partnership Against Cancer. Organized Breast Cancer Screening
Programs in Canada: Report on Program Performance in 2007 and 2008.
Toronto, ON: Canadian Partnership Against Cancer, 2013.
9. Nelson HD, O’Meara ES, Kerlikowske K, Balch S, Miglioretti D. Factors
associated with rates of false-positive and false-negative results from digital
mammography screening: An analysis of registry data. Ann Intern Med 2016;
164(4):226–35. PMID: 26756902. doi: 10.7326/M15-0971.
10. Brewer NT, Salz T, Lillie SE. Systematic review: The long-term effects of
false-positive mammograms. Ann Intern Med 2007;146(7):502–10. PMID:
17404352.
11. Johnson MM, Hislop TG, Kan L, Coldman AJ, Lai A. Compliance with the
screening mammography program of British Columbia: Will she return?
Can J Public Health 1996;87(3):176–80. PMID: 8771920.
12. Chiarelli AM, Moravan V, Halapy E, Majpruz V, Mai V, Tatla RK. False-positive
result and reattendance in the Ontario Breast Screening Program. J Med Screen
2003;10(3):129–33. PMID: 14561264.
13. Public Health Agency of Canada. Organized Breast Cancer Screening Programs in
Canada –Report on Program Performance in 2005 and 2006. Ottawa, ON: PHAC,
2011.
14. Alberta Society of Radiologists. What is the Alberta Society of Radiologists (ASR)?.
Available at: http://www.radiologists.ab.ca (Accessed September 1, 2016).
15. Taylor K, Britton P, O’Keeffe S, Wallis M. Quantification of the UK 5-point
breast imaging classification and mapping to BI-RADS to facilitate comparison
with international literature. Br J Radiol 2014;84(1007):1005–10. doi: 10.
1259/bjr/48490964.
16. Alberta Health Services Cancer Screening Programs. Bringing Good Breast
Health to You. Calgary, AB: Alberta Health Services Cancer Screening Programs.
Available at: http://screeningforlife.ca/screentest (Accessed September 1,
2016).
17. Towards Optimal Practice (TOP) Working Group for Breast Cancer Screening.
Breast Cancer Screening: Clinical Practice Guideline. Edmonton, AB: Toward
Optimized Practice, 2013. Available at: http://www.topalbertadoctors.org
(Accessed September 1, 2016).
18. Percy C, Muir CS, Holten VV, World Health Organization. International
Classification of Diseases for Oncology. Geneva, Switzerland: WHO, 1990.
19. Yuan Y, Li M, Yang J, Winget M. Using administrative data to estimate time to
breast cancer diagnosis and percent of screen-detected breast cancers –A
validation study in Alberta, Canada. Eur J Cancer Care 2015;24(3):367–75.
PMID: 25521706. doi: 10.1111/ecc.12277.
20. Canadian Partnership Against Cancer. Report From the Evaluation Indicators
Working Group: Guidelines for Monitoring Breast Cancer Screening Program
Performance, 3rd ed. Toronto, ON: Canadian Partnership Against Cancer,
2013.
21. Chiarelli AM, Halapy E, Nadalin V, Shumak R, O’Malley F, Mai V.
Performance measures from 10 years of breast screening in the Ontario
Breast Screening Program, 1990/91 to 2000. Eur J Cancer Prev 2006;15(1):
34–42. PMID: 16374227.
22. Alamo-Junquera D, Murta-Nascimento C, Macia F, Bare M, Galceran J,
Ascunce N, et al. Effect of false-positive results on reattendance at breast
cancer screening programmes in Spain. Eur J Public Health 2012;22(3):404–8.
PMID: 21558152. doi: 10.1093/eurpub/ckr057.
23. Bond M, Pavey T, Welch K, Cooper C, GarsideR, Dean S, et al. Systematic review
of the psychological consequences of false-positive screening mammograms.
Health Technol Assess 2013;17(13):1–170, v–vi. PMID: 23540978. doi: 10.3310/
hta17130.
24. Bond M, Pavey T, Welch K, Cooper C, Garside R, Dean S, et al. Psychological
consequences of false-positive screening mammograms in the UK. Evid Based
Med 2013;18(2):54–61. PMID: 22859786. doi: 10.1136/eb-2012-100608.
25. Drossaert CH, Boer H, Seydel ER. Monitoring women’s experiences during
three rounds of breast cancer screening: Results from a longitudinal study.
J Med Screen 2002;9(4):168–75. PMID: 12518007. doi: 10.1136/jms.9.4.168.
