The Survival Time of Terminal Cancer Patients : Prediction Based on Clinical Parameters and Simple Prognostic Scores = 말기 암환자의 임상변수 및 단일여명예측지수를 통한 생존기간 예측

김아솔 경북대학교 대학원 2013 국내석사

연구배경: 본 연구는 호스피스 치료를 받는 말기 암환자의 생존기간과 관련 있는 임상변수를 알아보고, 현재 사용되고 있는 단일여명예측지수를 통한 생존분석을 통해 말기 암환자의 치료에 도움이 되고자 하였다. 방법: 2009년 3월부터 2011년 8월까지 경북대학교병원 및 칠곡경북대학교병원 호스피스병동에 입원 중 원내 사망한 환자 415명을 대상으로 입원당시 임상변수 28가지 및 완화의료수행 지수, 완화의료예후지수 및 완화의료예후점수를 각각 조사하였다. 각 임상변수 및 예후지수별 생존기간 차이 분석 및 Cox의 비례위험모형을 이용한 상대위험도를 산출하였다. 단일여명예측지수들에 대해서는 Receiver operating characteristic 곡선 분석을 시행하였다. 결과: 조절되지 않는 통증, 호흡곤란, 오심, 부종, 수행능력저하, 빈혈, 백혈구증가증, 중성구증가증, 저알부민혈증, 고빌리루빈혈증, 간효소치증가, 프로트롬빈시간 연장, 활성화트롬보플라스틴시간 연장, 저나트륨혈증, C-반응성 단백 증가 등을 보이는 환자는 통계학적으로 유의하게 생존기간이 짧았다. Cox 회귀 분석에서 수행능력저하가 생존율에 영향을 주는 가장 주된 변수로 나타났다(odds ratio=4.242, P<0.001). 단일여명예측지수의 4주 생존예측에 대한 최적기준점은 완화의료예후지수는 4.5점 초과 (민감도 66.67%, 특이도 58.17%), 완화의료예후점수는 10점 초과 (민감도 66.18%, 특이도 82.69%), 완화의료수행지수는 30점 이하 (민감도 62.80%, 특이도 78.85%)로 나타났다. 결론: 수행능력저하가 생존율에 영향을 주는 가장 주된 변수였으며, 단일여명예측지수 중 완화의료예후점수가 4주 생존예측에 가장 유용한 도구였고 적절한 기준은 10점 초과로 나타났다. Background: This study aims to investigate the clinical parameters related to the survival time of terminal cancer patients to help a treatment through analysis of survival rates based on simple prognostic scores that are currently used. Method: This research was carried out with 28 clinical parameters, palliative performance scale, palliative prognostic index and palliative prognostic score of 415 inpatients who died in hospice at two hospitals from March of 2009 to August of 2011. The dissimilarities in the survival time were analyzed and relative risks using Cox Regression Modeling were determined. A Receiver Operating Characteristic Analysis was carried out to determine simple prognostic scores. Result: The patients with pain, dyspnea, nausea, edema, debilitation of performance status, anemia, leukocytosis, neutrophilia, hypoalbuminemia, hyperbilirubinemia, elevated liver enzyme levels, extended prothrombin time, extended activated partial thromboplastin time, hyponatremia, and increased levels of C-reactive protein all showed similarly short survival times and debilitation of performance status was the major factor affecting survival.(odds ratio=4.242, P<0.001) The optimal points for 4 week-survival expectation palliative prognostic index exceeded 4.5 points(sensitivity: 66.67%, specificity: 58.17%). This point for the palliative prognostic score exceeded 10 points(sensitivity: 66.18%, specificity: 82.69%), while the score for the palliative performance scale was below 30 points(sensitivity: 62.80%, specificity: 78.85%). Conclusion: Debilitation of performance status is a major factor affecting survival, and the palliative prognostic score is the most useful factor to predict 4 week-survival with the optimal level exceeding 10 points.

