Search

Background Follicular lymphoma (FL) is less prevalent in Asians, but detailed epidemiological analyses were not available. This study aimed to characterize the epidemiologic features of FL in Taiwan to explore the factors relevant to disease development and prognosis. Methods We obtained epidemiological data for Taiwanese citizens during 1990‐2012 from Taiwan's National Cancer Registry Database, and the corresponding data for US Caucasians from the Surveillance, Epidemiology, and End Results Program. Changes in incidence rates were evaluated with age‐period‐cohort (APC) analyses. Patient outcomes were compared with 5‐year relative survival rates (RS) estimates. Results Incidence rates of FL in Taiwan increased continuously during the study period (0.34 to 0.91 per 100 000 person‐year from 1993‐1997 to 2008‐2012 in men, and from 0.29 [1993‐1997] to 0.81 [2008‐2012] in women), while rates in the US remained stable in both sexes, ranging between 3.73 and 3.96 in men and between 3.24 and 3.55 in women. Estimates of average annual percentage changes in incidence were significantly positive in Taiwan, but not in US Caucasians. Notably, the APC analysis identified a strong birth‐cohort effect in Taiwan, corresponding to environmental alterations present during the study period. The estimated 5‐year RS rates in both populations showed steady improvement, but the RS in Taiwanese patients was consistently 10% lower than in US Caucasians. Conclusion A distinct increasing trend of incidence with a strong birth‐cohort effect was identified in Taiwan, providing evidence of the association between environmental factors and disease development.

Background: Invasive dental treatments (IDTs) can yield temporary bacteremia and have therefore been considered a potential risk factor of infective endocarditis (IE). It is hypothesized that, through the trauma caused by IDTs, bacteria gain entry to the bloodstream and may attach to abnormal heart valves or damaged heart tissue, giving rise to IE. However, the association between IDTs and IE remains controversial. The aim of this study is to estimate the association between IDTs and IE. Methods: The data in this study were obtained from the Health Insurance Database in Taiwan. We selected 2 case-only study designs, case-crossover and self-controlled case series, to analyze the data. The advantage of these methods is that confounding factors that do not vary with time are adjusted for implicitly. In the case-crossover design, a conditional logistic regression model with exposure to IDTs was used to estimate the risks of IE following an IDT with 4, 8, 12, and 16 weeks delay, respectively. In the self-controlled case series design, a conditional Poisson regression model was used to estimate the risk of IE for the risk periods of 1 to 4, 5 to 8, 9 to 12, and 13 to 16 weeks following an IDT. Results: In total, 9120 and 8181 patients with IE were included in case-crossover design and self-controlled case series design, respectively. In the case-crossover design, 277 cases and 249 controls received IDTs during the exposure period, and the odds ratio was 1.12 (95% confidence interval, 0.94–1.34) for 4 weeks. In the self-controlled case series design, we observed that 407 IEs occurred during the first 4 weeks after IDTs, and the age-adjusted incidence rate ratio was 1.14 (95% confidence interval, 1.02–1.26) for 1 to 4 weeks after IDTs. Conclusions: In both study designs, we did not observe a clinically larger risk for IE in the short periods after IDTs. We also found no association between IDTs and IE among patients with a high risk of IE. Therefore, antibiotic prophylaxis for the prevention of IE is not required for the Taiwanese population.

Objective: To evaluate the influence of age, screening interval, and histologic type on the effect of Pap smears in cervical cancer screening. Materials and methods: Data were retrieved from the Taiwan National Cancer Registry and Cervical Cancer Screening Registration System for the period from 2002 to 2010. Age, Pap smear interval, FIGO stage, and histology were further analyzed. Results: A total of 12,294 women with cervical cancer were enrolled, including 10,040 with squamous cell carcinoma (SCC), 1720 with adenocarcinoma (ADC), 401 with adenosquamous carcinoma (ASC), and 133 with small cell neuroendocrine carcinoma (SMC). Women who had a Pap smear at an interval of

Asian population-based data evaluating all-cause mortality and cause-specific mortality in patients with psoriasis are limited. This study aimed to evaluate the risk of all-cause mortality (stratified according to onset status, disease severity, and concomitant psoriatic arthritis) and cause-specific mortality in patients with psoriasis. Our study cohort consisted of 80,167 patients with newly diagnosed psoriasis between 2001 and 2012 in the National Health Insurance Database. Vital status and cause of death were ascertained from the National Death Registry of Taiwan. All-cause and cause-specific crude mortality rates and standardized mortality ratios were estimated. A total of 7,198 deaths were identified during the follow-up period (508,505 person-years). The standardized mortality ratios were 1.53 for severe psoriasis (95% confidence interval = 1.45-1.60), 1.47 for early-onset psoriasis (95% confidence interval = 1.34-1.61), and 1.47 for patients with psoriatic arthritis (95% confidence interval = 1.36-1.58). In the cause-specific mortality analysis, the absolute and excess risks of death were highest for malignancies (3.6 and 1.57 deaths per 1,000 patient-years, respectively) and circulatory system diseases (3.0 and 1.44 deaths per 1,000 patient-years, respectively). Patients with severe psoriasis, early-onset psoriasis, and psoriatic arthritis had higher all-cause mortality risks. In particular, patients with psoriasis had higher excess risks of mortality from malignancies and circulatory system diseases.

Background— Extracorporeal membrane oxygenation (ECMO) provides circulatory and respiratory support for patients with severe acute cardiopulmonary failure. The objective of this study was to examine the survival outcomes for patients who received ECMO. Methods and Results— Adult patients who received ECMO from September 1, 2002, to December 31, 2012, were identified from the Taiwan National Health Insurance Database associated with coronary artery bypass graft surgery, myocardial infarction/cardiogenic shock, injury, and infection/septic shock. A Cox regression model was used to determine hazard ratios and to compare 30-day and 1-year survival rates with the myocardial infarction/cardiogenic shock group used as the reference. The mean±SD age of the 4227-patient cohort was 57±17 years, and 72% were male. The overall mortalities were 59.8% and 76.5% at 1 month and 1 year. Survival statistics deteriorated sharply when ECMO was required for >3 days. Acute (30-day) survival was more favorable in the infection/septic shock (n=1076; hazard ratio, 0.61; 95% confidence interval, 0.55–0.67), coronary artery bypass graft surgery (n=1077; hazard ratio, 0.68; 95% confidence interval, 0.61–0.75), and injury (n=369, hazard ratio, 0.82; 95% confidence interval, 0.70–0.95) groups. The extended survival rapidly approached an asymptote near 20% for the infection/septic shock, myocardial infarction/cardiogenic shock (n=1705), and coronary artery bypass graft surgery groups. The pattern of survival for the injury group was somewhat better, exceeding 30% at year-end. Conclusions— Regardless of initial pathology, patients requiring ECMO were critically ill with similar guarded prognoses. Those in the trauma group had somewhat better outcomes. Determining the efficacy and cost-effectiveness of ECMO should be a critical future goal.