Figure 4. Adjusted risk ratios of fail-to-rescreen. Log binomial regression model includes test procedure, region, time to diagnostic
resolution, and age (adjusted as a cubic spline)
FALSE POSITIVE IMPACTS SCREENING RETENTION
e544 REVUE CANADIENNE DE SANTÉ PUBLIQUE •VOL. 108, NO. 5-6
26. Gur D, Sumkin JH, Hardesty LA, Clearfield RJ, Cohen CS, Ganott MA, et al.
Recall and detection rates in screening mammography. Cancer 2004;
100(8):1590–94. PMID: 15073844. doi: 10.1002/cncr.20053.
27. White E, Miglioretti DL, Yankaskas BC, Geller BM, Rosenberg RD, Kerlikowske
K, et al. Biennial versus annual mammography and the risk of late-stage breast
cancer. J Natl Cancer Inst 2004;96(24):1832–39. PMID: 15601639. doi: 10.1093/
jnci/djh337.
28. Jibaja-Weiss ML, Volk RJ, Kingery P, Smith QW, Holcomb JD. Tailored
messages for breast and cervical cancer screening of low-income and minority
women using medical records data. Patient Educ Couns 2003;50(2):123–32.
PMID: 12781927.
29. Taplin SH, Barlow WE, Ludman E, MacLehos R, Meyer DM, Seger D, et al.
Testing reminder and motivational telephone calls to increase screening
mammography: A randomized study. J Natl Cancer Inst 2000;92(3):233–42.
PMID: 10655440.
30. Alberta Health Services. Alberta Health Services Annual Report 2015–2016.
Edmonton, AB: Alberta Health Services, 2016.
31. Statistics Canada. Table 051-0001 –Estimates of population, by age group and sex
for July 1, Canada, provinces and territories, annual, CANSIM (database). Ottawa,
ON: Statistics Canada, 2016.
32. Vahabi M, Lofters A, Kumar M, Glazier RH. Breast cancer screening disparities
among urban immigrants: A population-based study in Ontario, Canada. BMC
Public Health 2015;15:679. PMID: 26194189. doi: 10.1186/s12889-015-2050-5.
Received: March 9, 2017
Accepted: May 28, 2017
RÉSUMÉ
OBJECTIFS : L’incidence des faux positifs sur la fidélisation au dépistage du
cancer du sein varie dans les études internationales. Nous étudions les
facteurs associés à la fidélisation au dépistage en Alberta, au Canada, dont
les résultats faux positifs au dépistage, le caractère invasif de la méthode
diagnostique et la variation spatiale.
MÉTHODE : En tout, 213 867 femmes de 50 à 67 ans ayant subi une
mammographie de dépistage indicielle entre juillet 2006 et juin 2008 ont
été évaluées 30 mois après pour déterminer le taux de fidélisation au
dépistage. Les associations entre la fidélisation au dépistage et le caractère
invasif de la méthode diagnostique, le délai de résolution du diagnostic et la
région de résidence ont été étudiées par régression log-binomiale
multivariée avec ajustement en fonction de l’âge des femmes.
RÉSULTATS : Les femmes ayant obtenu des résultats faux positifs au
dépistage étaient moins susceptibles de retourner subir leur prochain
dépistage recommandé que celles ayant obtenu des résultats vrais négatifs
(62,0 % c. 68,7 %). Comparativement aux femmes ayant obtenu des
résultats normaux au dépistage, les risques relatifs ajustés des femmes
n’ayant pas subi un dépistage ultérieur étaient de 1,08 (IC de 95 % :
1,05–1,12) pour le suivi avec imagerie seulement, de 1,72 (IC de 95 % :
1,44–2,07) pour le prélèvement à l’aide d’une aiguille et de 2,29
(IC de 95 % : 2,09–2,50) pour la biopsie ouverte. Les taux de fidélisation au
dépistage étaient légèrement plus élevés chez les résidentes des zones
rurales que chez celles des zones urbaines. Le délai de résolution du
diagnostic n’était pas associé à la fidélisation au dépistage. La fidélisation
au dépistage a culminé un an après la date indicielle du dépistage
précédent.
CONCLUSION : Il est nécessaire d’être plus sensibilisé à l’effet très
nuisible des biopsies sur la fidélisation au dépistage en cas de résultats
faux positifs. Une telle sensibilisation peut éclairer les stratégies
d’intervention pour atténuer cet effet et améliorer les taux de fidélisation
au dépistage.
MOTS CLÉS : cancer du sein; dépistage; fidélisation; faux positifs
FALSE POSITIVE IMPACTS SCREENING RETENTION
CANADIAN JOURNAL OF PUBLIC HEALTH •VOL. 108, NO. 5-6, 2017 e545