Random survival forest with intermediate event using multi-state modeling

노윤호 Graduate School, Yonsei University 2020 국내박사

Random forest는 다수의 의사결정나무에서 각 나무의 가지 선택 단계에 무작위성을 부여함으로써 예측력을 높이는 앙상블 학습 방법 중 하나이다. Random survival forest는 이를 우측 중도 절단된 생존 자료에 적합하도록 확장한 모형이며 특별한 분포나 통계학적 가정이 전제되지 않아도 분석이 가능하다는 장점을 가지고 있다. 임상분야에서 특정 사건에 대한 생존함수 추정 시 관찰기간동안 중간사건을 경험하는 경우가 종종 있다. 중간사건이란 관찰기간동안 장기이식, 다른 질병 발생 등 우리가 관심있는 사건에 대한 생존함수의 변화를 유발할 수 있는 사건을 말한다. 중간 사건을 고려하지 않을 경우 편의가 발생할 수 있으므로 이러한 편의를 고려하여 분석하는 것이 적절한 분석방법이다. 본 연구에서는 생존 자료에서 중간 사건 발생 시 random survival forest를 다상태 모형으로 적용하여 분석하는 방법을 제시하고자 한다. 기존의 우측중도절단에 적합하도록 고안된 Random survival forest는 중간 사건을 모형에 고려하지 않음으로 인한 편의가 존재한다. 중간사건을 다상태 모형의 형태로 적용함으로써 예측력 향상을 기대할 수 있으며 변수선택 과정에서 중간사건을 고려하지 않을 경우 중간사건에 영향을 미치는 변수를 놓지는 문제가 있어 이를 고려한 분석과의 비교를 통해 세밀한 해석을 가능하게 한다. 다상태 모형에 적용하기 위하여 중간 사건이 발생할 경우와 발생하지 않을 경우의 두 가지 상황에서 각각의 random survival forest를 모델링한다. 먼저, 중간 사건이 발생할 경우 중간사건발생 시점으로부터 좌측 절삭된 자료에서의 random survival forest를, 그리고 중간 사건이 발생하지 않는 경우 전체 자료에서 중간 사건 시점으로부터 우측중 도절단하여 분석한다. 각각의 상황에 따라 두 가지 모형이 나오게 되며 이를 토대로 통합된 예측 결과를 제시한다. 모의실험에 사용될 예측력 평가 척도로는 integrated AUC를 이용하였고 변수선택 평가 척도로는 Variable importance를 이용하였다. 다상태 모형을 적용한 random survival forest 에서는 중간 사건 발생 여부에 따른 두 가지 결과가 나오게 되므로 두 결과를 하나로 해석할 수 있도록 가중치로 통합된 평가지표를 제시하였다. 평가 모형은 중간 사건을 고려하지 않은 Random survival forest와 제안된 다상태모형을 적용한 random survival forest 이다. 모의실험 결과 중간사건을 고려하지 않은 모형에 비해 중간사건을 고려한 random survival forest 모형에서 변수선택 능력이 월등함을 확인할 수 있었다. 또한 예측력 평가 측면에서도 기존 중간모형을 고려하지 않은 모형에 비해 좋은 예측력을 나타내는 것으로 확인되었다. 임상에서 중간사건 발생시 이를 고려하지 않고 분석을 할 경우 중요한 변수를 놓질수 있다는 점에서 제안된 모형의 중요성을 확인할 수 있다. 또한 제안된 모형을 통해서 제한된 해석만이 가능한 random survival forest의 단점을 보완하며 보다 정교한 해석을 할 수 있는 점에서 큰 의의가 있다. Background: Random forest, developed by Breiman (2001), is an ensemble method that combines several classification trees using random variable selection for each node. Random survival forest is a special case of the random forest method for right-censored data and is free from the assumption of proportionality. Many situations exist where subjects may experience an intermediate event (e.g., transplantation or a different disease). These events have potential bias; therefore, it is appropriate to avoid this bias using the proper method. In this study, we propose a random survival forest method using multistate modeling when an intermediate event occurs. Methods: To apply the random survival forest method to multistate modeling, we develop two models. We apply the left-truncated random survival forest method for the subject with an intermediate event and the ordinary random survival forest method, modifying the intermediate time to right censoring for the subject without an intermediate event. Because we exhibit two models, we integrate the prediction. The integrated variable of importance (VIMP) is used to interpret the variable selection. Conclusions: The simulation results indicate that the discriminative ability performance in variable selection, which is evaluated using the area under the curve (AUC) of the VIMP (AUCVIMP), outperforms the original random forest model. Moreover, the prediction performance for the integrated AUC is more accurate.

(The) Effect of Entry Timing on Foreign Subsidiary Survival and Growth : An Empirical Study of Korean Firms in China

노운 전남대학교 대학원 2008 국내석사

Although there has been much earlier research conducted on entry timing and performance, the studies have had their limitations. Firstly, most of the earlier studies used measures such as market share or subsidiary profitability, but many neglected to control for a survival bias. In this paper we check the effect of entry timing on survival and subsidiary growth. Second, most of the earlier empirical studies on survival were based on American and western European firms. Because these countries are very developed and have well established industries structure and business environments, the results do not necessarily apply to Asian countries that have unique cultures, business environments and developing economies. Since the previous studies findings are limited in their scope, empirical research on survival in Asia is required. In addition to the shortcomings already mentioned, several of the earlier studies used technology-based asset advantage as a moderator in the relationship between entry timing and performance, but these studies ignored other factors that could play an important role. Firm-level variables can provide important strategic resources in foreign market entry, and help the firm in risk reduction. This paper tries to study the effect of entry timing on subsidiary survival and growth with firm-level advantages as moderators. Using a sample of 173 Korean FDI firms in China, the paper uses two regression models to develop and test hypotheses about firm-level advantage moderators of entry timing's influence on a subsidiary's growth and survival. The results show that early entry brings a higher long-term growth but has a lower survival likelihood than late entry. The magnitude of these effects depends on the type of firm-level advantages of a foreign investing firm possesses. First, when we use technology-based asset advantage as the moderator to check the effect of entry timing on survival and growth, the results show that it has positive influence on growth and survival, and early entrants with high technology-based asset advantage would be survived and get high sale growth in foreign entry. Second, if we use marketing intensity as the moderator, the results show that it can not ensure early entrants to get the high growth in local market, but very high marketing intensity may cause failure when a firm goes abroad to expand new market. Third, we use firm size as the moderator, the results show that firm size has positive relationship with the subsidiary survival. But it is not clear for the relationship between firm size and subsidiary growth. 지금까지 학자들은 해외자회사의 성과에 대한 연구를 많이 해 왔다. 하지만 선행연구를 살펴보면 아래와 같은 한계점이 있다. 첫째, 많은 선행연구에서 성과에 대한 측정할 때 항상 시장 점유율, 자회사의 이익소득 과 매출액 생장 3가지 변수를 이용해서 분석했고 자회사 생존 여부에 관한 문제 무시했다. 사실은 해외자회사 생존해야 그의 생장에 대한 연구만 이익적일 수 있다. 그래서 본 연구에서 생존과 생장 2가지 변수 분석한다. 둘째, 선행연구는 주로 미국, 유럽 등 선진국 투자국 아니면 피 투자국 대상으로 연구했다. 선진국 같은 경우는 완벽한 사회, 문화 와 경제체계를 제정했기 때문에 북미와 유럽을 바탕으로 실증연구 결과는 전 세계 적으로 적용할 수 없다. 셋째, 기술우위 조절변수로써 진입시점은 자회사의 성과에 미치는 영향에 대한 연구 많이 했지만 다른 중요한 변수 조절변수로써 연구하는 것은 많이 않다. 그래서 본 연구는 회사적 변수 조절변수로써 한국기업 대중국투자의 진입시점은 자회사의 생존과 생장에 미치는 영향을 연구한다. 173개 중국에 투자한 한국기업을 표본으로 의해서 2가지 회귀분석 방법을 이용해서 분석한다. 이런 분석을 통해서 얻은 결과는 선발진입자가 자회사 장기적인 생장을 유지할 수 있지만 생존가능성 후발진입자보다 낮다. 이런 관계가 조절변수에 따른 다를 수 있다.

