Jinheum Kim

Cost-effective mammography screening in Korea: high incidence of breast cancer in young women.

The epidemiological characteristics of breast cancer in Korean women are different from the characteristics reported in Western women. The highest incidence rate occurs in Korean women in their 40s. The purpose of this study was to determine the most cost‐effective screening interval and target age range for Korean women from the perspective of the national healthcare system. A stochastic model was used to simulate breast cancer screenings by varying both the screening intervals and the age ranges. The effectiveness of mammography screening was defined as the probability of detecting breast cancer in the preclinical state and the cost was based on the direct cost of mammography screening and the confirmative tests. The age‐specific mean sojourn times and the sensitivity of the mammography were applied in the stochastic model. An optimal cost‐effectiveness was determined by the incremental cost‐effectiveness ratio and lifetime schedule sensitivity. Sensitivity analyses were undertaken to assess parameter uncertainty. The selected cost‐effective strategies were: (1) the current biennial mammography screenings for women who are at least 40 years old; (2) biennial screening for women between the ages of 35 and 75 years; and (3) a combination strategy consisting of biennial screening for women aged between 45 and 54 years, and 3‐year interval screening for women aged between 40 and 44 years and 55 and 65 years. Further studies should follow to investigate the effectiveness of mammography screening in women younger than 40 years in Asia as well as in Korea. (Cancer Sci 2009; 100: 1105–1111)

A comparative study on gene-set analysis methods for assessing differential expression associated with the survival phenotype

BackgroundMany gene-set analysis methods have been previously proposed and compared through simulation studies and analysis of real datasets for binary phenotypes. We focused on the survival phenotype and compared the performances of Gene Set Enrichment Analysis (GSEA), Global Test (GT), Wald-type Test (WT) and Global Boost Test (GBST) methods in a simulation study and on two ovarian cancer data sets. We considered two versions of GSEA by allowing different weights: GSEA1 uses equal weights, yielding results similar to the Kolmogorov-Smirnov test; while GSEA2s weights are based on the correlation between genes and the phenotype.ResultsWe compared GSEA1, GSEA2, GT, WT and GBST in a simulation study with various settings for the correlation structure of the genes and the association parameter between the survival outcome and the genes. Simulation results indicated that GT, WT and GBST consistently have higher power than GSEA1 and GSEA2 across all scenarios. However, the power of the five tests depends on the combination of correlation structure and association parameter. For the ovarian cancer data set, using the FDR threshold of q < 0.1, the GT, WT and GBST detected 12, 6 and 8 significant pathways, respectively, whereas neither GSEA1 nor GSEA2 detected any significant pathways. In addition, among the pathways found significant by GT, WT, and GBST, three pathways - Purine metabolism, Leukocyte transendothelial migration and Jak-STAT signaling pathway - overlapped with those reported in previous ovarian cancer microarray studies.ConclusionSimulation studies and a real data example indicate that GT, WT and GBST tend to have high power, whereas GSEA1 and GSEA2 have lower power. We also found that the power of the five tests is much higher when genes are correlated than when genes are independent, when survival is positively associated with genes. It seems that there is a synergistic effect in detecting significant gene sets when significant genes have within-class correlation and the association between survival and genes is positive or negative (i.e., one-direction correlation).

Scheduling mammography screening for the early detection of breast cancer in Korean women.

Objective: To propose an efficient screening schedule for breast cancer among Korean women using the stochastic model in which the age-specific incidence rate was considered. Setting: Female breast cancer data in the Korea Central Cancer Registry 2002. Methods: The stochastic model was based on the threshold method, in which the schedule is determined by a pre-specified threshold value. The threshold value was defined as the probability of being in a preclinical state of breast cancer at age 40 years. The sensitivity of the mammography was specified as 0.7. Two models for mean sojourn time (MST) in the preclinical state were considered; MSTs for Model I were 2 (ages < 50 years), 3 (ages 50–59 years), and 4 years (ages ≥ 60 years), and MSTs for Model II were 3, 4, and 5 years for the corresponding age groups. Results: The threshold method for Model I generated 19 examinations within the screening ages of 40–69 years. Each screening time was determined at ages 40.0, 41.6, 43.2, 44.8, 46.0, 47.2, 48.4, 49.6, 50.7, 51.7, 52.7, 53.7, 54.7, 56.2, 57.8, 59.4, 61.3, 63.1, and 64.9 years. The schedule sensitivity of Model I was 64.2%, which was higher than that (57.5%) of the biennial periodic schedule. Model II included 11 screenings between the ages of 40 and 69 years and also showed a higher schedule sensitivity, especially for women aged 40 years as compared with the biennial screening. Conclusions: This finding suggests that the threshold screening schedule for breast cancer increase the schedule sensitivity by reflecting the age-specific incidence rate of a population.

