Background: The proposed study will contribute to an often-neglected area of cancer research, the evaluation of novel screening programs operating in low and middle-income countries (LMICs). There is an urgent need to improve early detection of cancer in LMICs, which typically do not have structured screening programs yet bear the highest burden of cancer morbidity and mortality [1, 2]. In Brazil, demographic shifts in the population composition and age distribution have led to an increase in the number of people at risk for developing cancer, thus contributing to the future challenges the healthcare system faces with respect to cancer prevention [3]. Inequalities in access to cancer screening are linked to geographic and socioeconomic disparities. Rural populations have fewer opportunities to participate in cancer screening and tend to receive cancer diagnoses upon clinical manifestation of disease (late stage) when therapy is less effective and survival probability is lower. In 2004, the Barretos Cancer Hospital (BCH) launched an innovative screening program for breast cancer, implemented with Mobile Units (MU) in rural, remote regions of Brazil. Today, there are 11 operating MUs equipped to perform diagnostic and clinical procedures. MUs provide screening exams free of charge across an immense geographic range of over 3 million square kilometers. Prior studies have assessed the active screening methods employed by the MUs and found they are effective in reaching vulnerable groups [4-7]. Epidemiologic evidence examining the cancer screening offered by the MUs is limited. Therefore, the objectives of my master's thesis research are: 1. To estimate the screening coverage of the mammography program in the Departmento Regional Saúde (DRS) of Araçatuba (II), Barretos (V), São Jose de Rio Preto (XV) and the micro-region of Jales. 2. To describe the overall performance of the MU breast cancer screening program in DRS II, V and XV by: a. Link program inputs (total number of screening exams) to program results, including the number of abnormal exams, biopsies and detected cases of cancer. b. Measure the efficacy/efficiency of the breast cancer referral system and identify factors associated with delay in response to abnormal test results using a time-to-event analysis. c. Investigate the change in stage-distribution of breast cancers detected by the BCH program. Data sources: The study population is comprised of women who have accessed mammography from a MU from 2010-2015 and who reside in one of 40 municipalities in DRS II, 18 municipalities in DRS V or 102 municipalities in DRS XV. De-identified patient data from the Prevention Institute of the Barretos Cancer Hospital for 155,632 mammography exams will be analyzed. Abnormal screening exams will be linked with data from SIS-ONCO medical health records system to extract data on follow-up examinations and TNM staging of detected cases. Population data on demographics and cancer epidemiology will be retrieved from the Brazil Institute of Geography and Economics (IBGE) and the Oncocentre Foundation of São Paulo - Tabnet program (FOSP - Tabnet). Methods: In objective 1, I will use population denominators from the 2010 National Census for women eligible for screening (aged 50-69) in order to determine the annual screening coverage rates attributable to the MU program per municipality from 2010 - 2015 [8]. A correlation matrix will be used to explore the effect of socioeconomic status (SES) per postal-code on screening coverage. SES will be measured at a municipal level using a social exclusion index which provides a composite measure of literacy, income and access to sanitation [9].In objective 2a, I will use routinely collected data from the mobile unit program from 2010-2015 to summarize the number of screening exams and BI-RADS results per region and municipality. Descriptive time-series statistics will be calculated for proportion of exams requiring further follow-up (recall rate), proportion receiving biopsies, case detection rate per 1000 screening exams, mean cancer size and cancer stage. In objective 2b, a Cox proportional hazard model will be used to identify explanatory variables associated with delays between receipt of an abnormal mammogram result and presentation at the fixed site clinic for diagnostic evaluations. Key co-variates in the Cox model will include distance from city of residence to the fixed city (Barretos or Fernandopolis), SES of postal-code, age, and BI-RADS score. In objective 2c, change in age-adjusted cancer stage distribution will be calculated using individual patient-level data from the MU screening program. Significance. Study findings will provide epidemiologic evidence of the benefits and harms of the MU program, expected to directly impact service delivery by providing information that can be used to improve coverage, quality, efficiency and effectiveness of screening services and referral systems. Findings will also be relevant to decision-making at local, institutional, national and international levels. Accordingly, new and similar interventions could be implemented geared at screening cancer in vulnerable populations. Facilitating and improving uptake of cancer screening would save lives, reduce costs of treating invasive cancer and potentially reduce inequalities in avoidable cancer mortality[10]. References 1. Bray F, McCarron P, Parkin DM. The changing global patterns of female breast cancer incidence and mortality. Breast Cancer Res. 2004;6(6):229-39. doi: 10.1186/bcr932. PubMed PMID: 15535852; PubMed Central PMCID: PMC1064079. 2. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127(12):2893-917. Epub 2011/02/26. doi: 10.1002/ijc.25516. PubMed PMID: 21351269. 3. IBGE. Indicadores Sociodemográficos e de Saúde no Brasil Rio de Janeiro: Instituto Brasileiro de Geografia e Estatística - IBGE, 2009. 4. Vieira RA, Lourenco TS, Mauad EC, Moreira Filho VG, Peres SV, Silva TB, et al. Barriers related to non-adherence in a mammography breast-screening program during the implementation period in the interior of Sao Paulo State, Brazil. J Epidemiol Glob Health. 2015;5(3):211-9. Epub 2015/08/02. doi: 10.1016/j.jegh.2014.09.007. PubMed PMID: 26231397. 5. De Castro Mattos JS, Mauad EC, Syrjanen K, Longatto-Filho A, Haikel RL, Da Costa Vieira RA, et al. The impact of breast cancer screening among younger women in the Barretos Region, Brazil. Anticancer Res. 2013;33(6):2651-5. Epub 2013/06/12. PubMed PMID: 23749923. 6. Haikel RL, Jr., Mauad EC, Silva TB, Mattos JS, Chala LF, Longatto-Filho A, et al. Mammography-based screening program: preliminary results from a first 2-year round in a Brazilian region using mobile and fixed units. BMC Women's Health. 2012;12:32. Epub 2012/10/04. doi: 10.1186/1472-6874-12-32. PubMed PMID: 23031787; PubMed Central PMCID: PMC3532077. 7. Silva TB, Mauad EC, Carvalho AL, Jacobs LA, Shulman LN. Difficulties in implementing an organized screening program for breast cancer in Brazil with emphasis on diagnostic methods. Rural Remote Health. 2013;13(2):2321. Epub 2013/04/20. PubMed PMID: 23597169. 8. IBGE. Atlas do censo demográfico 2010. Rio de Janeiro2013. 9. De Jesus Sousa Lemos J. Mapa da Exclusão Social No Brasil: Radiografia de um País Assimetricamente Pobre. Fortaleza: Banco do Nordeste do Brasil, 2012. 10. Miller A. Screening. In: Franco E, Rohan T, editors. Cancer Precursors: Epidemiology, Detection, and Protection. New York Springer; 2002. p. 365-73.