• Communications for Statistical Applications and Methods
• /
• v.24 no.6
• /
• pp.583-604
• /
• 2017

The most popular regression model for the analysis of time-to-event data is the Cox proportional hazards model. While the model specifies a parametric relationship between the hazard function and the predictor variables, there is no specification regarding the form of the baseline hazard function. A critical assumption of the Cox model, however, is the proportional hazards assumption: when the predictor variables do not vary over time, the hazard ratio comparing any two observations is constant with respect to time. Therefore, to perform credible estimation and inference, one must first assess whether the proportional hazards assumption is reasonable. As with other regression techniques, it is also essential to examine whether appropriate functional forms of the predictor variables have been used, and whether there are any outlying or influential observations. This article reviews diagnostic methods for assessing goodness-of-fit for the Cox proportional hazards model. We illustrate these methods with a case-study using available R functions, and provide complete R code for a simulated example as a supplement.

• Proceedings of the Technology Innovation Conference
• /
• 2005.02a
• /
• pp.42-56
• /
• 2005

Firms cooperating in R&D face a moral hazard problem, because with R&D effort not being observable each partner will focus on its own profit when choosing its effort level. This paper aims to explain the use of optimal license contract for R&D cooperation such as cross-licensing agreement. We argue that in the situations of asymmetric information, the optimal incentive scheme that can solve moral hazard problem is . a linear function of the likelihood ratio. Especially in the case of parallel research, each firm has an extra incentive for cooperative R&D effort, given by the license fee that considers the profit of the cooperating firm, which solely depends on his R&D success if the cooperating firm fails.

• Journal of Wetlands Research
• /
• v.7 no.1
• /
• pp.107-117
• /
• 2005

Soil erosion is influenced from a variety of factors such as rainfall distribution, soil type, land use, etc. This paper is aimed at analyzing the soil erosion hazard zone in cropland. RUSLE was used for an analysis of soil erosion amount, and for the spatial data of basin, soil erosion amount was calculated by extracting the respect topography space related factors of RUSLE using DEM, Landuse, Soil map as base map. This paper is targeting at the watershed of Gyeongan stream in Gyeonggi-do The result of an analysis of soil erosion amount showed that soil erosion occurred in the order of crop field(1210) planting area, orchard(1220), non-adjusted paddy fields(1120), and adjusted paddy fields(1110), and also the average soil erosion in these planting areas has the most amount in crop field planting area. As a result of analysis on soil erosion hazard zone of farm land by classifying it into 5 classes using the result of that result of analysis on the amount of soil erosion, in case of Class 5 in which the hazard of soil erosion is the highest, approximately 72.5ha that corresponds to 2.4% of the total farm land was decided as erosion hazard zone. For this erosion hazard zone, it was analyzed that dry field crop planting area was 72.4ha and orchard was 0.1ha, and Class 5 hazard zone did not appear in other farming areas. Also, it showed that Class II(1~50ton/ha/yr) area had the most ratio of the entire farm land, i.e., 70.2%, regardless of land use state. According to the result of analysis on soil erosion hazard zone of farm land by classifying it into 5 classes, the Class V has the highest soil erosion hazard, approximately 72.5ha that corresponds to 2.4% of the total farm land was estimated as an erosion hazard zone. This erosion hazard shows 72.4ha in dry field crop planting area, 0.1ha in an orchard, but the highest hazard zone, the Class V was not shown in other farming areas. Also, it showed that Class II area had the most ratio of the entire farm land, i.e., 70.2%, regardless of land use state.

