• Communications for Statistical Applications and Methods
• /
• v.30 no.4
• /
• pp.389-402
• /
• 2023

Multivariate or clustered failure time data often occur in many medical, epidemiological, and socio-economic studies when survival data are collected from several research centers. If the data are periodically observed as in a longitudinal study, survival times are often subject to various types of interval-censoring, creating multivariate interval-censored data. Then, the event times of interest may be correlated among individuals who come from the same cluster. In this article, we propose a unified linear regression method for analyzing multivariate interval-censored data. We consider a semiparametric multivariate accelerated failure time model as a statistical analysis tool and develop a generalized Buckley-James method to make inferences by imputing interval-censored observations with their conditional mean values. Since the study population consists of several heterogeneous clusters, where the subjects in the same cluster may be related, we propose a generalized estimating equations approach to accommodate potential dependence in clusters. Our simulation results confirm that the proposed estimator is robust to misspecification of working covariance matrix and statistical efficiency can increase when the working covariance structure is close to the truth. The proposed method is applied to the dataset from a diabetic retinopathy study.

• The Korean Journal of Applied Statistics
• /
• v.33 no.5
• /
• pp.627-637
• /
• 2020

In this paper, the 1.24 million elderly patient medical data (HIRA-APS-2014-0053) provided by the Health Insurance Review and Assessment Service and weather data are analyzed with generalized estimating equation (GEE) model and long short term memory (LSTM) based recurrent neural network (RNN) model to predict the number of disease occurrence. To this end, we estimate the patient's residence as the area of the served medical institution, and the local weather data and medical data were merged. The status of disease occurrence is divided into three categories(occurrence of disease of interest, occurrence of other disease, no occurrence) during a week. The probabilities of categories are estimated by the GEE model and the RNN model. The number of cases of categories are predicted by adding the probabilities of categories. The comparison result shows that predictions of RNN model are more accurate than that of GEE model.

• Journal of Korea Water Resources Association
• /
• v.49 no.8
• /
• pp.645-656
• /
• 2016

The study performed a regional flood frequency analysis and proposed a regression equation to estimate design floods corresponding to return periods for ungauged basins in Geum-river basin. Five preliminary tests were employed to investigate hydrological independence and homogeneity of streamflow data, i.e. the lag-one autocorrelation test, time homogeneity test, Grubbs-Beck outlier test, discordancy measure test ($D_i$), and regional homogeneity measure (H). The test results showed that streamflow data were time-independent, discordant and homogeneous within the basin. Using five probability distributions (generalized extreme value (GEV), three-parameter log-normal (LN-III), Pearson type 3 (P-III), generalized logistic (GLO), generalized Pareto (GPA)), comparative regional flood frequency analyses were carried out for the region. Based on the L-moment ratio diagram, average weighted distance (AWD) and goodness-of-fit statistics ($Z^{DIST}$), the GLO distribution was selected as the best fit model for Geum-river basin. Using the GLO, a regression equation was developed for estimating regional design floods, and validated by comparing the estimated and observed streamflows at the Ganggyeong station.

• Korean Journal of Social Welfare Studies
• /
• v.48 no.4
• /
• pp.33-57
• /
• 2017

As there is a growing concern about the steady increase in the consumer debt and its potential consequences on individuals and families, this study examined the association between personal debt and the risk of domestic violence, which in this study is referred to as violence between man and woman who have a spousal relationship. We used the data from the Korea Welfare Panel Study collected from 2009 to 2016. We applied a generalized estimating equation approach for the analysis of panel data. The results show that the higher the ratio of personal debt to disposable income and the ratio of debt payment to disposal income is, the greater the risk of domestic violence. While the debt to income ratio played a role regarding was related to a heightened risk of domestic violence among the poor group, the debt payment to income ratio was associated with a higher risk of domestic violence among the non-poor group. Implications of the study were discussed.

• Health Policy and Management
• /
• v.32 no.1
• /
• pp.53-62
• /
• 2022

Background: This study was designed to examine regional proportions for people who experienced unmet health care needs due to reduced mobility or unhealthiness and factors associated with experience of unmet health care needs by them. Methods: A total of 11,620 people were retrieved from the Korea Health Panel data (2014-2018). Regional proportions for people who experienced unmet health care needs due to reduced mobility or unhealthiness were estimated using cross-sectional weights and the factors associated with them were analyzed using generalized estimating equation. Results: The number of people who experienced unmet health care needs due to reduced mobility or unhealthiness was estimated as 278,083 in 2018. Women, the aged (65+), below elementary school, single as marital status, low income, bad self-rated health, people with disabilities, and long-term insurance beneficiaries were statistically significantly associated with experience of unmet health care needs due to reduced mobility or unhealthiness. Conclusion: Given high and dispersed demand for visiting health care, government need to expand the infrastructure and finance to facilitate visiting health care.

