Development of Predictive Models for Survival among Women with Breast Cancer in Malaysia

Abstract

Prediction of survival probabilities based on models developed by other countries has shown inconsistent findings among Malaysian patients. This study aimed to develop predictive models for survival among women with breast cancer in Malaysia. A retrospective cohort study was conducted involving patients who were diagnosed between 2012 and 2016 in seven breast cancer centres, where their survival status was followed until 31 December 2021. A total of 13 predictors were selected to model five-year survival probabilities by applying Cox proportional hazards (PH), artificial neural networks (ANN), and decision tree (DT) classification analysis. The random-split dataset strategy was used to develop and measure the models’ performance. Among 1006 patients, the majority were Malay, with ductal carcinoma, hormone-sensitive, HER2-negative, at T2-, N1-stage, without metastasis, received surgery and chemotherapy. The estimated five-year survival rate was 60.5% (95% CI: 57.6, 63.6). For Cox PH, the c-index was 0.82 for model derivation and 0.81 for validation. The model was well-calibrated. The Cox PH model outperformed the DT and ANN models in most performance indices, with the Cox PH model having the highest accuracy of 0.841. The accuracies of the DT and ANN models were 0.811 and 0.821, respectively. The Cox PH model is more useful for survival prediction in this study’s setting.

myBeST—A Web-Based Survival Prognostic Tool for Women with Breast Cancer in Malaysia: Development Process and Preliminary Validation Study

“…The website’s main feature is individualised survival prediction. The prediction was based on the previously-developed prognostic model [ 26 ]. The prediction model aimed to provide personalised five-year survival probability predictions based on women with breast cancer in Malaysia.…”

“…New predictors could not be added as the original prognostic model did not include the suggested predictors. We added a link to the recently published article describing the model’s development [ 26 ]. We kept the prognostic tool on a separate page with a link on the front page to accustom the user to the aim, target use, overview, and limitations of the web-based prognostic tool.…”

“…As any instrument or tool requires some form of validation for the intended user, the focus of our current paper was the content and face validity of the model presented as a web-based prognostic tool. The validity of the model algorithm has been described in our previous paper [ 26 ]. Preliminary validity assessments showed encouraging responses from the medical specialists and medical officers.…”

“…Input obtained from the previous steps was utilised to develop and deploy the website. The model’s algorithm [ 26 ] formed part of the backend code and was deployed as a functioning survival prediction tool (i.e., front-end or user interface) using “Shiny”, an R software package. Subsequently, the tool was embedded in a WordPress template website.…”

“…Due to the limitations of the aforementioned tools derived from the Western population [ 18 , 20 , 21 ], we have developed a locally adapted prognostic model. The model’s algorithm showed robust performance (c-index: 0.81, integrated Brier score: 0.122, and area under the receiver operating characteristics curve: 0.891) [ 26 ].…”