Strategies to Enhance Survival of Out-of-hospital Cardiac Arrest : From Individual to Systems-of-care perspectives

Jooyeong Kim 서울대학교 대학원 2022 국내박사

Background Out-of-hospital cardiac arrest (OHCA) is a sudden, unexpected happening with high incidence and poor neurologic intact survival, which makes it a major public health challenge to overcome. Efforts in various fields, including public health, emergency medicine and in-hospital critical care are essential for decreasing burden of OHCA. The first goal of this study is to estimate the effect of hypertension, which is the most prevalent cardiovascular risk factor, and treatment on incidence of emergency medical service (EMS)-assessed OHCA in Korea. The second goal of this study is to estimate the effect of mechanical chest compression device during transportation of patients with OHCA according to patient transport interval. Although mechanical chest compression device itself did not show significant survival benefit compared with manual compression, in an EMS system where all EMS-treated OHCA patients are transported to hospitals, CPR quality at moving ambulance may affect the survival. The third goal of this study was to evaluate the outcome of patients with OHCA who are transported to cardiac resuscitation centers (CRC) compared with non-tertiary centers. Method For the first study goal, case-control analysis was conducted using Korean Cardiac Arrest Research Consortium (KoCARC) database, a multicenter OHCA registry database as a case and community-based controls using Community Health Survey (CHS) conducted by the Korean Centers for Disease Control and Prevention (KCDC) as a control. Cases were defined as ES-treated adult OHCA patients presumed to have cardiac etiology from 2015 to 2017. Controls were matched at a 1:2 ratios with strata, including age, gender and county of residence. Multivariable conditional logistic regression was conducted after adjusting for diabetes, smoking and alcohol consumption habit. For the second study goal, KoCARC database from year 2015 to 2020 was used for the analysis. KoCARC database includes information such as patient demographics, prehospital and hospital factors based on Utstein criteria, and survival outcomes. Primary outcome was survival to admission. Multivariable logistic regression was conducted to analyze the effect of mechanical chest compression device (MCPRD) compared with that of manual CPR and calculated adjusted odds ratios (AORs) with 95% confidence intervals (CIs) after adjusting for potential confounders: age, sex, medical histories, initial electrocardiography, witness status, provision of bystander CPR, place pf arrest, provision of advanced airways, EMS response time interval and scene resuscitation time interval. The interaction effect of MCPRD according to patient transport interval (PTI) was evaluated. PTI was categorized into three groups: short PTI for 0–5 minutes, intermediate for 6–10 minutes, and long PTI for 11 min and longer. To control for selection bias and confounding factors, two methods of propensity score analysis was performed. For the third study gold, Hospital Assessment Survey and national OHCA databased was used. Hospital Assessment Survey was conducted through Gyeonggi Emergency Medical Support Center to all OHCA receiving EDs in Gyeonggi province at 2015. The data of prehospital and hospital factors of OHCA patients transported to hospitals in Gyeonggi province at 2012-2014 was collected using the national OHCA database constructed by KCDC. The hospitals were classified as definite cardiac resuscitation center (D-CRC) if they had a standard resuscitation protocol at the ED; had the device, manpower, and standard protocol for TTM; if PCI was available 24/7; if they had an OHCA registry system; if they could confirm brain death, had an available rehabilitation program for post-resuscitated patients, and provided community-based CPR programs. Otherwise, hospitals were classified as primary cardiac resuscitation center (P-CRC). The primary outcome was favorable neurologic outcome defined as cerebral performance category 1 or 2. Multivariable logistic regression was conducted to calculate adjusted odds ratios (AORs) with 95% confidence intervals (CIs) of D-CRCs and P-CRCs after adjusting for potential confounders. Results From the first study goal, a total 2,633 cases of OHCA patients and 5,266 cases of community controls were matched for the analysis after excluding pediatric patients (N=177) and patients with unknown information of hypertension (N=504). After adjusting for diabetes, smoking and alcohol consumption habits, hypertension was significantly associated with OHCA risk; the AOR (95% CI) for OHCA was 1.19 (1.07-1.32). In terms of hypertension treatment, the without hypertension treatment group was significantly associated with increased OHCA risk (3.41 [2.72-4.24]), whereas the hypertension treatment group was not significantly associated with OHCA risk (0.96 [0.86-1.08]) compared with the without hypertension group. For the second study goal, among 9,861 patients with OHCA after excluding pediatric patients (N=232), those who achieved prehospital return of spontaneous circulation (N=1,281), those without information of MCPRD or PTI (N=4,818), total 3,530 patients were eligible for the analysis. The AOR for survival to admission and