Background Despite the recommendations of statins treatment for secondary prevention of atherosclerotic cardiovascular disease (ASCVD), treatment adherence and persistence are still a concern. This study examined the real world practice of long-term adherence and persistence to statins treatment initiated after hospital discharge for ASCVD, and their associated factors in a nationwide cohort. Methods Post discharge statin prescriptions between 2006 and 2012 were extracted from the Taiwan National Health Insurance claims database. Good adherence, defined as proportion of days covered (PDC) ≥0.8 and mean medication possession ratio (MPR), was measured every 180-day period. Non-persistence was defined on the date patients failed to refill statin for 90 days after the end of the last prescription. Their associations with influential factors were analyzed using a generalized estimating equation and Cox’s proportional hazard model. Results There was a total of 185,252 post-discharge statin initiations (from 169,624 patients) and followed for 467,398 patient-years in the study cohort. Percentage of good adherence (mean MPR) was 71% (0.87) at 6-months; declined to 54% (0.68), 47% (0.59), and 42% (0.50) at end of year 1, 2, and 7, respectively. Persistence in statin treatment was 86, 67, 50, and 25% at 6-month, 1-, 2-, and 7-year, respectively. Comparing the statin-cohort initiated from year 2006 to 2012, 1-year persistence increased from 58 to 73%, and 1-year good adherence improved from 45 to 61%. Factors associated with sub-optimal adherence and non-persistence included: prescription by primary care clinics or non-cardiology specialties; patients’ age > 75 years; no history of previous statin use; ASCVD events with ischemic stroke diagnosis; comorbidities of renal disease, liver disease, depression, and chronic obstructive pulmonary disease. Conclusions Despite the improving trends, long-term adherence and persistence of statin treatment were suboptimal in Taiwan. Strategies to maintain statin treatment adherence and persistence need to be implemented to further enhance the positive trend.

The overall survival (OS) of patients with ovarian cancer is poor while epithelial ovarian cancer (EOC) is the most lethal gynecologic cancer. The aim of the present study was to evaluate the clinico-pathologic characteristics, especially the prognostic factors, for patients with epithelial ovarian cancer (EOC) in Taiwan. Information about newly diagnosed patients with EOC from 2009 to 2012 was retrieved from the database of the Taiwan Cancer Registry. Data from 2009 to 2013 for the respective cases from the claims database of Taiwan's National Health Insurance and National Death Registry were then retrieved. Potential prognostic factors were analyzed. The mean age at diagnosis of the 2,498 patients was 52.8 years. Serous carcinoma and clear cell carcinoma were diagnosed in 43.3% and 22.8% of the total patients, respectively. For patients with early-stage disease, taxane-based adjuvant chemotherapy, stage I, and younger age at diagnosis led to better overall survival (p = 0.030, p = 0.002, p

Background and objective Potential hepatoxicity is an important clinical concern when administering immunosuppressive therapies to patients infected by hepatitis B virus (HBV). Tumor necrosis factor inhibitors (anti-TNF) increase the likelihood of hepatitis consequent to HBV reactivation, but reported risks and outcomes vary. We determined the risks of liver enzyme elevation in anti-rheumatic drug users from an HBV-endemic region with differing HBV serostatus. Methods We established retrospective cohorts with rheumatoid arthritis, ankylosing spondylitis, or psoriasis/psoriatic arthritis who: 1) received anti-TNF agents from 1 January 2004 to 30 June 2013; 2) received care from 1 June 2011 to 30 June 2013 but only ever used conventional disease-modifying anti-rheumatic drugs (DMARDs). Serology results defined three subgroups: HBV surface antigen positive (HBsAg+), HBsAg negative/HBV core antibody positive (HBsAg−/HBcAb+), or uninfected. We compared incidences of serum alanine aminotransferase (ALT) exceeding twice the upper reference limit between HBV serostatus subgroups in each treatment cohort. Results Among 783 patients treated with anti-TNF (n = 472) or DMARDs only (n = 311), HBsAg−/HBcAb+ anti-TNF users had incidence of ALT elevation commensurate with uninfected counterparts (6.1 vs. 6.0/100 person-years), compared to 19.6/100 person-years in HBsAg+ patients (standardized rate ratio 3.3, 95% CI 1.3–8.2); none effected had severe or fatal hepatitis and ALT levels in all HBsAg−/HBcAb+ patients remained stable, mostly normalizing spontaneously, or after moderating treatment. Patterns of of ALT elevation associated with differing HBV serostatus in the DMARD cohort, resembled those in anti-TNF users. Conclusions In this large HBV-endemic cohort, the absolute incidence of ALT elevation in anti-TNF users was more than three-fold higher in HBsAg+ patients than in uninfected counterparts; however, no such association was evident in patients with HBsAg−/HBcAb+ serotype, whose risk and outcomes of liver enzyme elevation were similar to uninfected patients, suggesting that anti-TNF use by HBsAg−/HBcAb+ patients is probably safe.

BackgroundThe long-term outcome of patients with both diabetes and schizophrenia remains unclear.AimsTo explore whether having schizophrenia increases the risk of advanced complications and mortality in people with diabetes.MethodThis is a population-based matched cohort study using Taiwan's National Health Insurance Research Database. A total of 11 247 participants with diabetes and schizophrenia and 11 247 participants with diabetes but not schizophrenia were enrolled. We used Cox proportional hazard models to determine the effect of schizophrenia on macrovascular and microvascular complications, and all-cause mortality.ResultsThe adjusted hazard ratios were 1.49 (95% CI 1.32–1.68) for macrovascular complications, 1.05 (95% CI 0.91–1.21) for microvascular complications and 3.68 (95% CI 3.21–4.22) for all-cause mortality in patients with diabetes and schizophrenia compared with those patients with diabetes but not schizophrenia.ConclusionsPatients with both diabetes and schizophrenia had an increased risk of macrovascular complications and all-cause mortality but did not have statistically significant elevated risk of microvascular complications.

