• Korean Journal of Radiology
• /
• v.22 no.10
• /
• pp.1697-1707
• /
• 2021

The recent introduction of various high-dimensional modeling methods, such as radiomics and deep learning, has created a much greater diversity in modeling approaches for survival prediction (or, more generally, time-to-event prediction). The newness of the recent modeling approaches and unfamiliarity with the model outputs may confuse some researchers and practitioners about the evaluation of the performance of such models. Methodological literacy to critically appraise the performance evaluation of the models and, ideally, the ability to conduct such an evaluation would be needed for those who want to develop models or apply them in practice. This article intends to provide intuitive, conceptual, and practical explanations of the statistical methods for evaluating the performance of survival prediction models with minimal usage of mathematical descriptions. It covers from conventional to deep learning methods, and emphasis has been placed on recent modeling approaches. This review article includes straightforward explanations of C indices (Harrell's C index, etc.), time-dependent receiver operating characteristic curve analysis, calibration plot, other methods for evaluating the calibration performance, and Brier score.

• Journal of the Korean Statistical Society
• /
• v.27 no.3
• /
• pp.359-368
• /
• 1998

Let ξ$_{p}$(z$_{0}$) be the pth quantile of the distribution of the survival time of an individual with time-invariant covariate vector z$_{0}$ in the additive risk model. We propose an estimator of (ξ$_{p}$(z$_{0}$) and derive its asymptotic distribution, and then construct an approximate confidence interval of ξ$_{p}$(z$_{0}$) . Simulation studies are carried out to investigate performance of the proposed estimator far practical sample sizes in terms of empirical coverage probabilities. Also, the estimator is illustrated on small cell lung cancer data taken from Ying, Jung, and Wei (1995) .d Wei (1995) .

• Health Policy and Management
• /
• v.31 no.1
• /
• pp.91-99
• /
• 2021

Background: The Korean government introduced National Cancer Control Program and strengthening national health insurance coverage for cancer patients. Although many positive effects have been observed, there are also many concerns about cancer management such as patient concentration or time-to-treatment. Thus, we investigated the association between the time-to-treatment and survival of cancer patients, and compared regional differences by time trend. Methods: The data used in this study were national health insurance claims data that included patients diagnosed with lung cancer and received surgical treatment between 2005 and 2015. We conducted survival analysis with Cox proportional hazard model for the association between time-to-treatment and survival in lung cancer. Additionally, we compared the regional differences for time-to-treatment by time trend. Results: A total of 842 lung cancer patients were included, and 52.3% of lung cancer patients received surgical treatment within 30 days. Patients who received surgical treatment after 31 days had higher 5-year or 1-year mortality compared to treatment within 30 days (5-year: hazard ratio [HR], 1.566; 1-year: HR, 1.555; p<0.05). There were some regional differences for time-to-treatment, but it was generally reduced after 2010. Conclusion: Delayed surgical treatment after diagnosis can negatively affect patient outcomes in cancer treatment. To improve cancer control strategies, there are needed to analyze the healthcare delivery system for cancer care considering the severity and types of cancer.

• Asian Pacific Journal of Cancer Prevention
• /
• v.17 no.1
• /
• pp.159-163
• /
• 2016

Background: Colorectal cancer (CRC) is the third most common cancer in the world, and the fourth in Iran in both genders. The aim of this study was to find predictive factors for CRC survival. Materials and Methods: Medical records of 570 patients referred to the radiotherapy oncology department of Shiraz Namazi hospital from 2005 to 2010 were retrospectively analysed. Data were collected by reviewing medical records, and by telephone interviews with patients. Survival analysis was performed using the Cox's regression model with survival probability estimated with Kaplan-Meier curve. The log-rank test was used to compare survival between strata. Data was analyzed with Stata 12. Results: The five-year survival rate and the mean survival time after cancer diagnosis were 58.5% and $67{\pm}4months$. On multivariate analysis, age of diagnosis, disease stage and primary tumor site, lymphovascular invasion and type of treatment (in colon cancer) were significant factors for survival. Conclusions: Age of diagnosis and type of treatment (adjuvant therapy in patients with colon cancer) were two modifiable factors related to survival of CRC patients. Therefore earlier diagnosis might help increase survival.

