EMPIRICAL BAYES ESTIMATION OF RESIDUAL SURVIVAL FUNCTION AT AGE

Liang, Ta-Chen The Korean Statistical Society 2004 Journal of the Korean Statistical Society Vol.33 No.2

The paper considers nonparametric empirical Bayes estimation of residual survival function at age t using a Dirichlet process prior V(a). Empirical Bayes estimators are proposed for the case where both the function ${\alpha}$(0, $\chi$] and the size a(R$\^$+/) are unknown. It is shown that the proposed empirical Bayes estimators are asymptotically optimal at a rate n$\^$-1/, where n is the number of past data available for the present estimation problem. Therefore, the result of Lahiri and Park (1988) in which a(R$\^$+/) is assumed to be known and a rate n$\^$-1/ is achieved, is extended to a(R$\^$+/) unknown case.

Optimal Convergence Rate of Empirical Bayes Tests for Uniform Distributions

Liang, Ta-Chen The Korean Statistical Society 2002 Journal of the Korean Statistical Society Vol.31 No.1

The empirical Bayes linear loss two-action problem is studied. An empirical Bayes test $\delta$$_{n}$ $^{*}$ is proposed. It is shown that $\delta$$_{n}$ $^{*}$ is asymptotically optimal in the sense that its regret converges to zero at a rate $n^{-1}$ over a class of priors and the rate $n^{-1}$ is the optimal rate of convergence of empirical Bayes tests.sts.

Intravenous Immunoglobulin to Suppress Progression in a Patient With Advanced Breast Cancer

Kai-Liang Lin,Ta-Chung Chao,Ming-Han Chen 한국유방암학회 2022 Journal of breast cancer Vol.25 No.3

Intravenous immunoglobulin (IVIG) is used to treat various diseases and has anticancer effects that suppress metastases in animal models of sarcoma and melanoma. However, these effects have been observed in a limited number of clinical cases. We report the case of a patient with metastatic breast cancer in which long-term IVIG treatment stopped disease progression in the absence of salvage chemotherapy. The patient was treated with IVIG for the treatment of immune thrombocytopenia. Surprisingly, the lung and brain metastases were stabilized, and the patient achieved a progression-free interval of 29 months. More cases are needed to investigate and confirm the efficacy of IVIG in solid tumors in the future.

Adverse Events after Non-Hepatic Surgeries in Patients with Liver Cirrhosis Undergoing General or Neuraxial Anaesthesia

( Chien-chang Liao ),( Ta-liang Chen ) 대한간학회 2018 춘·추계 학술대회 (KASL) Vol.2018 No.1

Aims: Limited information is available on the outcomes after surgery in patients with liver cirrhosis receiving general or neuraxial anaesthesia. The purpose of this study is to compare the postoperative complications and mortality between cirrhotic patients with general and neuraxial anaesthesia. Methods: From the claims data from the National Health Insurance program in Taiwan, we identified cirrhotic patients aged ≥ 20 years who underwent non-hepatic surgery in 2008-2013. Using propensity-score matching procedure, we selected 8194 cirrhotic patients with neuraxial anaesthesia and 8194 cirrhotic patients with general anesthesia during the same period for comparison. The adjusted odds ratios (ORs) and 95% confidence intervals (CIs) of postoperative complications and mortality were calculated between cirrhotic patients with general anesthesia and neuraxial anaesthesia. Results: Compared with patients received neuraxial anaesthesia, patients received general anesthesia had increased risk of postoperative pneumonia (OR, 2.08; 95% CI, 1.48-2.90), septicaemia (OR, 1.24; 95% CI, 1.04-1.48), acute renal failure (OR, 2.31; 95% CI, 1.57-3.42), intensive care (OR, 4.63; 95% CI, 3.73-5.75), and 30-day mortality (OR, 2.09; 95% CI, 1.15-3.82). Longer length of hospital stay (P<0.0001) and higher medical expenditure (P<0.0001) were also noted in patients with general anesthesia than patients with neuraxial anaesthesia. Conclusions: Cirrhotic patients received general anesthesia had more complications and higher mortality after non-hepatic surgeries than those received neuraxial anaesthesia. Physicians in gastroenterology and perioperative care team need to pay attention to the anesthesia type for this susceptible population.