hospital ROSC was significant compared to manual CPR (AOR 1.36, 95% CI 1.07-1.77 and AOR 1.47, 95% CI 1.20-1.79). Interaction effect of long PTI on the effect of MCPRD was significant for both survival to admission (AOR 1.49, 95% CI 1.05-2.12) and hospital ROSC (AOR 1.70, 95% CI 1.26-2,29). Optimal and Greedy PS matched cohorts were extracted. From both propensity-matched cohort, interaction effect of long PTI on the effect of MCPR was both significant; AOR 1.60 (95% CI 1.06-2.24) and AOR 1.60 (95% CI 1.06-2.42), respectively. In third study, among the 90 cardiac receiving hospitals in Gyeonggi province, total 77 hospitals responded to the survey. According to the result of the survey, 15 hospitals were categorized as D-CRCs. Among 9,912 OHCA patients who were transported to cardiac arrest receiving hospitals, 4,036 patients (40.7%) were initially transported D-CRCs. After adjusting to potential confounders, patients transported to D-CRCs were significantly associated with good neurologic outcome (AOR 2.10, 95% CI 1.51-2.93) compared to those transported to P-CRCs. For patients who survived to admission, D-CRC was significantly associated with better neurologic outcome (AOR 1.48, 95% CI 1.02-2.14). Conclusion In this multicenter population-based OHCA study, individual, prehospital and hospital strategies to enhance survival were evaluated. HTN was an independent risk factor for OHCA presumed of cardiac etiology. The most high-risk population was the HTN diagnosed but without treatment group, and the magnitude of risk was observed to be reduced to the non-HTN-diagnosed population level when anti-hypertensive treatment was taken. Individualized and appropriate risk control management should be emphasized to reduce the burden of cardiovascular complications by HTN In EMS perspective, the use of mechanical chest compression device showed a significant survival benefit when patient transportation interval is prolonged. In systems-of-care perspective, regional CRC system significantly affected the neurologic outcome of patients with OHCA. Establishing an appropriate, optimal patient transport protocol for improving outcome considering EMS performance, prehospital special situations and regional hospital resources is necessary. 배경 병원 밖 심정지는 갑작스럽게 심장박동이 중지되는 상태로 발생률이 높고 사망률이 높아 전세계적인 보건학적 문제로 대두되고 있다. 병원 밖 심정지의 질병 부담을 줄이기 위해서는 공중보건학, 응급의학 그리고 중환자 의학 등 여러 다양한 분야에서의 노력이 필수적이다. 이 연구의 목적은 1) 가장 흔히 발생하는 심혈관질환 위험인자인 고혈압과 그것의 치료가 병원 밖 심정지 발생에 미치는 영향을 확인하고, 2) 병원 밖 심정지 환자에게 적용된 기계식 가슴 압박 장치와 생존율과의 관계를 환자 이송 시간에 따라 분석하며, 마지막으로 3) 심정지 치료 센터가 병원 밖 심정지 환자의 신경학적인 예후에 미치는 영향을 파악하고자 하였다. 방법 본 연구는 연구 대상은 병원 밖 심정지 환자로 2015-2019까지의 한국 심정지 연구 컨소시움 데이터베이스를 활용하였다. 한국 심정지 연구 컨소시움 데이터베이스는 다기관에서 수집하는 심인성 병원 밖 심정지 데이터로 환자들의 일반적 특성, Utstein 가이드 라인에 근거한 정보, 병원 전 단계와 병원 단계 정보를 수집하여 구축된 자료이다. 첫번째 연구 목적을 위해서는 환자-대조군 분석을 시행하였으며 환자군은 한국 심정지 연구 컨소시움 데이터베이스에 2015-2017년까지 등록된 병원 밖 심정지 환자이며 대조군은 같은 기간 성별, 연령, 거주지역을 1:2로 매칭하여 지역사회 건강조사 데이터에서 추출하였다. 고혈압 진단받지 않은 군과 비교하여 고혈압 진단 및 치료여부에 따른 병원 밖 심정지 발생의 오즈비 산출을 하였으며 당뇨, 흡연 그리고 알코올 소비 행태 등을 보정한 다변수 조건부 로지스틱 회귀 분석을 시행하였다. 두번째 연구 목적은 2015-2019까지 수집된 한국 심정지 연구 컨소시움 데이터베이스를 이용하여 현장에서 환자 이송거리에 따른 기계식 흉부 압박 장치의 효과를 분석하였다. 노출변수는 기계식 압박장치 사용여부와 환자 이송거리이며 결과변수는 생존 입원이다. 성별, 연령, 병력, 초기 심전도 리듬, 목격여부, 일반인 심폐소생술 제공 여부, 발생 장소, 구급대 처치 및 반응 시간 등을 보정한 다변수 로지스틱 회귀 분석을 통하여 보정 오즈비와 95% 신뢰구간을 산출하였다. 환자 이송거리에 따른 기계식 압박장치 효과는 교호분석을 하였다. 환자 이송시간은 기존 논문에 근거하여 0-5분, 6-10분, 10분 이상으로 분류하였다. 선택 편향과 혼란 변수를 보정하기 위해서 2가지 방법의 성향 점수 매칭을 시행하였다. 마지막 연구 목적은 경기도를 기반으로 시행되었으며 결과 산출을 위하여 질병관리청의 국가 급성 심장정지 데이터를 이용하였다. 경기응급의료지원센터에서 2015년 병원 밖 심정지 환자를 수용하는 병원 전체를 대상으로 병원 내 자원, 인력, 소생 후 치료 및 예후 평가 여부 그리고 시스템과 관련된 문항의 설문을 시행하였고 그 결과를 바탕으로 병원 밖 심정지 수용 병원을 최종 심정지 치료 병원과 초기 심정지 치료 병원으로 분류하였다. 노출 변수는 최종 심정지 치료 병원 여부 이며 결과 변수는 좋은 신경학적 결과로 뇌 기능 분류 척도 1 또는 2로 정의하였다. 최종 심정지 치료 병원과 초기 심정지 치료 병원의 좋은 신경학적인 결과는 다변수 로지스틱 회귀 분석을 통하여 보정된 오즈비와 95% 신뢰 구간으로 산출되었다. 결과 첫 번째 연구 목적을 위해서는 총 2,633건의 병원 밖 심정지 환자와 5,266건의 지역사회 대조군 매칭이 되었으며 당뇨, 흡연 그리고 음주 행태를 보정한 결과 고혈압 진단으로 인한 병원 밖 심정지 발생의 오즈비는 1.19 (1.07-1.32)이었다. 고혈압 치료를 받지 않은 군의 병원 밖 심정지 발생 오즈비는 3.41(2.72-4.24)이었으며 고혈압 치료를 받는 군의 병원 밖 심정지 발생 오즈비는 0.96 (0.86-1.08)이었다. 두번째 연구 목적을 위해 2015-2019년 구급대가 소생을 시도한 총 3,530명의 병원 밖 심정지 환자를 분석하였다. 환자 이송거리가 10분 초과인 경우 기계식 압박 장치 이용 생존 효과의 교호 작용의 오즈비가 1.49 (1.05-2.21)로 유의했다. Optimal과 Greedy 성향 점수 매칭으로 각각 2,049건의 코호트가 추출되었다. 두 성향 점수 매칭된 코호트 모두 환자 이송거리가 10분 초과인 경우 기계식 흉부압박 장치 사용에 의한 생존 효과의 교호 작용이 각각 오즈비 1.72 (1.24-2.46), 1.80 (1.28-2.53)으로 유의했다. 최종 심정지 치료 병원 분류를 위한 병원 자원 조사를 총 90개 병원 대상으로 실시하였으며 77개의 병원이 응답을 하였다. 병원 자원 조사 결과를 바탕으로 15개 병원을 최종 심정지 치료 병원으로 분류가 가능하였으며 그 외 병원을 초기 심정지 치료 병원으로 분류하였다. 2012-2014년 경기도 소개 병원으로 이송된 9,912건의 병원 밖 심정지 환자 중 4,036명의 환자가 병원 전에서 최종 심정지 치료 병원으로 이송되었다. 초기 심정지 치료 병원과 비교하여 최종 심정지 치료병원으로 이송된 환자들의 좋은 신경학적인 예후 오즈비는 2.10 (1.51-2.93)이었다. 생존 입원한 환자들만 추가 분석하였으며 최종 심정지 치료병원에 입원한 환자들의 좋은 신경학적인 예후 오즈비는 1.48 (1.02-2.14)로 분석되었다. 결론 본 연구는 다기관 병원 밖 심정지 환자를 대상으로 심정지 발생을 줄이고 생존률을 향상시키기 위한 개인수준, 병원 전 단계 그리고 치료시스쳄에서의 전략을 분석했다. 고혈압은 심인성 병원 전 심정지 발생에 독립적인 위험 요인이였으며 가장 높은 위험 인구군은 고혈압을 진단받았으나 치료를 받지 않은 군이었다. 기계식 압박 장치의 사용은 현재까지 근거가 미약하지만 환자 이송 거리가 긴 경우 등의 특별한 상황에서는 생존 효과는 수기 압박보다 높았다. 최종 심정지 치료병원의 지정과 효과적인 병원 환자 이송 및 병원 간 이송 프로토콜 수립, 적용한다면 병원 밖 심정지 환자의 생존율을 높일 수 있을 것이다. 병원 밖 심정지 환자의 질병 부담을 줄이기 위해서는 심정지 발생 예방 및 환자의 생존율 향상을 위한 전략 개발 및 자원 배분이 매우 중요하다. 따라서 심정지 발생 및 생존을 위한 여러 단계에서의 위험요인 및 고위험군을 파악하여 예방 전략을 수립하고 현재 가용중인 자원을 효과적으로 사용할 수 있는 전략 수립 및 적용이 필수적이다.