Objective: Because the number of long-term survivors of colorectal cancer has increased, second primary cancer has become an important issue. However, previous studies were heterogeneous in design, and few data for Asia-Pacific area were available. Methods: This was a retrospective population-based study conducted using the national database of the Taiwan Cancer Registry. Patients who have histology-proven primary colon cancer and rectal cancer from 1995 to 2005 were enrolled in this study. All second primary cancer events had to be histology proven. The standardized incidence ratio of second primary cancer was used as an indicator. Standardized incidence ratio was counted as the number of observed second primary cancer divided by the expected number of cancer cases in the general population. Results: A total of 65 648 eligible index patients were enrolled, and 3810 second primary cancer events were identified. The standardized incidence ratio for all of the patients was 1.03 (95% confidence interval: 0.99–1.06), which implied that the risk of second primary cancer was not significantly elevated in the index patients compared with that of the general population. The standardized incidence ratio for the patients aged 70 years was 2.52 (95% confidence interval: 2.28– 2.78), 1.18 (95% confidence interval: 1.12–1.23) and 0.80 (95% confidence interval: 0.76–0.84), respectively. In young patients (aged

Background. Many studies have shown that type 2 diabetes mellitus (DM) increases the risk for several types of cancer but not cervical cancer (CC). Although DM and insulin-like growth factor 1 have preclinical and clinical implications for CC, less is known about the prognostic impact of DM on patients with early stage CC. Patients and Methods. We used the nationwide Taiwan Cancer Registry database to collect the characteristics of stage I–IIA cervical cancer patients diagnosed between 2004 and 2008. DM and other comorbidities were retrieved from the National Health Insurance database. Cervical cancer-specific survival (CSS) and overall survival (OS) times of patients according to DM status were estimated using the Kaplan-Meier method. We used a Cox proportional hazards model to calculate adjusted hazard ratios (HRs) for the effects of DM and other risk factors on mortality. Results. A total of 2,946 patients had primary stage I–IIA CC and received curative treatments, and 284 (9.6%) had DM. The 5-year CSS and OS rates for patients with DM were significantly lower than those without DM (CSS: 85.4% vs. 91.5%; OS: 73.9% vs. 87.9%). After adjusting for clinicopathologic variables and comorbidities, DM remained an independent unfavorable prognostic factor for CSS (adjusted HR: 1.46) and OS (adjusted HR: 1.55). Conclusion. In Asian patients with early cervical cancer, DM is an independent unfavorable prognostic factor influencing both OS and CSS, even after curative treatments. Implications for Practice: Type 2 diabetes mellitus (DM) increases the incidence of several types of cancer but not cervical cancer (CC); however, less is known about the impact of DM on patients who already have CC. This study suggests that DM may increase the risk of cancer recurrence and death for early stage CC patients, even after curative treatments. Incorporating DM control should be considered part of the continuum of care for early stage CC patients, and close surveillance during routine follow-up in this population is recommended.

Metformin is the first-line oral therapy for type 2 diabetes with proven benefits against cardiovascular risk. Recent evidence suggested that acarbose might be similar to metformin in glucose-lowering efficacy and cardiovascular risk reduction. Therefore, international guidelines have suggested the use of acarbose as alternative first-line antidiabetic therapy.To compare the cardiovascular outcomes in the first-line users of acarbose vs metformin. DESIGN, SETTING, PATIENTS, AND OUTCOME MEASURES: A nationwide cohort study was conducted by analyzing the Taiwan National Health Insurance (NHI) Database. A total of 17,366 acarbose initiators and 230,023 metformin initiators were identified between January 1, 2009 and December 31, 2010. The primary outcome is hospitalization due to any cardiovascular events, including acute myocardial infarction, congestive heart failure, and ischemic stroke. The propensity score method was used to adjust for baseline differences between the two groups. Patients were followed from drug initiation to the earliest of outcome occurrence, death or disenrollment from NHI, or study termination.In intention-to-treat analyses, acarbose was associated with a higher risk of any cardiovascular event (adjusted hazard ratio [HR]: 1.05; 95% confidence interval [CI], 1.01-1.09), heart failure (HR, 1.08; 95% CI, 1.00-1.16), and ischemic stroke (HR, 1.05, 95% CI, 1.00-1.10) than metformin. No significant difference in risk was found in subgroups of patients with or without underlying hypertension, ischemic heart disease, or cerebrovascular disease. Similar results were found in auxiliary as-treated analyses or analyses stratified by propensity score quintiles.Our data do not support that acarbose has a cardio-protective effect similar to metformin as a first-line antidiabetic agent.

We performed this population-based study to investigate the effects of atorvastatin and rosuvastatin on renal function in patients with type 2 diabetes. From the Taiwan National Health Insurance Pay-for-Performance program for diabetes mellitus database, 2006 to 2009, type 2 diabetic patients aged 40 to 100 years with the first prescription of atorvastatin or rosuvastatin were identified. All the data were linked to the National Health Insurance claims database, 2000 to 2010, to construct longitudinal health care data. The Modification of Diet in Renal Disease equation was used to calculate the estimated glomerular filtration rate (eGFR), and the eGFRs between baseline and the end of follow-up (maximum 2 years) were compared. Totally, 3,601 new users of atorvastatin and 1,968 new users of rosuvastatin were included. The median follow-up was 238 days in atorvastatin users and 210 days in rosuvastatin users. The eGFR at baseline was 72.3 ± 25.9 ml/min/1.73 m(2) in atorvastatin users and 73.7 ± 27.3 ml/min/1.73 m(2) in rosuvastatin users. In both statin groups, we found no significant change in eGFR (+0.1 ml/min/1.73 m(2), 95% confidence interval -0.4 to 0.7, p = 0.62 in atorvastatin users; -0.1 ml/min/1.73 m(2), 95% confidence interval -0.8 to 0.6, p = 0.77 in rosuvastatin users). In conclusion, neither treatment with atorvastatin nor rosuvastatin was associated with a significant change of renal function in type 2 diabetic patients.

