Development of breast cancer risk assessment model

The research was carried out in cross-sectional and methodological design in order to develop a valid and reliable model for evaluating breast cancer risk in women. The sample of the research consisted of 800 people who applied to Adana City Education and Research Hospital and Çukurova University Balcalı Faculty of Medicine; 271 of them were breast cancer, 270 of them were at risk and 259 of them were healthy. The data were collected with the "Breast Cancer Risk Assessment Form" and "Healthy Life Style Scale" based on the literature information and benefiting from risk assessment tools such as Gail model, Claus model, Tyrer Cuizk(IBIS). Normality numerical variables were tested with Shapiro wilk test. In numerical data, Mann Whitney U test and Student t test were used to compare the 3 groups. Chi-square analysis was performed to investigate the relationship between categorical variables and breast cancer. Logistic regression with LASSO regularisation with 10-fold cross-validation was used to built models. PCA (Principal Component Analysis) scores were calculated for significant factors and performances of the models were compared in terms of their AUC values for test data. Sociodemographic characteristics, hormonal characteristics, family cancer history were found significant in the breast cancer group compared to other groups (p <0.05). The healthy lifestyle scale score was significantly lower in the breast cancer group compared to other groups (p <0.05). The healthy lifestyle scale score was significantly lower in the breast cancer group compared to other groups (p <0.05). The breast cancer risk calculation score of Gail and IBIS models was low. While LASSO logistic regression has different factors in classifying the patient-risk group, only 4 coefficients have been determined in the classification of the patient-healthy group with LASSO logistic regression. The model we developed has been found stronger in risk determination than the Gail and IBIS models. Keywords: Breast Cancer, Nursing, Risk Assessment, Model Development, Gail, IBIS