Serum total testosterone level predicts survival of male patients with terminal cancer

김세원 가천대학교 의학전문대학원 2017 국내석사

Cancer is a major cause of death and illness worldwide: an estimated 14.1 million new cancer cases and 8.2 million cancer deaths occurred in 2012. In Korea as well, cancer has been the leading cause of death since 1983 and has been paid much public health concern. Based on the nationwide cancer statistics, the incidence rate for all cancers showed an annual increase of 3.3% from 1999 to 2009 [1] (Figure 1). Although early detection and developed cancer treatments have led to increased survival rates, death is inevitable. With prolonged survival rates, concerns on the cancer trajectory have shifted to better quality of life from simply increased life expectancy. Palliative care is interdisciplinary approach that focuses on improving quality of life for patients with any serious illness and for their families [2]. It aimed to provide various support to patients and their families, by managing physical, psychological, or spiritual suffering. Ideally, palliative care should be initiated at the time of diagnosis. By contrast, hospice is a formal system that provides palliative care services to the dying in the last months of life. It has been developed in oncology setting, because survival prediction of cancer patients is relatively easier than those of other terminal illnesses. Among the end-of-life issues, estimation of remained life is one of the greatest concerns of terminally ill cancer patients and is essential in planning of palliative care. Identification of objective prognostic factors for survival will enhance the accuracy of predicted life remained, thus will improve end-of-life care. Many clinical or laboratory factors have been studied as prognostic indicators for survival in patients with terminal cancer. Other than tumor-related factors, patient related factors such as comorbidities, performance status, presence of certain symptoms, patient-reporting quality of life were associated with survival. Compared with other prognostic factors, laboratory indicators are objective and easy to estimate. Thus, laboratory parameters including leukocytosis, lymphocytopenia, anemia, hypoalbuminemia, elevated serum acute phase reactants;c-reactive protein (CRP) or vitamin B12, or elevated serum lactate dehydrogenase (LDH) were extensively studied as prognostic parameters for survival in patients with advanced cancer, of which consistency of evidence is variable across different many studies [3]. LDH catalyzes the change of lactate to pyruvic acid and transfers a hydride between molecules.LDH may be used as a tumor marker, because many cancers can raise LDH levels: measuring LDH levels can be helpful in monitoring cancer treatment [4]. Many noncancerous conditions (i.e., encephalitis/meningitis, acute pancreatitis, HIV and lung or liver disease, heart failure, hypothyroidism, anemia, hemolysis, pre-eclampsia) can also raise LDH levels. Clinical estimation of life span by physician was also a significant factor [5-9]. Many studies [10-13] constructed different prognostic scoring system for estimating life expectancy in these population, incorporating clinical prediction by physician, nutrition related symptoms, cognitive impairment, dyspnea, performance status or some laboratory values. Testosterone is metabolized in two ways: i)reductive metabolism leading to dihydrotestosterone; ii)oxidative metabolism to estradiol [14] (Figure 2). The reductive metabolism takes place in target tissues like skin, prostate, and liver. The enzyme 5a-reductase is especially active in the prostate giving 5a-dihydrotestosterone. In the liver, both 5a- and 5b-reductases are active and convert testosterone and other steroids into 5a- and 5b-steroid skeletons.In male, a very small portion of the testosterone is converted into estradiol by oxidative metabolism; most of this process happens in adipose tissue. Testosterone depletion in male adults has been suggested to be linked with decreased bone mass, muscle weakness, and sexual dysfunction [15]. Hypogonadism was also associated with poor survival in many studies. Low serum testosterone has also been associated with higher risk of cardiovascular disease, and has further been related to their poor survival [16, 17]. In aged over 50 years old, lower serum testosterone (<241 ng/dL) had higher risk of death, regardless of preexisting health conditions [18]. In men treated with hemodialysis, lower serum testosterone(≤8.1 nmol/L) was significantly related to increased all-cause mortality and cardiovascular disease mortality after adjustment for other risk factors [19]. Hypogonadism is common in cancer patients [20, 21]. The reason and clinical impact of hypogonadism in male cancer patients is unclear; hypothalamic-pituitary-gonadal axis dysfunction, chronic inflammation, chemotherapy, corticosteroids, or opioids are possible causes [22-24]. Hypogonadic or depressed serum total testosterone concentration(<12 nmol/L) was associated with greater tumor burden [25] or symptom burden [20, 21]. Only one study [21] examined testosterone and survival in cancer patients; lower total testosterone (≤200 ng/dL) was an independent poor prognostic factor for survival along with poor performance status, high CRP level, and hypoalbuminemia in male patients with advanced cancer, however, subjects were not all terminally ill patients and the timing of laboratory workup is not clear. In this study, we aimed to investigate the association between serum total testosterone concentrations and life expectancy, focusing terminally ill cancer patients admitted in hospice care unit.

