Managers in public health authorities must constantly encourage adherence to breast cancer screening programs to optimize early detection and reduce the associated morbidity and mortality at the population level
[18, 19]. Participation is naturally a strong prerequisite of the program’s impact; the lack of adherence to the programs can be attributed to a number of factors, including long distances to travel for access to medical facilities
[20–22], unavailability of services, low socioeconomic status
[21, 22], etc. Worldwide, participation in a first round of a breast cancer screening program varies widely, with a range of 10.6% to 54.2% in Canada
, 34% in Luxembourg
, 46% in Switzerland
, 48% in France
, 47% to 78.1% in Croatia
, 70.6% in Stockholm
, 71% in Copenhagen
, 73% in England
, 53% to 87% in Spain
, 75.7% in the Netherlands
 and 85% to 95% in Finland
. The participation rate of women in our program was 33.1% during the first 2-year round. This was lower than the rates reported in some nations
[17, 31], but it was similar to rates reported for other trials in the initial phases
[12, 23]. Non-participation in the screening program was partly due to the fact that some of the population had received mammography examinations outside the program. Those women opted to obtain private health insurance, rather than use public national health services
[32–34]. The lack of systematic data available on the performance of mammography screening has given rise to deep concern about the actual attendance in screening programs
In some European countries, women are identified from the national population registry with a unique personal identification number, and letters of invitation are sent efficiently with an appointment time. Thus, a high attendance rate is guaranteed
[13, 26, 27]. In Luxembourg, the attendance rate in opportunistic screening programs was reduced by adopting a policy of exclusively reimbursing women for mammographs performed in the context of an organized screening program
. However, this strategy was not feasible in Brazil, where the constitution guarantees all women the right to receive a mammography free of charge in the public health system.
In Canada, women in the target age range had access to mammography by participating in breast cancer screening programs organized provincially or territorially. The participation rate in seven of these provinces ranged from 10.6% to 54.2%, and five provinces had rates below 30%
. In one Canadian survey, 53.7% of women aged 50–69 years self-reported that they had had a screening mammogram within the previous 2 years; it was estimated that 21% of those women were screened through an organized program. The low attendance in organized screening programs may be attributed to: (1) a low capacity for mammography screening; (2) difficulty in shifting from preventive health care practices and opportunistic screening to organized, systematic screening; (3) an inability, in some jurisdictions, to access population information that would facilitate inviting women in the target age range; and (4) a lack of understanding or confidence in the potential benefits of regular mammography screening among women and physicians
In this study, there were several possible explanations for the low participation rate, including: (1) lack of a national registry that stored information on targeted women. The absence of this information may have hampered the ability to identify and access women targeted to participate in the program; (2) approximately 25% of the regional population covered by the breast cancer prevention program had private health insurance, which facilitated access to mammography outside our program; (3) lack of experience in community intervention. This was the first breast cancer screening program with mammography; (5) low adherence of physicians that worked in the municipalities; (6) lack of organization in the Brazilian Health System to accommodate breast cancer screening, and (7) inherent cultural concerns that the examinations may identify cancer and require a referral to and/or examination in a hospital, where they would be exposed to sick people. In the initial stages of this project, the Barretos Cancer Hospital offered mammography and confirmatory exams in the same building that provided the treatments, including chemotherapy, radiotherapy, and surgery. This created a problem due to the close association with confirmed cancer. The problem was resolved in December 2009, with the inauguration of the Barretos Cancer Hospital’s Prevention Department building, where only apparently healthy women were examined.
The number of participants in each age group in the Barretos Cancer Hospital screening program reflected the sizes of the age groups in the eligible population of the Barretos region. The 40 to 49-year-old groups comprised 48.2% of the screened population and 45.2% of the eligible population; the 50 to 59-year-old groups comprised 34.4% of the screened population and 31.9% of the eligible population; and the 60 to 69-year-old groups comprised 17.4% of the screened population and 22.8% of the eligible population. The results from mammogram examinations performed in women aged 40 to 49 years were not significantly different from those performed in women aged 50 to 59. Despite the low number of cases, the results suggested that it would be prudent to initiate mammographic screening in relatively young women. This would optimize the detection of early tumors, despite differences in causality and ethnicity among different regions in Brazil.
This study ratified the large demand for MU examinations; our results corroborated the assumption that the MU was an important tool for facilitating access to mammography
. The MU was a critical part of the strategy to overcome obstacles represented by remote areas, poor resources, and/or difficulty in accessing the public health system
. A national survey of mammography facilities in the US revealed that 2.4% of mammograms were performed in a MU, but the number of units assessed was low
. On the other hand, the Netherlands had 66 units, mostly mobile units, with an annual national coverage of 80%
; this provided evidence for the importance and feasibility of MUs.