Poster Session:PS 0179 ; Endocrinology : Preoperative Risk Assessment for Surgical Mortality in Patients with Diabetes: A Nationwide Population-Based Retrospective Cohort Study

( Chien Chang Liao ),( Ta Liang Chen ) 대한내과학회 2014 대한내과학회 추계학술대회 Vol.2014 No.1

Background: To identify and assess the preoperative risk factors for postoperative mortality in patients with diabetes. Methods: Using reimbursement claims, we performed a nationwide retrospective cohort study of 339,053 patients with diabetes undergoing major surgery. Sociodemographic factors, preoperative coexisting medical conditions, and types of surgery were evaluated for the 30-day postoperative mortality rates. Using conditional logistic regression, adjusted odds ratios (ORs) with 95% confi dence intervals (CIs) and risk scores were analyzed. Results: Among surgical patients with diabetes, 30-day postoperative in-hospital mortality was 1.31%. Gastrectomy (OR 8.02 [95% CI 6.42-10.0]), cardiovascular surgery (OR 6.57 [5.51-7.83]), and neurosurgery (OR 6.01 [5.06-7.14]) were associated with the highest postoperative mortality. Renal dialysis (OR 2.60 [2.38-2.85]) and low income (OR 2.19 [1.90-2.54]) were also important predictors for postoperative mortality in diabetic patients. Diabetic patients with risk scores higher than 15 had the highest postoperative mortality compared with those with baseline scores (OR 27.6 [18.0-42.6]). Thirty-day postoperative mortality correlates closely with increased risk scores (P for trend <0.0001). Conclusions: This study provides preoperative risk scores to predict 30-day postoperative mortality rates in diabetic patients. Type of surgery, renal dialysis and low income were the most infl uential factors for postoperative mortality in diabetic patients.

Poster Session : PS 0150 ; Neurology : Increased Risk of Fracture and Post-Fracture Adverse Events in Patients with Parkinson`s Disease: Two Nationwide Population-Based Retrospective Cohort Studies

( Chien Chang Liao ),( Ta Liang Chen ) 대한내과학회 2014 대한내과학회 추계학술대회 Vol.2014 No.1

Background: The relationship between Parkinson`s disease (PD) and fracture is not completely understood. This study evaluated fracture risk and post-fracture mortality in patients with diabetes. Methods: We identifi ed 1624 adults aged 40 years and older newly diagnosed with PD in 2000-2003 using the Taiwan National Health Insurance Research Database. Comparison cohort consisted of 6496 adults without PD randomly selected from the same dataset, frequency matched by age and sex. Events of fracture in 2000-2008 were ascertained from medical claims. Adjusted hazard ratios (HR) and 95% confidence interval (CI) of fracture associated with PD was calculated. Another nested cohort study consisting of 397,766 patients with fracture admission between 2004 and 2010 calculated the adjusted odds ratios (ORs) and 95% CIs of adverse events after fracture in patients with and without PD. Results: During 51,287 person-years of follow-up, there were 1402 newly diagnosed fracture cases. The incidences of fracture for people with diabetes and without diabetes were 40.4 and 24.1 per 1000 person-years, respectively (P < 0.0001). Compared with people without PD, the adjusted HR of fracture was 1.96 (95% CI 1.74-2.21) for people with PD. The ORs of post-fracture urinary tract infection, pneumonia, septicemia, stroke, and mortality associated with PD were 1.61 (95% CI 1.54-1.69), 1.48 (95% CI 1.40-1.56), 1.45 (95% CI 1.37-1.54), 1.50 (95% CI 1.41-1.60), and 1.27 (95% CI 1.18-1.38), respectively. Conclusions: PD was associated with fracture. Patients with PD had more adverse events and subsequent mortality after fracture. Prevention of fracture and post-fracture adverse events is needed in this susceptible population.

