Association of treatment delay and stage with mortality in breast cancer: a nationwide cohort study in Taiwan


Study design

This retrospective cohort study was conducted to evaluate the impact of DFTI and other related factors on mortality among female breast cancer patients. Confirmed breast cancer cases diagnosed between 2011 and 2017 were sampled and their survival followed until the end of 2018. The start date of follow-up was the date of first diagnosis and the end date was the date of loss of monitoring, death or the end of the study period (December 31, 2018), whichever occurred first.

Data sources

To select our sample participants, we accessed the Taiwan Cancer Registry published by the Health Promotion Administration, Ministry of Health and Welfare. The Taiwan Cancer Registry is a nationwide population-based cancer registry system that was established in 1979. The database includes detailed information on cancer stage, site-specific factors of cancer, treatment, and recurrence13. In addition, we linked these data to 2009–2018 data from the National Health Insurance Research Database (NHIRD), the Registry of Patients with Catastrophic Illness Database (RCIPD) and the Cause of Death File , Ministry of Health and Welfare. The RCIPD is a sub-database of the NHIRD, which contains the data of 99.99% of Taiwan’s population14. The RCIPD includes healthcare records of patients diagnosed as having any of 30 specified catastrophic diseases (such as malignancies, severe hereditary diseases, diseases or immune disorders)15. People with any catastrophic illness are exempt from consultation, pharmacy, treatment and hospitalization fees under the National Health Insurance. We extended 2 years of data (2009 and 2010) from the NHIRD to evaluate the Charlson comorbidity index (CCI) scores and other catastrophic diseases of the included patients. This study was approved by the institutional review board of Cheng Ching Hospital (IRB number: HP150004) and was conducted in accordance with the Declaration of Helsinki. Informed consent was waived by the Ethnic Research Committee of China Medical University Hospital.

Study participants

In total, 81,906 patients were newly diagnosed with breast cancer between 2011 and 2017. Their diagnosis was determined based on International Classification of Diseases for Oncology, Third Edition codes C50.0–C50.9. Of all patients, only 49,426 patients were included for further analysis (Fig. 3) as the following patients were excluded: male patients, patients diagnosed with any type of cancer before or after breast cancer diagnosis, patients who had any kind of catastrophic disease before. breast cancer, patients aged < 20 years, patients without a known date of initial treatment, patients at an unknown stage of cancer, patients with carcinoma in situ and patients without any immunohistochemical information.

Figure 3
figure 3

The process of selecting participants in the study.

Variable definitions and descriptions

We defined DFTI as the interval between the date of diagnosis of breast cancer based on biopsy and the date on which the first treatment was initiated. Four DFTI groups were used to separate included patients: ≤ 30, 31–60, 61–90, and ≥ 91 days. We then defined age as the age at which a patient received a confirmed diagnosis of breast cancer based on pathological findings. We categorized marital status as single, married, divorced, widowed, and missing (persons with unknown marital status) and grouped education levels into six categories. Patients’ income was based on their monthly salary. Environmental factors were based on the level of urbanization of patients’ areas of residence before cancer diagnosis; a total of seven levels – from highly developed urban cities (level 1) to remote districts (level 7) – were employed16.

The degree of comorbidity—a weighted index based on the presence of comorbid conditions within 2 years before cancer diagnosis—was classified into four levels based on CCI (Deyo’s CCI) scores.17. Tumor subtypes were classified based on estrogen receptor, progesterone receptor, and human growth factor/neu receptor status, as recorded by pathologists’ interpretation of the assays. The largest tumor dimension (in centimeters) determined through pathological examination was considered the tumor size. Regional lymph nodes were defined as the most proximal lymph nodes that served as immediate drainage sites for tumors, and these included axillary nodes, ipsilateral intramammary nodes, internal mammary nodes, and supraclavicular nodes. Cancer stages were categorized according to American Joint Committee on Cancer Staging Manual, Eighth edition. Patients were also classified based on the treatment they received within 6 months of breast cancer diagnosis. Other variables included hospital level and ownership of the treating hospital.

Main outcome measures

The primary outcome was cancer-specific mortality in patients with breast cancer. Duration of follow-up was defined as the duration from the date of diagnosis to the date of death or the end point of follow-up (December 31, 2018). Confirmation of death was made on the basis of administrative data (Case of Death File).

statistical analysis

For descriptive statistics, we included baseline characteristics, income, environmental factors, health status, tumor characteristics, stage, type of treatment, primary hospital information, and DFTI distribution.

The log-rank test and the Cox proportional hazards model were adopted for inferential statistics. First, a bivariate analysis was performed using the log-rank test to determine significant differences between survival status by the end of 2018 and DFTI or other variables. The adjusted Cox proportional hazards model was used to analyze the relative risk of mortality in breast cancer patients with different DFTI, after controlling for relevant variables. Next, we analyzed the impact of DFTI on the survival of patients at different cancer stages and of different ages. Finally, we estimated survival time according to adjusted survival curves for all breast cancer patients; stratification analyzes by tumor stage were performed to investigate the impacts of different DFTIs on patient survival.

In this study, SAS (version 9.4; SAS Institute, Cary, NC, USA) was used for data analysis, with the significance level (α) set at 0.05.


Leave a comment

Your email address will not be published. Required fields are marked *