The MU has been used in developed countries to service population groups with restrictions that prevented participation in screening for breast cancer. These groups generally comprised women with low incomes, low levels of education, and older age, in addition to other limitations, such as residing in a rural area
[38–41]. The MU approach has also been used to aid in organized screenings. In small counties, the MU was assumed to provide easy access to a mammogram examination for women who lived at a distance from cities with appropriate resources for breast examinations
. Our findings firmly supported this assumption. We found that 10 521 (58.6%) women sought mammograms in the MU, compared with 7443 (41.4%) women that were examined in the FU.
Radiologists in the MU conducted a daily average of 26.3 examinations, with an occupancy rate of 65.7%. In contrast, the FU showed an occupancy rate of 15.5%, with an average of 18.6 examinations per day. Clearly, the units were not operating at full capacity, but the MU in fact answered an underlying need of the population resident in the rural areas more than the FU did with respect to the urban population, which was already partially covered by opportunistic screening outside the program. This highlighted the urgency of implementing more comprehensive communication between the population and public health system to optimize the potential benefits of mammographic screening.
Women were invited to participate in the screening program through various sources, including the media, press, radio, letters, lectures, etc. Some of these methods were previously used in developed countries, such as the invitation letter, which resulted in a participation rate of 85%
. Other programs have combined this strategy with media campaigns; these approaches achieved success rates greater than 70%
. In our study, the best strategy for promoting participation was an individual suggestion given particularly by community healthcare workers (46.8%). The next most successful strategy was the recommendation from a physician; this persuaded 27.0% of women to have a mammogram. This finding is very important, because a significant percentage of women received information from their doctors.
In the present study, the recall rate of screening was 9.4%, slightly above the 7% recall rate that is most commonly recommended in population-based programs as the upper acceptable limit for the first screening round according to the European Guidelines for Breast Cancer Screening
. The Breast Cancer Surveillance Consortium of the U.S. National Cancer Institute, the national breast cancer early detection program of the U.S. Centers for Disease Control and Prevention, and the UK National Health Service Breast Screening Program had recall rates of 13.1%, 11.2%, and 7.4%, respectively, for the first round of screening
. The positive predictive value for biopsies was low (16%) compared with the indication from BI-RADS, which ranged from 25% to 40%; this indicated that many patients referred for biopsy had no lesions. This suggested that, although the radiologists had proven experience performing diagnostic mammography, apparently, they did not feel comfortable assessing and diagnosing patients with minimal mammographic alterations. This may have been because they had no specific training in breast radiology; in addition, we used stringent quality assurance and quality control guidelines
. The issue of professional training of medical and technical personnel in the assessment units of the screening programs should be among the priorities for consideration in further development of this project.
This program diagnosed 76 cases of breast cancer; the detection rate was 4.23 cases per 1000 examinations. In a large series from UK, a country with higher breast cancer incidence than Brazil, the detection rate ranged from 8.6 to 10.1 cases per 1000 examinations
. The absence of a population-based registry in this region of Brazil prevents the determination of breast cancer prevalence in these women for comparison. However, estimates of the INCA show a prevalence of 68.9 cases for 100,000 women in the State of Sao Paulo
One primary objective of screening programs is to identify early stage cancer. We found that 43.4% of tumors were in the early stages of development (EC 0 and EC I). Of note, both symptomatic and asymptomatic women were included in this class, because symptoms were not known initially. This frequency was higher than the 20% reported in the state of Sao Paulo, according to the Oncocentro Foundation of São Paulo
, and higher than the 14.5% observed in the Barretos region, according to the Cancer Registry of Barretos Cancer Hospital in the period prior to the start of the present screening program. Conversely, in patients with more advanced breast cancer, we observed significant improvements over previous observations, as a result of early screening. We found that tumors in advanced stages (CS III and CS IV) accounted for 15.8% of cancers detected in both units, compared with 39.0% from the state of São Paulo
 and 40.5% from previous reports of the Barretos region. These findings are important, because they demonstrate the efficiency and relevance of this screening program and they indicate the actual status of breast cancer in Brazil. This approach can lead to new screening programs, or it can be used to improve existing programs, even in the first round. Thus, our findings confirmed, in part, the importance of implementing an organized screening program for the prevention of breast cancer.
Although we obtained accurate data and successfully implemented a system of early detection for breast cancer, the major methodological limitation of this study was that we failed to separate symptomatic from asymptomatic women. This initiative was consistent with the philosophy of the Barretos Cancer Hospital, which included the Department of Prevention. Currently, there is a registry that distinguishes symptomatic from asymptomatic women, but both groups continue to be served by the screening program. In addition, there was a high demand for mammography in the region, which was confirmed by the high number of referrals from doctors for early detection screening. Consequently, the overall problem was large, due to the lack of a health policy priority system for mammographic screening. The early detection program described here was functional and provided acceptable results. The MU approach offers potential for progressive improvement in early detection of breast cancer.