• Journal of Gastric Cancer
• /
• v.12 no.3
• /
• pp.164-172
• /
• 2012

Purpose: The purpose of this study is to investigate the prognostic significance of tumor size for 5-year survival rate in patients with gastric cancer. Materials and Methods: A total of 1,697 patients with gastric cancer, who underwent potentially curative gastrectomy, were evaluated. Patients were divided into 4 groups as follows, according to the median size of early and advanced gastric cancer, respectively: small early gastric cancer (tumor size ${\leq}3$ cm), large early gastric cancer (tumor size >3 cm), small advanced gastric cancer (tumor size ${\leq}$ 6 cm), and large advanced gastric cancer (tumor size >6 cm). The prognostic value of tumor size for 5-year survival rate was investigated. Results: In a univariate analysis, tumor size is a significant prognostic factor in advanced gastric cancer, but not in early gastric cancer. Multivariate analysis showed that tumor size is an independent prognostic factor for 5-year survival rate in advanced gastric cancer (P=0.003, hazard ratio=1.372, 95% confidence interval=1.115~1.690). When advanced gastric cancer is subdivided into 2 groups, according to serosa invasion: Group 1; serosa negative (T2 and T3, 7th AJCC), and Group 2; serosa positive (T4a and T4b, 7th AJCC), tumor size is an independent prognostic factor in Group 1 (P=0.011, hazard ratio=1.810, 95% confidence interval=1.149~2.852) and in Group 2 (P=0.033, hazard ratio=1.288, 95% confidence interval=1.020~1.627), respectively. Conclusions: Tumor size is an independent prognostic factor in advanced gastric cancer irrespective of the serosa invasion, but not in early gastric cancer.

• Journal of Chest Surgery
• /
• v.42 no.6
• /
• pp.710-718
• /
• 2009

Background: The Cox maze procedure has been used as a standard surgical treatment for atrial fibrillation for about 20 years. Recently, the creators have used a bipolar radiofrequency electrode (Cox maze IV procedure) instead of the incision and suture (cut-sew) technique to make atrial ablation lesions for persistent atrial fibrillation. We investigated clinical outcomes for the Cox maze procedure with a bipolar radiofrequency electrode and cryoablator in patients with persistent atrial fibrillation, and compared results with clinical outcomes of the cut-sew procedure. Material and Method: Between April 2005 and July 2007, 40 patients with persistent atrial fibrillation underwent Cox maze IV procedure with a bipolar radiofrequency electrode and cryoablator (bipolar radiofrequency group). Surgical outcomes were compared with those of 35 patients who had the cut-sew technique for the Cox maze III procedure. All patients had concomitant cardiac surgery. Postoperatively, the patients were followed up every 1 to 2 months. Result: At 6 months postoperatively, the conversion rate to regular sinus rhythm was not significantly different between the two groups: 95.0% for the bipolar radiofrequency ablation group; 97.1% for the cut-sew technique (p=1.0). At the end of the follow-up period, the conversion rate to regular sinus rhythm was also not significantly different (92.5% vs. 91.6%, p=1.0). In multivariate analysis using a Cox-regression model, the postoperative atrial dimension was an independent determinant of sinus conversion in the bipolar radiofrequency ablation group (hazard ratio 31, p=0.005). In the Cox-regression model for both groups, atrial fibrillation at 6 months postoperatively (hazard ratio 92.24, p=0.003) and the postoperative left atrial dimension (hazard ratio 16.05, p=0.019) were independent risk factors of continuance or recurrence of atrial fibrillation after Cox maze procedures. Aortic cross-clamp time and cardiopulmonary bypass time were significantly shorter in the radiofrequency group than in the cut-sew group. Conclusion: In the Cox maze procedure for patients with persistent atrial fibrillation, the use of bipolar radiofrequency ablation and a cryoablator is as good as the cut-sew technique for conversion to sinus rhythm. The postoperative left atrial dimension is an independent determinant of postoperative continuance and recurrence of atrial fibrillation.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.7 no.5
• /
• pp.11-18
• /
• 2007

A laboratory investigation was conducted to characterize the flexural fatigue behavior of high performance fiber reinforced cement mortar. Five specimens for statics flexural test and fourteen specimens for the flexural fatigue test were made based on the fiber mixing ratio. Static flexural tests were firstly performed to obtain magnitudes of static failure loads and stress levels before flexural fatigue tests. The flexural fatigue behaviors were investigated based on the stress level and fiber mixing ratio. Also, the equations for the interrelation of the flexural fatigue stress levels with the number at loading cycle were proposed.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.6 no.3 s.22
• /
• pp.37-45
• /
• 2006

In this study, displacement ductility capacity and plastic hinge regions of reinforced concrete multi-column bent with different transverse reinforcement ratio are investigated. The ductility increases remarkably as transverse reinforcement ratio increase and the multi-column bent loaded along transverse direction is more ductile. The plastic hinge length for special detailing requirements of transverse reinforcement is estimated. For high target ductility, plastic hinge length for confinement should be extended with increased transverse reinforcement ratio. The plastic hinge length of multi-column bent loaded along transverse direction is shorter than that along longitudinal direction, because of the different moment distribution.