A national viral hepatitis therapy program was launched in Taiwan in October 2003. This study aimed to assess the impact of the program on reduction of end-stage liver disease (ESLD) burden. Profiles of national registries of households, cancers, and death certificates were used to derive incidence and mortality of ESLDs from 2000 to 2011. Age-gender–adjusted incidence and mortality rates of hepatocellular carcinoma (HCC) and chronic liver diseases (CLDs) and cirrhosis of adults ages 30-69 years were compared before and after launching the program using Poisson's regression models. A total of 157,570 and 61,823 patients (15%-25% of those eligible for reimbursed treatment) received therapy for chronic hepatitis B and C, respectively, by 2011. There were 42,526 CLDs and cirrhosis deaths, 47,392 HCC deaths, and 74,832 incident HCC cases occurred in 140,814,448 person-years from 2000 to 2011. Male gender and elder age were associated with a significantly increased risk of CLDs and cirrhosis and HCC. Mortality and incidence rates of ESLDs decreased continuously from 2000 to 2003 (before therapy program) through 2004-2007 to 2008-2011 in all age and gender groups. The age-gender–adjusted rate ratio (95% confidence interval; P value) in 2008-2011 was 0.78 (0.76-0.80; P

Background Angiotensin-converting enzyme (ACE) inhibitors have been widely used in the treatment of hypertension, but the comparative effectiveness in reducing mortality among different drugs is seldom reported. Methods We identified hypertensive patients who started captopril, enalapril, lisinopril, fosinopril, perindopril, ramipril, or imidapril therapy from Taiwan's National Health Insurance database between 1 January 2004 and 31 December 2009. Overall and cause-specific mortalities were ascertained through a linkage to Taiwan's National Death Registry. Patients were followed from the initiation of ACE inhibitors to death, disenrollment, or study termination (31 December 2010). A Cox proportional hazard regression model was used to calculate the hazard ratio (HR) and 95% confidence interval (CI), using ramipril as the reference group. Results A total of 989,489 hypertensive patients were included, with a mean follow-up ranging from 3.5 years for imidapril to 4.5 years for enalapril. Captopril initiators had the highest overall mortality rate (117.8 per 1,000,000 person-days) as compared to other ACE inhibitors (54.3-79.4 per 1,000,000 person-days). Patients who started captopril therapy had a significantly increased risk of overall mortality (HR: 1.28, 95% CI: 1.24-1.31) when compared with ramipril. Enalapril (HR: 1.08, 95% CI: 1.05-1.11) and fosinopril (HR: 1.08, 95% CI: 1.05-1.12) were also associated with a modestly increased risk. No difference in mortality was found for lisinopril, perindopril, and imidapril, as compared with ramipril. Conclusions There are differences in the mortality risk associated with different ACE inhibitors. However, potential residual confounding effects might still exist.

BACKGROUND Although irinotecan and oxaliplatin are both standard treatments for advanced colon cancer, it remains unknown whether either is effective for patients with resectable synchronous colon cancer and liver-confined metastasis (SCCLM) after curative surgery. PATIENTS AND METHODS A population-based cohort of patients diagnosed with de novo SCCLM between 2004 and 2009 was established by searching the database of the Taiwan Cancer Registry and the National Health Insurance Research Database of Taiwan. Patients who underwent curative surgery as their first therapy followed by chemotherapy doublets were classified into the irinotecan group or oxaliplatin group accordingly. Patients who received radiotherapy or did not receive chemotherapy doublets were excluded. RESULTS We included 6,533 patients with de novo stage IV colon cancer. Three hundred and nine of them received chemotherapy doublets after surgery; 77 patients received irinotecan and 232 patients received oxaliplatin as adjuvant chemotherapy. The patients in both groups exhibited similar overall survival (median: not reached vs. 40.8 months, p=0.151) and time to the next line of treatment (median: 16.5 vs. 14.3 months, p=0.349) in both univariate and multivariate analyses. Additionally, patients with resectable SCCLM had significantly shorter median overall survival than patients with stage III colon cancer who underwent curative surgery and subsequent adjuvant chemotherapy, but longer median overall survival than patients with de novo stage IV colon cancer who underwent surgery only at the primary site followed by standard systemic chemotherapy (p

Background Liver enzyme elevation is an important and common adverse effect among patients with immune-mediated diseases who receive tumour necrosis factor inhibitors (anti-TNF), and has various causes. Hence, we evaluated the relative risks of developing liver enzyme elevation in anti-TNF users with differing hepatitis B virus (HBV) infection status. Methods At a hospital in central Taiwan, 407 patients with rheumatoid arthritis, ankylosing spondylitis, or psoriasis/psoriatic arthritis received anti-TNF therapy between 1 January 2004 and 30 June 2012. We performed a nested case-control study (n = 368) of cases with serum alanine aminotransferase (ALT) > 40 international units/L ≤ 12 months after starting anti-TNF therapy, and corresponding controls without liver enzyme elevation. Conditional logistic regression was used to evaluate associations between liver enzyme elevation and HBV serostatus, as well as other risk factors. Results Thirty cases were compared to 338 controls. After adjustment for potential confounders, HBV surface antigen-positive (HBsAg+) serostatus was associated with substantially higher likelihood of developing elevated ALT (adjusted odds ratio 7.91, 95% confidence interval (CI) 2.16–31.31) relative to those with an uninfected HBV status; no such association was observed among HBsAg-negative/HBV core antibody-positive (HBsAg–/HBcAb+) patients (adjusted odds ratio 1.00, 95% CI 0.33–3.25). Increased risk of ALT elevation was associated with methotrexate used alone, without folic acid (adjusted odds ratio 11.60, 95% CI 2.52–56.46), and history of ALT elevation (adjusted odds ratio 13.71, 95% CI 4.32–45.75). Conclusions HBsAg+ patients with immune-mediated diseases who received anti-TNF therapy had an approximately eight-fold higher likelihood of liver enzyme elevation than those without HBV infection, whereas patients with HBsAg–/HBcAb+ serostatus had a risk similar to that of uninfected patients. Electronic supplementary material The online version of this article (doi:10.1186/s13075-017-1413-y) contains supplementary material, which is available to authorized users.