• Journal of Information Technology Services
• /
• v.18 no.4
• /
• pp.151-170
• /
• 2019

This study conducted an empirical analysis through the Kaplan-Meier method, which is mainly used for clinical experiment analysis, on the survival rate and the survival duration of small and medium-sized enterprises using B2B e-commerce guarantee provided by credit guarantee institutions for activating B2B e-commerce transactions. The variables presented in this study are analyzed by the subdivision of the survival characteristics of enterprises using B2B e-commerce guaranteee by year, enterprises attribute, representative attribute, and guarantee use amount based on the variables tested through previous studies. According to the empirical analysis, SMEs using B2B e-commerce guarantees have a higher survival rate compared to general enterprises Simply by year and have a variety of survival characteristics, and most of the variables have a significant effect except for some variables. The implication of this study is that the researches conducted on enterprises participating in B2B e-commerce for a long period of time to support the establishment of stable business environment of SMEs and the results of empirical analysis on the survival characteristics are useful information to the stakeholders of B2B e-commerce And it can contribute to enhance the survival rate of related enterprises.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.19
• /
• pp.8483-8488
• /
• 2014

Background: Breast cancer is the most common cancer and the second most common cause of cancer-induced mortalities in Iranian women, following gastric carcinoma. The survival of these patients depends on several factors, which are very important to identify in order to understand the natural history of the disease. Materials and Methods: In this retrospective study, 313 consecutive women with pathologically-proven diagnosis of breast cancer who had been treated during a seven-year period (January 2006 until March 2014) at Towhid hospital, Sanandaj city, Kurdistan province of Iran, were recruited. The Kaplan-Meier method was used for data analysis, and finally those factors that showed significant association on univariate analysis were entered in a Cox regression model. Results: the mean age of patients was $46.10{\pm}10.81$ years. Based on Kaplan-Meier method median of survival time was 81 months and 5 year survival rate was $75%{\pm}0.43$. Tumor metastasis (HR=9.06, p=0.0001), relapse (HR=3.20, p=0.001), clinical stage of cancer (HR=2.30, p=0.03) and place of metastasis (p=0.0001) had significant associations with the survival rate variation. Patients with tumor metastasis had the lowest five-year survival rate (37%)and among them patients who had brain metastasis were in the worst condition (5 year survival rate= $11%{\pm}0.10$). Conclusions: Our findings support the observation that those women with higher stages of breast malignancies (especially with metastatic cancer) have less chance of surviving the disease. Furthermore, screening programs and early detection of breast cancer may help to increase the survival of those women who are at risk of breast cancer.

• Asian Pacific Journal of Cancer Prevention
• /
• v.14 no.1
• /
• pp.321-324
• /
• 2013

Background: Intra- and extrahepatic cholangiocarcinoma (CCA) is the most common cancer in Thailand, especially in the northeast region. Most extrahepatic CCA patients consult a doctor at a late stage. Surgery is still the best treatment. Objectives: The aim of this study was to evaluate survival rates and factors affecting survival in extrahepatic CCA patients following surgery at Srinagarind Hospital, Khon Kaen University, Thailand. Materials and Methods: A retrospective cohort study was conducted with 58 patients who were diagnosed and treated by surgical resection by the same surgeon at Srinagarind Hospital between 2005 and 2009. The patients were followed up until death or the end of the study (31 December, 2011). Survival rates were calculated by the Kaplan-Meier method, and the Cox proportional hazard model was used to identify independent prognostic factors. Results: The total follow-up time was 1,215 person-months, and the mortality rate was 50 per 100 person-years. The cumulative 1-, 3-, and 5-year survival rates were 62.1%, 21.7% and 10.8%, respectively. The median survival time after resection was 15 months. After adjusting for age, gender, lymph node metastasis and histological type, resection margin remained as a statistically significant prognostic factor for survival following surgery. A positive resection margin was associated with a 2.3-fold higher mortality rate than a negative margin. Conclusions: Resection margins are important prognostic factors affecting survival of extrahepatic CCA patients after surgery. A negative resection margin can reduce the mortality rate by 56%.

