• 한국신뢰성학회지:신뢰성응용연구
• /
• 제16권4호
• /
• pp.323-330
• /
• 2016

Purpose: Recently, propensity score matching method is used in a large number of research paper, nonetheless, there is no research using fitness test of before and after propensity score matching. Therefore, comparing fitness of before and after propensity score matching by logistic regression analysis using data from 'online survey of adolescent health' is the main significance of this research. Method: Data that has similar propensity in two groups is extracted by using propensity score matching then implement logistic regression analysis on before and after matching separately. Results: To test fitness of logistic regression analysis model, we use Model summary, -2Log Likelihood and Hosmer-Lomeshow methods. As a result, it is confirmed that the data after matching is more suitable for logistic regression analysis than data before matching. Conclusion: Therefore, better result which has appropriate fitness will be shown by using propensity score matching shows better result which has better fitness.

• 사회복지연구
• /
• 제37호
• /
• pp.171-197
• /
• 2008

본 연구는 국민기초생활보장법 하의 조건부 수급자가 자활직업훈련 프로그램의 참여로 인해 임금이 향상되었는지에 관해 실증적으로 분석 하였다. 분석 자료로는 한국노동연구원에서의 자활사업 참여자를 대상으로 한 '자활대상자 직업훈련 실태'조사와 한국보건사회연구원에서 실시하는 2004년 '자활패널'조사 자료를 결합하여 '자활직업훈련 실태'조사의 응답자를 실험그룹으로'자활패널'을 비교그룹으로 구성하여 분석하였으며, 기존 연구의 통계적 취약점을 극복하기 위해 Propensity Score Matching 방법을 이용하여 자활직업훈련의 임금효과를 분석하였다. 분석 결과 자활직업훈련 참가자들이 그와 비슷한 특성을 가진 비교그룹에 비해서 약 12만원 정도 소득이 낮음을 확인함으로써, 이러한 부정적인 효과가 나타난 원인과 의미에 대해서 논의하였다.

• Communications for Statistical Applications and Methods
• /
• 제30권1호
• /
• pp.1-19
• /
• 2023

Observational data with missing or incomplete data are common in biomedical research. Multiple imputation is an effective approach to handle missing data with the ability to decrease bias while increasing statistical power and efficiency. In recent years propensity score (PS) matching has been increasingly used in observational studies to estimate treatment effect as it can reduce confounding due to measured baseline covariates. In this paper, we describe in detail approaches to competing risk analysis in the setting of incomplete observational data when using PS matching. First, we used multiple imputation to impute several missing variables simultaneously, then conducted propensity-score matching to match statin-exposed patients with those unexposed. Afterwards, we assessed the effect of statin exposure on the risk of heart failure-related hospitalizations or emergency visits by estimating both relative and absolute effects. Collectively, we provided a general methodological framework to assess treatment effect in incomplete observational data. In addition, we presented a practical approach to produce overall cumulative incidence function (CIF) based on estimates from multiple imputed and PS-matched samples.

• Physical Therapy Rehabilitation Science
• /
• 제13권1호
• /
• pp.134-142
• /
• 2024

Objective: This study aimed to compare the subjective health status, physical activity index, and perceived stress rate of obese children with those of healthy children through propensity matching analysis and to use the results to help manage obese children. Design: Descriptive correlation study. Methods: This study conducted a secondary analysis using raw data from the National Health and Nutrition Examination Survey. Propensity matching (1:2) was conducted between obese and healthy adolescents. Results: The number of days of physical activity was 1.22 days for the obese adolescent group and 1.01 days for the non-obese adolescent group, which was higher in the obese group (p=0.003). The subjective health status was 3.43 and 3.81 in the obese and non-obese adolescent groups, respectively, showing a higher difference in the normal group (p<0.001). No statistical difference was observed in the perceived stress rate (p=0.871) or strength exercises (p=0.190) between the two groups. Conclusion: This study suggests the need for development measures to effectively improve youth physical activity, muscle-strengthening programs, stress management, and subjective health status in obese and normal groups.

• Journal of the Korean Data and Information Science Society
• /
• 제27권1호
• /
• pp.91-100
• /
• 2016

성향점수 매칭이란 선택편의가 존재 할 수 있는 두 집단의 데이터를 성향 점수로 매칭하여 비슷한 성향을 갖는 데이터를 추출하는 방법이다. 본 논문은 그 중 하나인 퍼지 매칭 방법을 제시하였다. 성향 점수를 만들기 위해 통제변수를 선정하는 방법과 로지스틱 회귀분석을 통하여 성향 점수를 구하는 방법을 제시하였으며, 이 점수로 퍼지 매칭을 통해 성향이 비슷한 실험 집단과 통제 집단을 추출할 수 있었다. 본 논문에서는 허용오차 범위를 달리하여 분포도와 표준화 차이를 통해 두 집단이 동일한 집단임을 증명했으며, 허용오차 범위 점수가 작아질수록 선택되어 지는 케이스 수도 작아지는 것을 확인 할 수 있었다.

• 수산경영론집
• /
• 제53권3호
• /
• pp.85-101
• /
• 2022

This study analyzed the performance of the national fishing port development project, which lacked ex-post impact evaluation despite a lot of investment in terms of fishery income opportunities. Using micro data from the Census of Agriculture, Forestry, and Fisheries, the sales amount of fishery products and the proportion of fishery-related businesses were used as performance indicators. The fishery households in the fishing port area (treatment group) and those not in the area (control group) were classified through data pre-processing, and factors unrelated to the fishing ports were controlled using the propensity score matching difference-in-difference method. The analysis target is six fishing ports with large investment in from 2010 to 2014. As a result of the analysis, it was confirmed that the sales of fishery products increased significantly in four of the six fishing ports, and the proportion of fishery-related businesses increased in two fishing ports. The analysis method of this study can be fully utilized in the evaluation of the Fishing Community New Deal 300 Project, which is in need of performance analysis.