Background: Although cancer treatment information has been collected through the Cancer Registry system in Taiwan for more than 10 years, the accuracy of such data has never been evaluated. This study examined the accuracy rate between registrar experience and on-site chart review for the first course of cancer treatment. Methods: In this retrospective chart review study, 392 randomly selected medical records from 14 hospitals were re-abstracted by experienced abstractors. The kappa coefficients of accuracy for the abstracting data were calculated against the gold standard. Correlations between registrar background and workload were then identified through regression analysis. Results: Regarding surgery type, low accuracy rates were noted for gastric cancer (84.0%), oral cavity cancer (84.6%), and bladder cancer (88.9%). For chemotherapy, low accuracy rates were observed for hematopoietic diseases (81.3%) and esophageal cancer (88.0%). For radiotherapy, low accuracy rates were noted for esophageal cancer (80.0%), cervical cancer (81.8%), and lymphoma (85.7%). When stratifying by surgery type after adjustment for hospital caseload, a high accuracy rate was found for cancer registrars who had progressed from basic to advanced licenses within 5 years of graduating. Conclusion: The accuracy rate for the first course of cancer treatment was affected by the cancer type and the experience of cancer registrars, but it was not affected by the workload of cancer registrars. We recommend that cancer registrars with basic licenses upgrade to advanced licenses as soon as possible. Medical record collaboration should establish documentation for checklist of radiotherapy and surgical operation records. Keywords: Data quality, Peer review

Background To facilitate priority-setting in health policymaking, we compiled the best available information to estimate the adult mortality (>30 years) burden attributable to 13 metabolic, lifestyle, infectious, and environmental risk factors in Taiwan. Methods We obtained data on risk factor exposure from nationally representative health surveys, cause-specific mortality from the National Death Registry, and relative risks from epidemiological studies and meta-analyses. We applied the comparative risk assessment framework to estimate mortality burden attributable to individual risk factors or risk factor clusters. Results In 2009, high blood glucose accounted for 14,900 deaths (95% UI: 11,850–17,960), or 10.4% of all deaths in that year. It was followed by tobacco smoking (13,340 deaths, 95% UI: 10,330–16,450), high blood pressure (11,190 deaths, 95% UI: 8,190–14,190), ambient particulate matter pollution (8,600 deaths, 95% UI: 7,370–9,840), and dietary risks (high sodium intake and low intake of fruits and vegetables, 7,890 deaths, 95% UI: 5,970–9,810). Overweight-obesity and physical inactivity accounted for 7,620 deaths (95% UI: 6,040–9,190), and 7,400 deaths (95% UI: 6,670–8,130), respectively. The cardiometabolic risk factors of high blood pressure, high blood glucose, high cholesterol, and overweight-obesity jointly accounted for 12,120 deaths (95% UI: 11,220–13,020) from cardiovascular diseases. For domestic risk factors, infections from hepatitis B virus (HBV) and hepatitis C virus (HCV) were responsible for 6,300 deaths (95% UI: 5,610–6,980) and 3,170 deaths (95% UI: 1,860–4,490), respectively, and betel nut use was associated with 1,780 deaths from oral, laryngeal, and esophageal cancer (95% UI: 1,190–2,360). The leading risk factors for years of life lost were similar, but the impact of tobacco smoking and alcohol use became larger because the attributable deaths from these risk factors occurred among young adults aged less than 60 years. Conclusions High blood glucose, tobacco smoking, and high blood pressure are the major risk factors for deaths from diseases and injuries among Taiwanese adults. A large number of years of life would be gained if the 13 modifiable risk factors could be removed or reduced to the optimal level. Electronic supplementary material The online version of this article (doi:10.1186/s12963-017-0134-4) contains supplementary material, which is available to authorized users.

Background In 2011, two national policies aiming to foster hospice services for terminal cancer patients took effect in Taiwan. The single-payer National Health Insurance of Taiwan started to reimburse full hospice services. The national hospital accreditation program, which graded all hospitals, incorporated hospice utilization in its evaluation. We assessed the impact of these national policies. Methods A cohort of 249,394 patients aged ≥18 years who died of cancer between 2008 and 2013 were identified from the National Death Registry. We retrieved utilization data of medical services and compared the health care utilization in the final month of life before and after the implementation of the new policies. Results After the policy changes, hospice utilization increased from 20.8% to 36.2%. In a multivariate analysis adjusting for patient demographics, cancer features, and hospital characteristics, hospice utilization significantly increased after 2011 (adjusted odds ratio [AOR] 2.35, p Conclusion The national policy changes fostering hospice care significantly increased hospice utilization, decreased invasive end-of-life care, and reduced the medical costs of terminal cancer patients.

Objectives A number of observational studies have reported that, in patients with chronic obstructive pulmonary disease (COPD), β blockers (BBs) decrease risk of mortality and COPD exacerbations. To address important methodological concerns of these studies, we compared the effectiveness and safety of cardioselective BBs versus non-dihydropyridine calcium channel blockers (non-DHP CCBs) in patients with COPD and acute coronary syndromes (ACS) using a propensity score (PS)-matched, active comparator, new user design. We also assessed for potential unmeasured confounding by examining a short-term COPD hospitalisation outcome. Setting and participants We identified 22 985 patients with COPD and ACS starting cardioselective BBs or non-DHP CCBs across 5 claims databases from the USA, Italy and Taiwan. Primary and secondary outcome measures Stratified Cox regression models were used to estimate HRs for mortality, cardiovascular (CV) hospitalisations and COPD hospitalisations in each database after variable-ratio PS matching. Results were combined with random-effects meta-analyses. Results Cardioselective BBs were not associated with reduced risk of mortality (HR, 0.90; 95% CI 0.78 to 1.02) or CV hospitalisations (HR, 1.06; 95% CI 0.91 to 1.23), although statistical heterogeneity was observed across databases. In contrast, a consistent, inverse association for COPD hospitalisations was identified across databases (HR, 0.54; 95% CI 0.47 to 0.61), which persisted even within the first 30 days of follow-up (HR, 0.55; 95% CI 0.37 to 0.82). Results were similar across a variety of sensitivity analyses, including PS trimming, high dimensional-PS matching and restricting to high-risk patients. Conclusions This multinational study found a large inverse association between cardioselective BBs and short-term COPD hospitalisations. The persistence of this bias despite state-of-the-art pharmacoepidemiologic methods calls into question the ability of claims data to address confounding in studies of BBs in patients with COPD.

Objective To measure therapeutic inertia by characterizing prescription patterns using secondary data obtained from the nationwide diabetes mellitus pay-for-performance (DM-P4P) programme in Taiwan. Methods Using reimbursement claims from Taiwan’s National Health Insurance Research Database, a nationwide retrospective cohort study was undertaken of patients with diabetes mellitus who participated in the DM-P4P programme from 2006–2008. Glycosylated haemoglobin results were used to evaluate modifications in therapy in response to poor diabetes control. Prescription patterns were used to assign patients to either a therapeutic inertia group or an intensified treatment group. Therapeutic inertia was defined as the failure to act on a known problem. Results The research sample comprised of 168 876 patients with diabetes mellitus who had undergone 899 135 tests. Of these, 37.4% (336 615 visits) of prescriptions were for a combination of two types of drug and 27.7% (248 788 visits) were for a combination of three types of drug. The proportion of patients in the intensified therapy group who were prescribed more than two types of drug was considerably higher than that in the therapeutic inertia group. Conclusion In many cases in the therapeutic inertia group only a single type of hypoglycaemic drug was prescribed or the dosage remained unchanged.