An Elite Analysis of the Disintegration and Survival of State Socialism : The North Korean Political Elite in Comparative Perspective

권소영 University of Cambridge 2003 해외박사

This thesis examines the divergent outcomes of the disintegration and survival of state socialism within an elite paradigm, focusing on elite composition and functioning. The thesis pursues the following objectives. Firstly, it studies an elite paradigm to draw a trajectory of system change and survival. Secondly, it examines the cases of system disintegration within the elite paradigm to generalise about the conditions underpinning system change and regime breakdown. Thirdly, within the same paradigm, it places the survival cases in a comparative framework and investigates how system survival of state socialism was possible in some countries. Elite theorists define the political elite in a state socialist formation as being ideologically unified, with narrow differentiation and strong unity. My thesis challenges this contention by demonstrating that the nature and structure of the political elite underwent evolutionary changes under state socialism. The thesis further argues that the degree of change from the original 'ideocratic' configuration determines system disintegration and two types of survival - reform survival and system survival. Three cases are considered (the Party Central Committees of the Soviet Union, China, and North Korea), each of them representing a different regime outcome. A more detailed case study is made of the North Korean political elite, 1980-2000, in comparison to the Soviet and Chinese political elites. The study reveals that the structure of the Soviet and Chinese political elites was significantly diversified in the 1980s; it becoming larger, functionally specialised and socially heterogeneous. Integration and unity of the Soviet political elite was weakened considerably at the end of the 1980s; it was no longer ideologically unified, but severely fragmented. The Chinese political elite maintained relatively strong unity and cohesiveness despite structural differentiation and differing viewpoints on policy. It changed from an ideocratic elite type to a consensual one in which elites agreed on the fundamental structures of the political and social system. Compared to the USSR and China, North Korea showed a reverse pattern of elite change. Facing both exogenous and endogenous crises in the 1990s, the North Korean political elite became smaller in size, exclusive and homogeneously centralised. It formed a compact unified elite in which elites were interconnected socially, economically and ideologically and were able to act in a conscious and cohesive manner. A high level of ideological consensus amongst the ruling group was evident in the 1990s, which was reflected in the political rhetoric and slogans fabricated by the elite to legitimise regime survival. The case study suggests that the North Korean political elite in the 1990s was of an . ideologically unified type with little elite differentiation and strong unity. On the basis of the comparative study, the thesis establishes a connection between elite features and system outcomes of state socialism. It is concluded that a state socialist regime survives in its original form when the political elite maintains narrow differentiation and strong cohesion; it disintegrates when elite division prevails concurrently with substantial elite differentiation; and it survives through reforms when elite differentiation is wide but elite unity remains strong.