• Management Science and Financial Engineering
• /
• v.12 no.1
• /
• pp.65-77
• /
• 2006

This study proposes a Bayesian dynamic model in a hierarchical way to assess the time-varying effect of risk factors on the likelihood of corporate bankruptcy. For the longitudinal data, we aim to describe dynamically evolving effects of covariates more articulately compared to the Generalized Estimating Equation approach. In the analysis, it is shown that the proposed model outperforms in terms of sensitivity and specificity. Besides, the usefulness of this study can be found from the flexibility in describing the dependence structure among time specific parameters and suitability for assessing the time effect of risk factors.

• Communications for Statistical Applications and Methods
• /
• v.23 no.6
• /
• pp.555-562
• /
• 2016

Interval censored data often occur in an observational study where the subject is followed periodically. Instead of observing an exact failure time, two inspection times that include it are available. There are several methods to analyze interval censored failure time data (Sun, 2006). However, in the presence of competing risks, few methods have been suggested to estimate covariate effect on interval censored competing risk data. A sub-distribution hazard model is a commonly used regression model because it has one-to-one correspondence with a cumulative incidence function. Alternatively, Klein and Andersen (2005) proposed a pseudo-value approach that directly uses the cumulative incidence function. In this paper, we consider an extension of the pseudo-value approach into the interval censored data to estimate regression coefficients. The pseudo-values generated from the estimated cumulative incidence function then become response variables in a generalized estimating equation. Simulation studies show that the suggested method performs well in several situations and an HIV-AIDS cohort study is analyzed as a real data example.

• Journal of Nutrition and Health
• /
• v.28 no.1
• /
• pp.71-83
• /
• 1995

The purpose of the study was to develop generalized equations for estimating stature and weight for the nonambulatory elderly persons. Height weight recumbent knee height total ann length, midarm, waist and calf circumferences, triceps and subscapular skinfolds were measured from over 60 years old 315 ambulatory elderly. The equations to predict stature and weight were derived from participants in the validation sample and were applied to the participants in the cross-validation to test the accuracy and validity of equations. Stature and weight were significantly and negatively associated with age of women and similar patterns observed in men but associated to a slight degree. Knee height and total arm length were highly correlated with stature but the majority of the variances in stature was accounted for by knee height for both the men and women. In men, waist circumference was the most significantly correlated with weight and am, calf circumferences and so forth. But in women arm circumference was the highest then waist and calf circumference in order. The possible predictor variables to estimate of stature were knee height total arm length and age for both elderly men and women. Predictor variables to estimate of weight were recumbent measures of waist am, calf circumferences and knee height for both sexes. Inclusion of skinfold thickness measurements did not improve the prediction power of estimation for weight. When both equations developed from the present study and Chumlea's study were applied to cross-valida-tions samples, the equations derived from present study showed better accuracy and validity. The presentation of prediction equations using two, three, or four recommended measurements allows the selection of an equation based upon the measurements that are possible to collect on an individual basis.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.6
• /
• pp.4067-4076
• /
• 2015

This study aims to examine the trend in household catastrophic medical expenditure according to the healthcare coverage types and its associated factors based on the raw data of the Korean Health Panel over the years 2008 to 2011. Correspondence analysis was used to investigate the trend in the incidence rates of annual catastrophic medical expenditure and generalized estimating equation to examine the factors influencing the incidence of catastrophic medical expenditure. The annual mean incidence rates of household catastrophic medical expenditure were 25.1%, 15.4%, 10.1%, 5.4% and 3.2% in the threshold levels of 10%, 15%, 20%, 30%, and 40% respectively. The incidence rate of household catastrophic medical expenditure was higher when the total annual household income was lower, the education level of the householder was lower, the healthcare coverage type was National Health Insurance, the householder had disability, the age of the householder was older, the number of household members was smaller, the subjective health status of household members was lower, and the prevalence rate of the chronic disease of the household was higher(p<0.05). Therefore, a policy for vulnerable households with older or patient members of chronic diseases should be established.

• Health Policy and Management
• /
• v.31 no.1
• /
• pp.125-139
• /
• 2021

Background: In Korea, the health gap widens due to the number of medical resources and access to medical services between metropolitan and rural. The purpose of this study is to identify the impact of residential migration on medical utilization and accessibility. Methods: This study extracted 528,516 claimed cases in the National Health Insurance Service-Cohort Sample Database from 2006 to 2015. Subjects were classified into two groups by the magnitude of the region, the metropolitan and the rural. The inversed probability weights were calculated for each group. And coefficients of the two-part model were estimated by generalized estimation equation. Results: Those who moved region from metropolitan to rural tend to increase the length of stay and inpatients with ambulatory care sensitive conditions (ACSC) disease. Contrariwise, those who moved areas from rural to metropolitan tend to decrease the total medical cost, the adjusted patient days, the number of outpatients and the number of outpatients and inpatients with ACSC disease. Conclusion: This study identified that between the residents who continued to reside in the region and the migrants, there were significant differences in the medical accessibility, quality of primary care, and unmet medical need.