• 보건행정학회지
• /
• 제25권1호
• /
• pp.3-10
• /
• 2015

Background: Previous studies showed differences in healthcare utilization among insurance types. This study aimed to analyze the difference in healthcare utilization for percutaneous transluminal coronary angioplasty inpatients by insurance types after controlling factors affecting healthcare utilization using propensity score matching (PSM). Methods: The 2011 national inpatient sample based on health insurance claims data was used for analysis. PSM was used to control factors influencing healthcare utilization except insurance types. Length of stay and total charges were used as healthcare utilization variables. Patients were divided into National Health Insurance (NHI) and Medical Aid (MA) patients. Factors representing inpatients (gender, age, admission sources, and Elixhauser comorbidity index) and hospitals (number of doctors, number of beds, and location of hospitals) were used as covariates in PSM. Results: Tertiary hospitals didn't show significant difference in length of stay and total charges after PSM between two insurance types. However, MA patients showed significantly longer length of stay than that of NHI patients after PSM in general hospitals. Multivariate regression analysis provided that admission sources, Elixhauser comorbidity index, insurance types, number of doctors, and location of hospitals (province) had significant influences on the length of stay in general hospitals. Conclusion: Study results provided evidences that healthcare utilization was differed by insurance types in general hospitals. Health policy makers will need to prepare interventions to influence the healthcare utilization differences between insurance types.

• 보건행정학회지
• /
• 제26권4호
• /
• pp.343-351
• /
• 2016

Background: This study aims to analyze the effects of insurance types on the medical service uses for heart failure inpatients using propensity score matching (PSM). Methods: 2014 National inpatient sample based on health insurance claims data was used in the analysis. PSM was applied to control factors influencing the service uses except insurance types. Negative binomial regression was used after PSM to analyze factors that had influences on the service uses among inpatients. Subjects were divided by health insurance type, national health insurance (NHI) and medical aid (MA). Total charges and length of stay were used to represent the medical service uses. Covariance variables in PSM consist of sociodemographic characteristics (gender, age, Elixhauser comorbidity index) and hospital characteristics (hospital types, number of beds, location, number of doctors per 50 beds). These variables were also used as independent variables in negative binomial regression. Results: After the PSM, length of stay showed statistically significant difference on medical uses between insurance types. Negative binomial regression provided that insurance types, Elixhauser comorbidity index, and number of doctors per 50 beds were significant on the length of stay. Conclusion: This study provided that the service uses, especially length of stay, were differed by insurance types. Health policy makers will be required to prepare interventions to narrow the gap of the service uses between NHI and MA.

• Journal of Gastric Cancer
• /
• 제19권2호
• /
• pp.212-224
• /
• 2019

Purpose: Splenic hilar lymph node dissection (LND) during total gastrectomy is regarded as the standard treatment for proximal advanced gastric cancer (AGC). This study aimed to investigate whether splenic hilar LND or D2 LND is essential for proximal AGC of pT2- 4aN0M0 stage. Materials and Methods: Data of curative total gastrectomies (n=370) performed from 2000 to 2010 for proximal AGC of pT2-4aN0 stage were retrospectively reviewed. Clinicopathological characteristics and long-term outcomes were compared using propensity score matching between patients who underwent splenectomy (n=43) and those who did not (n=327) and between patients who underwent D2 LND (n=122) and those who underwent D1+ LND (n=248). Results: Tumors of larger size and a more advanced T stage and significantly lower overall and relapse-free survival (P<0.001) were observed in the splenectomy group than in the 2 spleen-preserving groups. Before propensity score matching, worse overall and relapse-free survival (P<0.001) was observed in the splenectomy group than in the non-splenectomy group. After matching, although the overall survival became similar (P=0.123), relapse-free survival was worse in the splenectomy group (P=0.021). Compared with D1+ LND, D2 LND had no positive impact on the overall (P=0.619) and relapse-free survival (P=0.112) after propensity score matching. Conclusions: Splenic hilar LND with or without splenectomy may not have an oncological benefit for patients with pathological AGC with no LN metastasis.

• 응용통계연구
• /
• 제36권5호
• /
• pp.415-445
• /
• 2023

생존 자료에서 Hade 등 (2020) 은 시간-의존 교란 변수가 환자의 처치 시점에 영향을 미칠 때, 해당 효과를 보정하여 treatment delay effect를 올바르게 추정하기 위해 성향 점수 매칭 방법을 이용하였다. 이 때, treatment delay effect란 환자가 관심 있는 지연 시점만큼 늦게 처치를 받는 경우 제 때 받는 경우에 비해 사건 발생 위험에 미치는 영향을 의미한다. 본 연구에서는 또 다른 성향 점수 기반 모형인 Cox-MSM 모형 또한 해당 효과를 올바르게 추정할 수 있는지 모의 실험을 통해 확인 및 기존 매칭 모형과 비교하였다. 모의실험 결과, 세 가지 모형 모두 다양한 시나리오 내에서 treatment delay effect를 올바르게 추정함을 확인하였다. 특히 모든 시나리오 내에서 Cox-MSM의 제곱근평균제곱오차의 값이 가장 낮았으며, restricted Cox matching 모형에서 가장 큰 값을 가지는 것으로 나타났다. 결론적으로, 성향 점수에 기반하나 매칭이 아닌 방법 또한 treatment delay effect 적용이 가능하다는 결과를 제공한다. 추후 G-formula과 같이 성향 점수 기반이 아닌 모형에서도 적용이 가능한지에 대한 상세 연구가 필요하다고 사료된다.