• Journal of Korean Academy of Nursing
• /
• v.52 no.2
• /
• pp.214-227
• /
• 2022

Purpose: This study was aimed at investigating the incidence and risk factors of dyslipidemia in menopausal women using a Korean community-based longitudinal study. Methods: The subjects were 245 postmenopausal women without dyslipidemia who had participated in the Ansan-Ansung cohort study from 2001~2002 (baseline) to 2015~2016 (seventh follow-up visit). The dyslipidemia incidence was measured as incidence proportion (%) and incidence rate per 100 person-years. The predictors of developing dyslipidemia were analyzed with Cox's proportional hazard model. Results: The incidence of new dyslipidemia during the follow-up period was 78.4% (192 patients), and 11.9 per 100 person-years. Mean duration from menopause to developing dyslipidemia was 5.3 years in new dyslipidemia cases. The triglyceride/high density lipoprotein (TG/HDL-C) ratio at baseline (hazard ratio = 2.20; 95% confidence interval = 1.39~3.48) was independently associated with developing dyslipidemia. Conclusion: Dyslipidemia occurs frequently in postmenopausal women, principally within five years after menopause. Therefore, steps must be taken to prevent dyslipidemia immediately after menopause, particularly in women with a high TG/HDL-C ratio at the start of menopause.

• The Journal of the Petrological Society of Korea
• /
• v.26 no.1
• /
• pp.55-62
• /
• 2017

A Lahar on the volcanic area is one of the important hazard that can cause the loss of life and property damage. In order to estimate lahar hazard area at Ulleung Island, we simulated lahar inundation area using Laharz_py. We assumed 400 m of additional elevation for DEM to draw proximal hazard zone of Ulleung Island that H/L ratio were selected 0.45 and 0.5. And lahar volumes for simulation were estimated to 30,000, 50,000, 70,000, $100,000m^3$, respectively. In the results, 5 streams are located near a proximal hazard zone, Jeodong (east), Sadong and Okchon (southeast), Namyang (southwest), and Chusan (north), Nari basin is also considered that has a possibility of lahar during downpour. The results of this study can be used as basic data to make a hazard map for reduce the damage that can be caused by volcanic hazards occurred on Ulleung Island.

• Journal of Hospice and Palliative Care
• /
• v.20 no.4
• /
• pp.235-241
• /
• 2017

Purpose: The Palliative Performance Scale (PPS) is a widely used prognostic tool in patients with advanced cancer. This study examines the association between changes in PPS score and survival in patients with advanced cancer. Methods: We identified a cohort of 606 inpatients who died at a Korean university hospital's hospice/palliative care center. For each patient, the PPS score was measured twice according to a standard procedure: 1) upon admission, and 2) three days after admission (D3). "Change on D3" was defined as a difference between initial PPS and PPS on D3. We used a Cox regression modeling approach to explore the association between this score change and survival. Results: The changes in scores were associated with survival. A score change of >30% yielded a hazard ratio for death of 2.66 (95% CI 2.19~3.22), compared to a score change of ${\leq}30%$. PPS of ${\leq}30$ on D3 also independently predicted survival, with a hazard ratio of 1.67 (95% CI 1.38~2.02) compared to PPS of >30. Conclusion: A change of over 30% in PPS appears to predict survival in hospitalized patients with terminal cancer, even after adjustment for confounders. Changes in PPS may be a more sensitive indicator of impending death than a single PPS measured on the day of admission in terminal cancer patients. Further prospective study is needed to examine this important finding in other populations.