BACKGROUND Sorafenib is the current standard treatment for advanced hepatocellular carcinoma (HCC). We analyzed national prescription patterns and treatment outcomes of patients who received sorafenib for advanced HCC. PATIENTS AND METHODS We established a nation-wide cohort of patients who started receiving treatment with sorafenib for advanced HCC between August 2012 and July 2013 from the National Health Insurance Research Database of Taiwan and also retrieved demographic and prescription data. The databases of National Death Registry and Taiwan Cancer Registry were used for survival outcomes and cancer diagnosis information, respectively. RESULTS A total of 3,293 patients were enrolled. The median overall survival (OS) and time to treatment discontinuation (TTD) of all patients were 6.8 and 2.6 months, respectively. Upon the first prescription of sorafenib, 58.4% of patients received the standard dose (800 mg/day). Among them, 61.9% had subsequent dose reduction. A total of 41.6% of patients initially received lower than standard doses; 36.1% of them had subsequent dose escalation to 800 mg/day. Being male (odds ratio=1.41; p

Aim To evaluate whether aspirin can reduce the risk of colorectal cancer in people with diabetes. Methods We studied ≥ 30-year-old people with diabetes, included in the Longitudinal Health Insurance Database 2005 in Taiwan, who were treated with hypoglycaemic drugs. We used a time-varying Cox regression model to adjust for immortal time bias and to estimate the adjusted hazard ratio and 95% CI for the association between aspirin use and colorectal cancer occurrence. Results We studied a total of 60 828 people with diabetes (31 176 men and 29 652 women). Their mean (sd) age was 58.72 (13.33) years. A total of 26 494 people were taking aspirin. Aspirin use 3–5 times/week (moderate frequency) for > 5 years (long duration) was found to reduce the risk of colorectal cancer by 46% (hazard ratio 0.54, 95% CI 0.34–0.86). Aspirin use > 5 times/week (high frequency) for 4–5 years (moderate duration) and > 5 years reduced the risk of colorectal cancer by 56 and 68%, respectively (hazard ratio 0.44, 95% CI 0.24–0.80; hazard ratio 0.32, 95% CI 0.20–0.50). Low frequency (≤ 2 times/week) and/or short duration (≤ 3 years) of aspirin use did not reduce the risk of colorectal cancer. Conclusions Aspirin use with high frequency and long duration reduced the risk of colorectal cancer in people with diabetes in a frequency- and duration-dependent manner, whereas low frequency and/or short duration of aspirin use did not. The dosage, frequency and duration of aspirin use that are sufficient to prevent the incidence of colorectal cancer in people with diabetes require further study.

Purpose Long-acting insulin analogues were developed to facilitate consistent glycemic control without excessive hypoglycemia. However, structural modifications of the insulin molecule can alter biological responses and binding characteristics with specific receptors. The aim of this study was to estimate the risk of sight-threatening diabetic retinopathy (STDR) associated with treatment using long-acting insulin analogues compared with intermediate-acting insulin in patients with type 2 diabetes mellitus (T2DM). Methods A retrospective cohort consisting of patients with T2DM aged 20 years of age and older with newly initiated treatment with long-acting insulin analogues (glargine and detemir) and intermediate-acting human insulin was identified from the National Health Insurance database between January 2004 and December 2006 and was subdivided into different cohorts. The risk of the development of STDR was determined by Cox regression models and compared between different cohorts. Findings Of the 46,739 eligible patients, initiators of insulin glargine, insulin detemir, and neutral protamine Hagedorn (NPH) insulin were identified for comparison using propensity-score matching methods. Long-acting insulin analogues were not associated with changed risk for STDR by intention-to-treat and time-varying use approaches between either matched or unmatched cohorts. Implications The strategies that aim at preventing diabetic retinopathy by treating T2DM patients with long-acting insulin analogues remain further prospective studies with longer follow-up period to validate our observations within an appropriate dosage range and to further evaluate the safety of long-acting insulin analogues on reducing the progression of diabetic retinopathy.

Background. A rapid surge of young-female breast cancer (YFBC) has been observed in Taiwan and other East Asian countries. We recently reported that these cases of YFBC, in contrast to their Western counterparts, are predominantly luminal A subtype. YFBC in Asia may have distinct clinicopathological features and outcomes. Methods. Data collected prospectively by participating hospitals were retrieved from the Taiwan Cancer Database. A total of 15,881 women with newly diagnosed stage I–III breast cancer in 2002–2006 were included. The age at diagnosis was categorized into nine 5-year groups (from Results. The rates of stage I, estrogen receptor-positive (ER+), and progesterone receptor-positive breast cancer were higher in the younger patients ( Conclusion. Emerging YFBC in Taiwan is uniquely associated with favorable pathological features and better outcomes and should not be regarded as the mirror image of its Western counterpart.

Concerns about an increased cardiovascular risk with the angiotensin receptor blocker, olmesartan, prompted the current study to examine associations between olmesartan and other angiotensin receptor blockers with overall and cause-specific mortalities. We collected patients who started to use losartan, valsartan, irbesartan, candesartan, telmisartan, and olmesartan between January 1, 2004, and December 31, 2009, from Taiwan’s National Health Insurance claims database. Prescribed drug types, dosage, and other clinical information were collected. Overall mortality and cause-specific mortality were ascertained through linkages with Taiwan’s National Death Registry. Two follow-up analyses, labeled intention-to-treat and as-treated, were conducted. A Cox proportional hazard regression model was used to calculate the hazard ratio (HR) and 95% confidence interval (CI) using losartan as the reference group. A total of 690 463 subjects were included, with a mean follow-up ranging from a low of 2.8 years for olmesartan to a high of 4.1 years for irbesartan. Subjects who began with valsartan had a modest but significantly increased risk of overall mortality (HR, 1.04; 95% CI, 1.02–1.06) compared with losartan. Irbesartan (HR, 0.96; 95% CI, 0.94–0.99), candesartan (HR, 0.95; 95% CI, 0.92–0.99), telmisartan (HR, 0.93; 95% CI, 0.90–0.96), and olmesartan (HR, 0.93; 95% CI, 0.88–0.97) were associated with a slightly lower overall mortality risk than losartan. The analysis indicates that the differences in mortality risk among individual angiotensin receptor blockers were only marginal and thus less likely to be clinically important. Although uncontrolled confounding might still exist, olmesartan does not seem to increase cardiovascular risk compared with losartan.

