Community health workers and nurses affiliated with 7 health centers received the breast cancer early detection trainings in April-May 2015

Community health workers and nurses affiliated with 7 health centers received the breast cancer early detection trainings in November-December 2015

6 health centers served as controls and did not receive training during the study period

Weeklong training in signs, symptoms and treatability of breast cancer, performance of clinical breast exam, and structured algorithms for management of breast complaints followed by regular mentorship for health center nurses. CHWs received a one-day training in breast awareness and how to educate communities about breast health.

Nurses' knowledge will be assessed through written tests administered immediately before and immediately after trainings, and delayed post-tests administered 3 to 6 months after training. The investigators will use an adapted instrument that incorporates the Breast Cancer Knowledge Test and other surveys to assess clinician knowledge about breast cancer risk, signs and symptoms, and treatability. Questions were developed through review of existing instruments used in the United States, Mexico, the United Kingdom, and Nigeria, and expert input, feedback from clinical colleagues in Rwanda, and pilot testing with Rwandan hospital-based nurses. We will assess each nurse's overall test score as the total percentage of questions correct and assess scores within specific knowledge domains. The more questions answered correctly and the higher the score, the better the outcome (the score represents the percent of questions answered correctly).

CHWs' knowledge will be assessed through written tests administered immediately before and immediately after trainings, and delayed post-tests administered 3 to 6 months after training. The investigators will use an adapted instrument that incorporates the Breast Cancer Knowledge Test and other surveys to assess clinician knowledge about breast cancer risk, signs and symptoms, and treatability. Questions were developed through review of existing instruments used in the United States, Mexico, the United Kingdom, and Nigeria, and expert input, feedback from clinical colleagues in Rwanda, and pilot testing with Rwandan hospital-based nurses. We will assess each CHW's overall test score as the total percentage of questions correct and assess scores within specific knowledge domains. The more questions answered correctly and the higher the score, the better the outcome (the score represents the percent of questions answered correctly).

Nurses' performance scores on clinical breast exam checklists that assess completion of the steps required for high-quality clinical breast exam and are based on other published checklists adapted to the rural Rwandan setting. CBE materials were based on CBE guidelines and published and unpublished curricula adapted to our setting. During her regular health center visits, the breast health mentor will assess nurses' clinical skills in their clinics by using a checklist that includes the same questions as the pre- and post-test assessment form. When possible, the mentor will assess individual nurses by using the same checklist two or more times during a given clinical session, including one at the beginning of a session. We will analyze checklists completed between May 2015 and September 2016. The more actions performed correctly and the higher the score, the better the outcome (the score represents the percent of actions performed correctly).

Number of patients per month with breast concerns presenting at health centers before versus after the intervention, and compared with control health centers.