Goodness-of-fit tests for the additive risk model with (p > 2)-dimensional time-invariant covariates.

This paper presents methods for checking the goodness-of-fit of the additive risk model with p(> 2)-dimensional time-invariant covariates. The procedures are an extension of Kim and Lee (1996) who developed a test to assess the additive risk assumption for two-sample censored data. We apply the proposed tests to survival data from South Wales nikel refinery workers. Simulation studies are carried out to investigate the performance of the proposed tests for practical sample sizes.

Logrank-type tests for comparing survival curves with interval-censored data

We propose logrank-type tests for comparing several survival functions from interval-censored data. The proposed tests do not require use of the so-called EM algorithm because we introduce uniform weights that depend only on the size of the risk set at each observed time instead of the weights involving estimated survivals. This technique reduces computation time. As alternatives for the estimated asymptotic variance of a proposed test statistic, we introduce estimates that mimic the logrank test and the multiple imputation method. Results from simulation studies show that our proposed tests are very satisfactory in terms of size and powers. We illustrate the proposed tests with breast cosmesis data from Finkelstein and Wolfe [1985. Biometrics 41, 933-945] and lung cancer post-operative data from the Yonsei Cancer Center in Korea.

Median Regression Model with Interval Censored Data

Quantile regression methods have been used to estimate upper and lower quantile reference curves as the function of several covariates. Especially, in survival analysis, median regression models to the right-censored data are suggested with several assumptions. In this article, we consider a median regression model for interval-censored data and construct an estimating equation based on weights derived from interval-censored data. In a simulation study, the performances of the proposed method are evaluated for both symmetric and right-skewed distributed failure times. A well-known breast cancer data are analyzed to illustrate the proposed method.

A review on mathematical models for estimating indoor radon concentrations.

Radiation from natural sources is one of causes of the environmental diseases. Radon is the leading environmental cause of lung cancer next to smoking. To investigate the relationship between indoor radon concentrations and lung cancer, researchers must be able to estimate an individual’s cumulative level of indoor radon exposure and to do so, one must first be able to assess indoor radon concentrations. In this article, we outline factors affecting indoor radon concentrations and review related mathematical models based on the mass balance equation and the differential equations. Furthermore, we suggest the necessities of applying time-dependent functions for indoor radon concentrations and developing stochastic models.

Modeling Clustered Interval-Censored Failure Time Data with Informative Cluster Size

We propose two estimating procedures to analyze clustered interval-censored data with an informative cluster size based on a marginal model and investigate their asymptotic properties. One is an extension of Cong et al. (2007) to interval-censored data and the other uses the within-cluster resampling method proposed by Hoffman et al. (2001). Simulation results imply that the proposed estimators have a better performance in terms of bias and coverage rate of true value than an estimator with no adjustment of informative cluster size when the cluster size is related with survival time. Finally, they are applied to lymphatic filariasis data adopted from Williamson et al. (2008).

Analyzing Clustered and Interval-Censored Data based on the Semiparametric Frailty Model

We propose a semi-parametric model to analyze clustered and interval-censored data; in addition, we plugged-in a gamma frailty to the model to measure the association of members within the same cluster. We propose an estimation procedure based on EM algorithm. Simulation results showed that our estimation procedure may result in unbiased estimates. The standard error is smaller than expected and provides conservative results to estimate the coverage rate; however, this trend gradually disappeared as the number of members in the same cluster increased. In addition, our proposed method was illustrated with data taken from diabetic retinopathy studies to evaluate the effectiveness of laser photocoagulation in delaying or preventing the onset of blindness in individuals with diabetic retinopathy.

Haplotype-based score test for linkage in nuclear families

BackgroundTo look for genetic linkage between angiotensin-I converting enzyme(ACE) gene and hypertension in a Korean adolescent cohort, we developed a powerful test using the covariances between marginal differences and their variances in a transmission/non-transmission table.ResultsWe estimated haplotype frequencies using the parental and affected offsprings genotypes and then constructed a transmission/non-transmission table for the parental haplotypes transmitted to the offspring. We then proposed a test for checking the marginal homogeneity in the table. Because the cells in the table were dependent due to the uncertainty of the parental haplotypes, we adopted a randomization procedure to estimate the significance of the observed test statistic. Simulations show that our test performs well on a nominal level and has a monotone power, which increases as the relative risk increases. With our test, there was no evidence of genetic linkage between the ACE gene and hypertension in the Korean adolescent cohort.ConclusionWe developed a score test for linkage and used simulations to demonstrate that our test performs well at a nominal level. Under some situations where the diversity of haplotypes is low, the proposed test gained a little power over the method based on only variances between marginal differences in a transmission/non-transmission table.