Slide Session : OS-RHEU-04 ; Rheumatology : Increased Risk of Acute Myocardial Infarction and Mortality in Patients with Systemic Lupus Erythematosus: Two Nationwide Retrospective Cohort Studies

( Chien Chang Liao ),( Ta Liang Chen ) 대한내과학회 2014 대한내과학회 추계학술대회 Vol.2014 No.1

Background: This study evaluated the risk of acute myocardial infarction (AMI) and mortality among patients with systemic lupus erythematosus (SLE) in two nationwide retrospective cohort studies. Methods: Using Taiwan`s National Health Insurance Research Database, we conducted a retrospective cohort study and identifi ed 1207 adults newly diagnosed with SLE in 2000-2004. Non-SLE cohort consisted of 9656 adults without SLE, frequency- matched for age and sex and randomly selected from the same data set. Events of AMI were considered as outcome during the follow-up period between 2000 and 2008. Another nested cohort study of 6900 patients with AMI receiving cardiac surgeries was conducted to analyze the impact of SLE on post-AMI mortality. Results: During the follow-up period, there were 52 newly diagnosed AMI cases. The incidences of AMI for SLE cohort and non-SLE cohort were 2.10 and 0.49 per 1000 person-years, respectively, with an adjusted hazard ratio (HR) of 5.11 (95% confi dence interval [CI] 2.63-9.92). For females, the adjusted HR of AMI associated with SLE was as high as 6.28 (95% CI 2.67-14.7). Further analyses in the nested cohort showed SLE was signifi cantly associated with post-AMI mortality (odds ratio, 2.83; 95% CI 1.21- 6.64). Conclusions: Patients with SLE had higher risk of AMI compared with non-SLE control, and this risk was more signifi cant in females. In addition, SLE is an independent risk factor for post-AMI mortality.

Poster Session : PS 0145 ; Neurology : Stroke Risk and Outcomes in Patients with Epilepsy: Two Nationwide Studies

( Chien Chang Liao ),( Ta Liang Chen ) 대한내과학회 2014 대한내과학회 추계학술대회 Vol.2014 No.1

Background: The association between epilepsy and stroke was not completely understood. The purpose of this study is to investigate whether patients with epilepsy have an increased risk of stroke or post-stroke mortality. Methods: Using Taiwan`s National Health Insurance Research Database, we conducted a retrospective cohort study of 3,284 patients with new-diagnozed epilepsy and 13,136 persons without epilepsy between 2000 and 2004. The risk of stroke was compared between 2 cohorts through December 31, 2008. To investigate the association between in-hospital mortality after stroke and history of epilepsy, we conducted a case-control study of 135,220 hospitalized patients with newly diagnosed stroke between January 1, 2005, and December 31, 2008. The post-stroke in-hospital mortality and complications were compared between stroke patients with and without previous epilepsy. Results: The epilpesy cohort had an increased stroke risk (hazard ratio [HR] 2.18; 95% CI 1.87-2.54). Among patients with stroke, those with a history of epilepsy had a higher risk of post-stroke mortality compared withthose without epilepsy (odds ratio 1.26; 95% CI 1.17-1.36). Epilepsy was also associated with post-stroke pneumonia (odds ratio 1.52; 95% CI 1.44-1.62) and urinary tract infection (odds ratio 1.48; 95% CI 1.39- 1.56). The association between epilepsy and stroke risk remains signifi cant in every age group (20-39, 40-49, 50-59, 60-69, >=70) and both sexes. Conclusions: Epilepsy was associated with risk of stroke and post-stroke pneumonia, urinary tract infection and mortality. This research shows the importance of prevention, early recognition, and treatment of stroke in epileptic population.