Background The association between psoriasis and diabetes mellitus (DM) has been explored previously. However, no studies have been reported regarding the severity of psoriasis, comorbidities, and concomitant medications on the risks of DM in patients with psoriasis. Objective We sought to evaluate the impact of the severity of psoriasis, comorbidities, and concomitant medications on the risk of type 2 DM in patients with psoriasis. Methods We conducted a cohort study with 14,158 adults with psoriasis and adults without psoriasis using data from the Taiwan National Health Insurance Research Database. Cox regression models using time-varying covariates were used. Results After the comorbidities and concomitant medications were adjusted for, psoriasis was found to be independently associated with an increased risk of DM (severe: hazard ratio, 2.06 [95% confidence interval, 1.58-2.68] vs mild: hazard ratio, 1.28 [95% confidence interval, 1.05-1.55]). Other independent risk factors included age, Cushing disease, and the increased cumulative doses of the thiazide and methotrexate. Limitation The National Health Insurance Research Database did not have information regarding the Psoriasis Area and Severity Index, diet, obesity, body mass index, exercise status, and family history of diabetes. Conclusion Patients with psoriasis have a higher risk of developing DM. The risks vary depending on the severity of psoriasis, comorbidities, and concomitant medications.

Objective Antidepressant drugs might induce weight gain and increase diabetes risk. We examined the diabetes risk with long-term antidepressant use in a general population. Method This study was a population-based, nested case-control study using Taiwan's National Health Insurance Research Database between 1998 and 2009. A total of 47,885 patients with type 2 diabetes mellitus (ICD-9 codes: 250.x; excluding 250.x 1 and 250.x 3) and 95,770 controls were identified. We used a conditional logistic regression model for data analysis and 1-year latent period before the diabetes diagnosis to account for the quantification of treatment duration of antidepressant (defined by Anatomic Therapeutic Chemical classification code N06A). Sensitivity analyses were performed using a propensity score matching method, as well as different lengths of latent periods. Results Compared with nonusers, patients with cumulative antidepressant use (> 2 years) had an increased risk of diabetes (adjusted OR = 1.20; 95% CI, 1.05-1.37). Moreover, increasing mean daily dose or use of selective serotonin reuptake inhibitors or serotonin antagonist and reuptake inhibitors was associated with increased diabetes risk. The increased diabetes risk with long-term antidepressant therapy in patients aged 44 years or less (adjusted OR = 2.39; 95% CI, 1.46-3.90) was higher than that in older adults (adjusted OR = 1.15; 95% CI, 1.00-1.32). Conclusions The findings suggest that long-term antidepressant use may be associated with an increased risk of type 2 diabetes mellitus, especially for young adults. Therefore, long-term antidepressant use should be evaluated more cautiously for its benefits and the potential risk of diabetes.

Taiwan's Bureau of National Health Insurance launched the National Antiviral Treatment Program (NATP) in 2003 to reimburse patients for antiviral drugs and interferons for chronic hepatitis B and C. The objective was to examine the impact of the NATP on the incidence and mortality due to hepatocellular carcinoma (HCC).The cumulative numbers of NATP participants were retrieved from the National Health Insurance claims database. The national incidence and mortality rates of HCC were obtained from the Taiwan Cancer Registry in each quarter from 1979 to 2009. An interrupted time-series analysis was applied to test the temporal trend change before and after NATP.From 1979 to 1995, the HCC incidence increased in men and women of all age groups. From 2003 to 2009, 31,155 men and 10,769 women received anti-hepatitis B virus therapy, whereas 13,939 men and 10,721 women received anti-hepatitis C virus therapy. The incidence of HCC reached a plateau and then started to decline in men aged 30-39 (slope change P=0.003), 50-59 (P=0.051), and 60-69 years (P0.001). A similar trend was noted in women aged 50-59 (P=0.035), 60-69 (P=0.006), and 70-79 years (P=0.052). The HCC mortality rate had been decreasing since 1996 and a further decline was observed after 2004 in men aged 60-69 years and women aged 60-79 years.There is a strong temporal relationship between NATP and the stabilization of the HCC incidence and related mortality. The cost-effectiveness of the NATP needs further evaluation.

Background Recent studies have shown that use of angiotensin-converting enzyme (ACE) inhibitors may decrease pneumonia risk in various populations. We investigated the effect of ACE inhibitors and angiotensin II receptor blockers (ARBs) on pneumonia hospitalization in the general population of Taiwan. Methods We conducted a case-crossover study using the Taiwan Longitudinal Health Insurance Database for the year 2005. Data from patients hospitalized for the first time for pneumonia during 1997-2007 were analyzed. The case period was defined as the 30 days before admission; the periods 90 to 120 days and 180 to 210 days before admission were used as control periods. Prescribing status of ACE inhibitors and ARBs during the 3 periods was assessed for each patient. Conditional logistic regression was used to estimate the odds ratio (OR) for pneumonia associated with use of ACE inhibitors and ARBs. Results We identified 10 990 cases of hospitalization for new pneumonia. After adjustment for time-variant confounding factors, pneumonia was not associated with use of ACEI or ARBS: the ORs were 0.99 (95% CI, 0.81-1.21) and 0.96 (0.72-1.28), respectively. No association was seen for cumulative defined daily doses (DDDs), as compared with nonusers, for 0 to 30, 31 to 60, or more than 60 DDDs. The results were found to be robust in sensitivity analysis. Conclusions Neither the use nor cumulative dose of ACE inhibitors or ARBs was associated with pneumonia among the Taiwanese general population.