Influence of Vicarious Learning on Start-up enterprises Survival

CHEN CAN 가천대학교 일반대학원 2022 국내박사

Abstract The enterprise is the main body of the market, the survival and development of the enterprise is not only related to its own, but also will have an impact on social stability and economic development. The problems of short survival time and difficult access to capital are common in Chinese enterprises, especially in small and medium-sized private enterprises. Therefore, it is very necessary to study the factors affecting the survival of enterprises. This study explores the relationship between vicarious learning and firm survival, tries to divide firm vicarious learning into leader learning and codelearning, and focuses on the relationship between codelearning, leader learning and firm survival, as well as the effect of vicarious learning on firm survival under the influence of policy incentives and firm size. Furthermore, this paper studies the following questions :(1) does the vicarious learning behavior of start-ups significantly improve the survival time of enterprises? (2) Does codelearning and leader learning intensity increase the survival time of enterprises? (3) Does policy incentive and support play a moderating role in the relationship between indirect learning and enterprise survival time? (4) Does the internal resources of the firm -- the size of the firm affect the relationship between indirect learning and the duration of the firm? This article first to the leader of the study, codelearning, the concept of enterprise survival is defined, then the logical relationship between the three variables for literature, then, in the vicarious learning theory, system theory, signal theory, enterprise based on the analysis of the competitive advantage theory put forward in this paper, the hypothesis that build regression model. Select China Industrial enterprise database enterprises as the research object, collect and manually sort out the relevant data, eliminate the incomplete data of enterprise samples, and put forward the hypothesis of this paper on the basis of the analysis of the logical relationship between variables, take the 2011-2015 China industrial enterprise database related data as the research sample. It was processed and analyzed using Excel2010 and Stata 16.0. Descriptive statistical analysis, correlation analysis, multiple regression analysis and robustness test were conducted for the relationship among leader learning, codelearning and firm survival. After research and analysis, this paper draws the following conclusions: Start-up companies first, instead of learning behavior significantly influence the survival time of the enterprise, including codelearning is significant positive enhancement effect, leader of the study is significant negative effects, because the leader of the enterprise has the area lies between sex, leader of the enterprise itself has the key resource is different from a competitor's unique resources, is the enterprise sustainable development support, That is to say, the leader enterprise has key resources that the codeenterprise does not have. The simple learning and imitation of the leader enterprise by the start-up enterprise may not improve the performance of the enterprise, because there is no early basis for the implementation of its strategy. Second, government incentive has a significant positive moderate effect on the main effect. When the industrial policy environment is in a positive and encouraging situation, enterprises will form a positive judgment on the future development of their own industry, and form a good incentive effect on the survival and development of enterprises. Thirdly, the enterprise size has a significant positive moderate effect on the main effect. The larger the enterprise has a higher sunk cost, which means a higher exit barrier, leading to a relatively long survival time of the enterprise. Finally, this part analyzes the empirical results and conclusions, and puts forward relevant suggestions from the perspective of the country, the government and enterprises respectively. The innovation of this paper includes the following points: First, this paper divides the codevicarious learning behavior of enterprises into two types, namely "leader learning" and "codelearning", and verifies the empirical analysis to enrich the existing research on the effect of organizational learning. Most of the previous studies only examined the impact of vicarious learning behavior on firm performance, but few further distinguished codevicarious learning and studied the relationship between such behavior and firm survival. In this regard, this paper expands the previous research basis, which is helpful to deepen the understanding of learning imitation effect of enterprises, and provides some reference for enterprises' strategic decision-making. Second, corporate behavior is embedded in a certain social context, and social context factors will also have an important impact on corporate behavior. This paper focuses on the convergence learning and imitation behavior of enterprises in the context of macro policies and enterprises, especially on the two important moderating factors of policy incentive and enterprise size, which provides a new perspective for in-depth understanding of the formation of the mechanism of the influence of enterprise learning behavior on enterprise survival. Third, most of the existing studies measure the impact of policies on enterprise decisions with dummy variables, and fail to carry out quantitative analysis on policy intensity, persistence and enterprise policy solutions, which weakens the practical application of relevant conclusions. Moreover, it cannot show the industry heterogeneity preference of government policies, so there is a lack of relevant research on the response of industry differentiation policies. Therefore, this paper adopts text analysis tools and word frequency analysis as the main means to measure policy incentive variables, so as to explore the industry heterogeneity of policy measurement and deepen the understanding and recognition of the impact of firm learning behavior on firm survival. Fourthly, existing studies on the relationship between vicarious learning and firm survival in start-ups are scarce, and a systematic research framework has not yet been formed, making it difficult to fully reveal the causal relationship and conditional effect between the two. Combining with the Chinese government policy situation and enterprise internal resources of special research, emphatically analyzes the policy incentives, firm size moderate under the action of start-ups vicarious learning behavior and the relationship between the enterprise survival, enrich the research of enterprise relationship between learning and enterprise survival theme, also for related research provides a relatively complete theoretical framework. Key words: Vicarious Learning business survival policies encourage business size

Creative destruction mechanism of Korean industry from the perspective of industrial dynamics

이헌준 서울대학교 대학원 2017 국내박사

Schumpeter's study on economic growth and technological progress in the capitalist system as a process of creative destruction has influenced many studies on industry dynamics. In addition to Schumpeterianism, evolutionary economics and organizational ecology also attempted to grasp the sources of industrial dynamics. Although each perspective is slightly different, they all perceived competition in the market as a major source of industrial dynamics. In particular, Schumpeterianism emphasized Schumpeterian competition that firms are competing their competitive advantage originated from innovation as a main source of industry dynamics. In this perspective, this study attempted to analyze the creative destruction mechanism of Korean industry. In particular, we tried to describe the selection criteria exists in the Korean industry through empirical analysis of exit firms. First, we reviewed theoretical background and the empirical analysis on the survival of firms and derived stylized facts on firm survival. The stylized facts were classified into individual level, firm level, industry level, and macroeconomic level. At the individual level, it was possible to derive a stylized fact that the higher the level of education and experience of the organization members, the better the survival of the firm. At the firm level, the firm size, age, R&D investment, and exporting and were identified as significant determinants on the firm survival. At the industrial level, it was reported that the firm entry rate, industrial growth rate, which determine the degree of competition of the industry, and technology intensity as determinants on the survival of the firm. At the macroeconomic level, we were able to derive a stylized fact that firms' survival rate is procyclic to upturns and downturns of the economy. Second, survival analysis was implemented to describe the selection criteria of Korean industry through firm level micro data. The results showed that the stylized facts on the survival such a as firm size, age, and R&D investment is also found in Korean industry. In addition, we found the changes in the firm selection criteria as a result of restructuring of the financial sector and the industry sector in the process of overcoming the Asian financial crisis. More specifically, it was found that there was a change in firm financial management behavior before and overall incentive in terms of firm survival for the firm's investment activity was reduced after the Asian financial crisis. Third, we focused on the cleansing effect hypothesis in economic recessions. This study investigated two recessions in Korea, the Asian financial crisis and the global financial crisis. We measured total factor productivity using micro level manufacturing plant data from 1993 to 2013 and decomposed the source of the changes in total factor productivity to measure the cleansing effect in two large recessions. During the Asian financial crisis, there was no evidence to support a cleansing effect hypothesis. In contrast, during the global financial crisis, we found the evidence of a cleansing effect. Additionally, we found differences in market selection criteria in the two recessions; by the global financial crisis, the market selection criteria had changed to enable a more conducive environment for the creative destruction process. Fourth, the problem of zombie companies was investigated from a different perspective. Previous studies have recognized zombie companies as a factor that hinders the creative destruction process and recognized that they should be exited through restructuring. However, this study focuses on the fact that the problem of zombie firm may be different according to the financial system of the country. Specifically, we analyzed the characteristics of recovering firms and exiting firms in the credit based financial system such as Korea. Based on the firm level micro data, it was found that the firms with high amount of accumulated knowledge showed higher probability of recovering to the normal firms. Also, it is confirmed that the financial sector was not able to identify and support selectively firms between recovering firms and exiting firms.

