Preliminary Study On Risk Assessment Model Of Breast Cancer For High Risk Women In Bengbu

Objective: To screen women the risk factors of breast cancer in Bengbu, the risk assessment model of breast cancer is initially established, which is distinguished by cut-off point with low, medium and high risk groups in Bengbu. The probability of breast cancer is also assessed by the risk assessment model.Methods: The case-control study was performed at surgical oncology and general surgical unit in the first affiliated hospital of Bengbu medical college(47),the second affiliated hospital(9) of Bengbu medical college and the third people’s hospital of Bengbu(15) from March to November in 2015. It contains 71 cases of breast cancers who have been diagnosed by pathologically and 91 controls who were selected by age(±2 years old) according to the cases, They are all local people. The controls were physical examined healthy during the same period in the same hospital healthy center.Using binary Logistic regression to screen risk factors of breast cancer and establish risk assessment model based on the risk factors. Using Fisher discriminant analysis to evaluate the model; Calculating the proportion under the ROC curve to observe the effectiveness of the model and finding the cutoff point for low, medium and high risk of breast cancer from the ROC curve.Results: 1. The single influence factors associated with breast cancer are:(1) general demographic information: the educational status, occupation, family monthly income,medical insurance and body mass index.(2)reproductive factors: the number of birth and the frequency of abortion.(3) diet: beans and bean products, eggs and milk, fried food, white meat, potato and the types of drinking water.(4) sleep condition: time of sleep, sleeping with a bra and the satisfaction of sleep.(5) behavior habits: physical exercise and the amount of exercise.(6)environmental factors: the residence and the presence of pollution sources around living environment.(7)psychological factors: the satisfaction of life.(8)The cognition of breast cancer and screening behavior: cognitive condition, screening for breast cancer.2. Multiple factors binary Logistic regression of the main risk factors are: family economic conditions, consumption of soy products, negative emotional comfort, breastcancer screening.3. Risk assessment model predict the low, medium and high-risk groups with probability value P≤0.49 for low-risk sub-populations, P≥0.51 for high-risk sub-populations, 0.49<P<0.51 for medium.Conclusion: The model can assess the probability of breast cancer of Bengbu women with specific risk factors as well as provide the probable basis for screening criteria.