• Asian Pacific Journal of Cancer Prevention
• /
• v.17 no.1
• /
• pp.95-100
• /
• 2016

Background: Survival time of lymphoma patients can be estimated with the help of microarray technology. In this study, with the use of iterative Bayesian Model Averaging (BMA) method, survival time of Mantle Cell Lymphoma patients (MCL) was estimated and in reference to the findings, patients were divided into two high-risk and low-risk groups. Materials and Methods: In this study, gene expression data of MCL patients were used in order to select a subset of genes for survival analysis with microarray data, using the iterative BMA method. To evaluate the performance of the method, patients were divided into high-risk and low-risk based on their scores. Performance prediction was investigated using the log-rank test. The bioconductor package "iterativeBMAsurv" was applied with R statistical software for classification and survival analysis. Results: In this study, 25 genes associated with survival for MCL patients were identified across 132 selected models. The maximum likelihood estimate coefficients of the selected genes and the posterior probabilities of the selected models were obtained from training data. Using this method, patients could be separated into high-risk and low-risk groups with high significance (p<0.001). Conclusions: The iterative BMA algorithm has high precision and ability for survival analysis. This method is capable of identifying a few predictive variables associated with survival, among many variables in a set of microarray data. Therefore, it can be used as a low-cost diagnostic tool in clinical research.

• Asian Pacific Journal of Cancer Prevention
• /
• v.17 no.4
• /
• pp.2055-2060
• /
• 2016

Background: Colorectal cancer (CRC) is one of the most common causes of death worldwide and in Thailand. The X-ray repair cross-complementary protein 1 (XRCC1) is required for efficient DNA repair. The effects of this gene on survival in colorectal cancer remain controversial and have not been reported in Thailand. The aim of this study was to investigate the association of the XRCC1 gene with survival of colorectal cancer patients in a Thai population. Materials and Methods: Data and blood samples were collected from 255 newly diagnosed and pathologically confirmed CRC patients who were recruited during the period 2002 to 2006 and whose vital status was followed up until 31 October, 2014. Real-time PCR-HRM was used for genotype identification. The Kaplan-Meier method, the log-rank test, and Cox proportional hazard regression were used to estimate cumulative survival curves and compare various survival distributions and adjusted hazard ratios. Results: Most of the cases were males, and the median age was 55 years. The median survival time was 2.43 years. The cumulative 1-, 3-, 5-, 7-, and 10 year survival rates were 76.70%, 39.25%, 26.50%, 16.60% and 3.56%, respectively. After adjustment, female gender, ages 50-59 and ${\geq}60years$, tumour stage III+IV, a signet-ring cell carcinoma, and poor differentiation had significant associations with increased risk of CRC death. While the XRCC1 Arg/Arg homozygote appeared to be a risk factor for CRC death, the association was not significant. Conclusions: The genetic variant in the XRCC1 may not be associated with the survival of CRC patients in Thailand. Further studies are needed to verify our findings.

• Asian Pacific Journal of Cancer Prevention
• /
• v.14 no.2
• /
• pp.1107-1110
• /
• 2013

Intrahepatic cholangiocarcinoma (ICC), one of the primary liver cancers, is frequent in the northeastern part of Thailand. Surgical resection remains the best method of treatment, but patients suffering from ICC usually present at a late stage of the disease. Studies of survival and prognostic factors after surgery remain rare. The aim here was to evaluate the survival rate and factors affecting the survival of patients with intrahepatic cholangiocarcinoma after surgery. The study used a retrospective cohort design. The subjects were 73 consecutive patients with ICC, who were admitted for surgery to Srinagarind Hospital, Khon Kaen University, during the period 2005-2009. The censoring date was 31 December, 2011, data being evaluated using uni- and multivariate analyses. Postoperative survival analysis was performed by the Kaplan-Meier method, and the Cox proportional hazard model was used to identify independent prognostic factors. The total follow-up time was 99 person-years. The total number of deaths was 59, giving a mortality rate of 59 per 100 person-years. The cumulative 1-, 3-, and 5-year survival rates were 52.1%, 21.7%, and 11.2%, respectively. The median duration of survival after resection was 12.4 months. Univariate analysis revealed stage of disease, lymph node metastasis, histological type, histological grade and macroscopic classification to be statistically significant (p-value<0.05) prognostic factors. In the multivariate analysis, only macroscopic classification was statistically significant (p-value<0.05). In conclusion, macroscopic classification was the only independent factor found to be significantly associated with survival following surgical treatment of ICC.