Identification of synthetic survival and synthetic survival burden in multiple cancer

임영균 서울대학교 대학원 2015 국내석사

Background: According to cancer genomics become prevalent, large cancer genome projects are launched and almost finished. From these project, a huge amount of genomic, transcriptomic, epigenomic data is released in public. This public data have provided quite some studies of somatic mutations in cancer cell, then many researchers have identified several cancer-related genes. However, some existing cancer genomics studies have a limitations. Many studies are restricted to single cancer type or single cancer driver gene. It can limit study’s scope to identify how cancer occur, develop and progress. To solve this problem, a concept of synthetic lethality is arisen. Synthetic lethality is a concept that a combination of mutations in two genes lead to cell death, while a mutation in an only one of these genes remains cell survival. We propose a concept synthetic survival based on the idea of synthetic lethality. Synthetic Survival (SS) is a concept that a combinations of mutations in two gene lead to patients survival, whereas a mutation in only one gene lead to patients bad prognosis. The goal of this dissertation is to identify synthetic survival and synthetic survival burden in multiple cancers. Method: We downloaded 9,184 patients’ clinical information and 6,936 patients’ somatic mutation data in 20 cancer types from The Cancer Genome Atlas. all of data is open-accessed data. To make core data set for further analysis, several criteria is applied to filter out data which is not able to analysis. All genes harboring at least one non-synonymous mutations are annotated to gene damaging score. Gene damaging score is a measure to quantify each gene’s deleteriousness. Gene damaging score is derived from both scores in Sorting Intolerant From Tolerant and classification of LoF mutations. By gene damaging score, patients are stratified to four groups. Survival analysis are conducted for all gene pairs’ group with cox proportional hazard model with penalized likelihood. From the cox model, candidates of potential synthetic survival pairs are decided. Result: The result from filtering raw data, we build the core data set for 4,844 patients including clinical information and somatic mutation profiles. For candidates of potential synthetic survival pairs, 436 gene pairs are identified in 5 cancer types. We also identified synthetic survival burden which means that the group which have more SS pairs and triplets have higher survival rate than the group which have less SS pairs Discussion: In this dissertation, Candidate SS pairs are identify by only performing survival analysis with cox proportional hazard model and also ensure that the number of SS pairs is related to patient’s prognosis. Genes consisting SS pairs are usually related to cell-cell interaction or migration so might be action to prevent metastasis, then shows good prognosis. This study might be approach to reduce cost for screening and to apply practical clinical uses.

Prognostic significance of microvascular invasion and related microRNAs after hepatic resection for hepatocellular carcinoma

박용근 아주대학교 일반대학원 2018 국내박사

Prognostic significance of microvascular invasion in tumor stage for hepatocellular carcinoma Background: The presence of microvascular invasion (McVI) in hepatocellular carcinoma (HCC) has been proposed as a cause of recurrence and poor survival, although this has not been officially emphasized in staging systems. Thus, we conducted a retrospective study to investigate the prognostic importance of McVI in tumor staging in patients with HCC who underwent hepatic resection. Patients and Methods: A retrospective analysis was performed of patients who underwent hepatic resection for HCC at our center from 1994 to 2012. Patients with HCC were classified into four groups based on the presence of McVI and extent of gross vascular invasion (VI). Results: The 5-years overall and recurrence-free survival rates of 676 patients were 63.3% and 42.6%, respectively. There was no difference in tumor recurrence or survival rate between patients with HCC and McVI without gross VI and those with gross VI confined to segmental/sectional branches. Multivariate analysis revealed that the extent of VI based on the presence of McVI and gross VI was independently associated with tumor recurrence and overall survival. Conclusions: McVI was revealed to be an important risk factor similar to gross VI confined to a segmental/sectional branch in patients with HCC who underwent hepatic resection. This finding should be considered when estimating the stage for prognosis. Conditional survival analysis demonstrates that recurrence risk of surgically treated hepatocellular carcinoma evolves with time Objective: The study aim was to investigate long-term change in tumor recurrence risk in patients with hepatocellular carcinoma (HCC) after hepatic resection. Recurrence probability over time was estimated by conditional survival (CS) analysis. Patients and Methods: Early-stage HCC patients with hepatic resection were selected for inclusion from our surgery database. Variables predictive of tumor recurrence were identified by univariate and multivariate analyses. Five-year recurrence-free CS probability was calculated for all patients and for risk groups stratified by independent predictors. Results: In this series of 436 patients, tumor size >5 cm, microvascular invasion, positive resection margin, liver cirrhosis, and a indocyanine green retention ratio at 15 minutes (ICG-R15) >20% were independently predictive of tumor recurrence. The estimated 5-year recurrence-free CS probability improved with each additional year of recurrence-free survival, and the improvement was significantly greater in the high risk than in the low or intermediate risk groups. Conclusion: CS provides added value during follow-up of early-stage HCC patients treated by surgical resection. MicroRNA-9 overexpression is associated with microvascular invasion and poor survival after hepatic resection for hepatocellular carcinoma Purpose: Although microvascular invasion (McVI) has prognostic value for patients with hepatocellular carcinoma (HCC) who have undergone hepatic resection, few studies have investigated the relationship between McVI and the aberrant expression of microRNAs. This study identified microRNAs selectively expressed in HCC with McVI and investigated their prognostic roles. Patients and methods: Clinical data and microRNA expression profiles for 355 HCC patients were extracted from The Cancer Genome Atlas database. MicroRNAs that were differentially expressed in the patients with McVI and those without vascular invasion were identified and investigated as potential prognostic factors for HCC. Results: MicroRNA-9-5p was upregulated more (fold change [FC] 2.30; false discovery rate [FDR] < 0.001) and microRNA-675-5p was downregulated more (FC 0.52; FDR = 0.005) in the patients with McVI. Multivariate analysis revealed that the types of surgery (HR 1.718, 95% CI 1.169–2.524, P = 0.006), the presence of a residual tumor (HR 3.475, 95% CI 1.507–8.013, P = 0.003) and advanced TNM stage (HR 1.817, 95% CI 1.163–2.839, P = 0.009) were independently associated with tumor recurrence, and that advanced TNM stage and overexpression of microRNA-9 were independent risk factors for overall survival after hepatic resection for HCC (HR 1.671, 95% CI 1.114–2.508, P = 0.013 and HR 3.451, 95% CI 1.796–6.630, P < 0.001, respectively). Conclusion: Overexpression of microRNA-9-5p was associated with McVI and poor survival of patients after hepatic resection for HCC.