Poster Session : PS 0318 ; Hematology : Increased Mortality after Stroke in Patients with Immune Thrombocytopenia: A Nationwide Population- Based Study

( Chien Chang Liao ),( Ta Liang Chen ),( Chun Chuan Shih ) 대한내과학회 2014 대한내과학회 추계학술대회 Vol.2014 No.1

Background: Limited information was available on the outcomes after stroke in patients with immune thrombocytopenia (ITP). This study evaluated the adverse events after stroke in patients with ITP. Methods: From the claims data of National Health Insurance Database, we conducted a nationwide retrospective cohort study consisted of 2183 stroke patients with ITP between 2002 and 2009. Using matching procedure by propensity score with age, sex, low income, urbanization, teaching hospital, diabetes, hypertension, hyperlipidemia, myocardial infarction, mental disorders, liver cirrhosis, dementia, renal dialysis, anticoagulants, anti-platelet agents, lipid-lowering agents, and types of stroke, we selected 21830 non-ITP patients with stroke during the index date for comparison. Results: After adjustment for sociodemographics, coexisting medical conditions, medications, and types of stroke in the multivariate logistic regressions, patients with ITP had higher 30-day mortality after stroke compared with stroke patients without ITP (odds ratio=1. 41, 95% confi dence interval=1. 15-1. 74). ITP was also associated withincreased medical expenditure after stroke (odds ratio=1. 13, 95% confi dence interval= 1. 01-1. 27). Conclusions: Our nationwide retrospective cohort study investigated the increased risk of mortality after stroke in patients with ITP.

Poster Session : PS 0435 ; Infectious Disease ; Adverse Outcomes after Major Surgery in HIV-Infected Patients: A Nationwide Matched Cohort Study

( Chien Chang Liao ),( Ta Liang Chen ) 대한내과학회 2014 대한내과학회 추계학술대회 Vol.2014 No.1

Background: Patients with infection of human immunodeficiency virus (HIV) were known to have a higher risk of complications during hospitalization. However, the association between HIV-infection and postoperative adverse events were not completely understood. Methods: We conducted a population-based study in 2566 HIV-infected patients and 10,264 non-HIV controls receiving major surgery from the Taiwan National Health Insurance Research Database in 2004 and 2010. Propensity-score methods were used to construct a matched-pairs cohort that reduced important baseline differences between patients who with and without HIV infection. Adjusted rate ratios (RRs) and 95% confi dence intervals (CIs) of major postoperative complications and 30-day mortality were evaluated among HIV-infected patients with different severity in the multivariable logistic regressions. Results: After the adjustment, HIV-infected patients had higher risks of postoperative pneumonia (RR 1.66; 95% CI 1.25-2.20), septicemia (RR 1.32; 95% CI 1.08-1.62), acute renal failure (RR 2.15; 95% CI 1.27-3.66), surgical site infection (RR 1.59; 95% CI 1.38-1.82) and overall complications (RR 1.46; 95% CI 1.31-1.61). Compared with surgical patients without HIV infection, HIV-infected patients had increased 30-day postoperative mortality (RR 2.71; 95% CI 1.82-4.05), particularly in those with viral hepatitis (RR 4.14; 95% CI 2.23-7.70), pulmonary tuberculosis (RR 11.1; 95% CI 3.39- 36.4), diabetes (RR 6.20; 95% CI 2.11-18.2), preoperative HIV-related hospitalization (RR 3.70; 95% CI 2.38-5.74), opportunistic infection (RR 6.31; 95% CI 3.50-11.4) and highly active antiretroviral therapy (RR 4.14; 95% CI 2.34-7.33). Conclusions: HIV-infected patients showed signifi cantly higher postoperative adverse outcome rates with risk of 30-day mortality nearly fi vefold when compared with patients without HIV infection. Our fi ndings suggest the urgency revising the protocol of postoperative care for this specifi c population.