OBJECTIVES Emerging evidence shows that methicillin-resistant Staphylococcus aureus (MRSA) infections caused by isolates with higher vancomycin MICs within the susceptibility range are associated with adverse outcomes. No study, however, has examined different susceptibility tests in predicting treatment outcomes of MRSA infections. METHODS This retrospective cohort study included 393 patients with MRSA bacteraemia. Vancomycin MICs for all MRSA isolates were determined simultaneously by agar dilution and the Etest, and using the MicroScan, VITEK-2 and Phoenix automated systems, and categorized into low- and high-MIC isolates at a breakpoint of ≥ 2 mg/L. The essential and categorical agreement between testing methods was compared. The method-specific ability to predict in-hospital mortality was examined by multivariate logistic regression analysis controlling for other potential confounders using clinical data from 310 vancomycin-treated MRSA bacteraemia patients. RESULTS The agar dilution, Etest, MicroScan, VITEK-2 and Phoenix methods assessed 14.2% (56/393), 9.7% (38/393), 28.8% (113/393), 22.6% (89/393) and 3.1% (12/393) of MRSA isolates as having high (≥ 2 mg/L) vancomycin MICs. The essential and categorical agreement between testing methods ranged from 98.5% to 100% and from 73.8% to 91.9%, respectively. High vancomycin MICs for isolates determined using agar dilution and the Etest independently predicted mortality when controlling for confounding factors [adjusted OR, 2.321; 95% CI, 1.160-4.641; and adjusted OR, 3.121; 95% CI, 1.293-7.536, respectively]. High vancomycin MICs determined using all three automated systems failed to predict mortality. CONCLUSIONS Vancomycin MICs generated by the agar dilution and Etest methods, but not the automated systems, independently predicted mortality among vancomycin-treated MRSA bacteraemia patients. Clinicians should incorporate this information with clinical assessment for decisions on appropriate anti-MRSA treatment.

Object We sought to explore factors associated with persistence of antidepressant treatment in Taiwan and to compare persistence rates across various antidepressants. Method This was a retrospective cohort study using medical claims in Taiwan. We collected data of all new antidepressant users with depressive disorder, aged 18 years or older, during the study period from January 1, 1998, to July 4, 2009. Overall antidepressant treatment persistence was defined as undergoing treatment for 180 days without exceeding a 30-day gap. We also assess the cause of treatment change of initial monotherapy, including discontinuation, switching or combination. The competing risks method was used to estimate cause-specific cumulative incidence and to determine associated factors. Results Only 17.4% of patients continued overall antidepressant treatment and 7.6% continued initial antidepressant monotherapy for 180 days or more. Most patients change initial monotherapy through discontinuation, followed by switching and combination. Male gender, older age, comorbidity with anxiety or sleep disorders, and more concomitant use of drugs were associated with lower discontinuation rate. In terms of antidepressant comparisons, we found that patients treated with selective serotonin reuptake inhibitors are less likely to change initial monotherapy. Conclusion The overall persistence of antidepressant treatment in Taiwan was lower than in other countries.

Emerging literature has documented the effectiveness of intralesional steroid injection as an alternative treatment for benign vocal fold disorders. However, clinical application is frequently limited by the associated technical demands for adequate anesthesia and precise needle placement. This study investigated the applicability and effectiveness of the more practical and less technically demanding method of transnasal endoscopic steroid injection (TESI).Prospective case series.This study recruited 30 patients with vocal nodules and polyps. Dexamethasone was injected into the Reinke's space under local anesthesia via the operating channel of a transnasal flexible laryngoscope in an office setting. Treatment outcome were measured before, 1 month after, and 3 months after the injection, using videolaryngostroboscopy (VLS), maximal phonation time (MPT), 10-item voice handicap index (VHI-10), acoustic analysis, and perceptual evaluation.VLS examinations at 3 months post-treatment demonstrated that vocal lesions of 10 and 19 patients were resolved or reduced, respectively. Objective measurements showed increased MPT and decreased VHI-10 (P.05 and P.01, respectively). Acoustic analysis revealed significant decrease in jitter and shimmer (P.05). Perceptual evaluation using the GRB (grade, roughness, breathiness) scale also showed improved voice quality (P.01). Treatment outcomes were similar between vocal nodules and polyps (P.05). Mild vocal hematoma occurred in three patients following TESI, but resolved spontaneously within 1 month.TESI is a simple and practical office-based treatment modality for benign vocal fold lesions, suitable for most otolaryngologists. Treatment outcomes showed significant subjective and objective improvements that were comparable to the results of other injection procedures reported in the literature.4.

Purpose The epidermal growth factor receptor inhibitors, gefitinib and erlotinib, are used as standard salvage therapy for advanced non-small-cell lung cancer (NSCLC). The aim of the present study was to compare their efficacies in this population. Patients and methods The Taiwan Cancer Registry and the National Health Insurance claim databases were searched for newly diagnosed patients with NSCLC from 2004 to 2007 who received gefitinib or erlotinib as third-line therapy. Overall survival (OS) and time to treatment failure (TTF) were determined from registered parameters. Treatment efficacies were compared by the log-rank test in total population and subsets with different clinical characteristics. The Cox’s proportion hazard model was used to estimate the adjusted hazard ratios in multivariate analyses. Results A total of 984 patients who received gefitinib (67%) or erlotinib (33%) were included. Patients receiving gefitinib or erlotinib had similar OS (median, 10.2 versus 9.9 months, p = 0.524) and TTF (median, 5.5 versus 3.4 months, p = 0.103). In multivariate analyses, both treatment groups had similar risk of overall mortality (adjusted hazard ratio [HR] = 1.04, p = 0.629) and treatment failure (adjusted HR = 0.94, p = 0.417). Comparing the treatments in subgroups based on age, tumour histology and gender also revealed no differences in OS and TTF. For patients who received gefitinib or erlotinib for more than 3 or 6 months, there was no difference in TTF but patients who received erlotinib had longer OS. Conclusions Gefitinib and erlotinib had similar efficacies as salvage therapy for advanced NSCLC in Taiwan.

The objective of this study was to use instrumental variable (IV) analyses to evaluate the clinical effectiveness of percutaneous stent revascularization versus bypass surgery in the treatment of peripheral artery disease (PAD) among type 2 diabetic patients. Type 2 diabetic patients who received peripheral artery bypass surgery (n = 5,652) or stent revascularization (n = 659) for lower extremity arterial stenosis between 2000 and 2007 were identified from the Taiwan National Health Insurance claims database. Patients were followed from the date of index hospitalization for 2 years for lower-extremity amputation, revascularization, and hospitalization for medical treatment. Analysis using treatment year, patients’ monthly income level, and regional difference as IVs were conducted to reduce unobserved treatment selection bias. The crude analysis showed a statistically significant risk reduction in favor of stent placement in lower extremity amputation and in the composite endpoint of amputation, revascularization, or hospitalization for medical treatment. However, peripheral artery stent revascularization and bypass surgery had similar risk of lower limb amputation and composite endpoints in the analyses using calendar year or patients’ monthly income level as IVs. These two treatment modalities had similar risk of lower limb amputation among DM patients with PAD.

OBJECTIVE Data from the National Taiwan Cancer Registry have been widely used since 2002 to assess quality of health, but its quality of coding in cancer staging data has not been discussed. This study assessed the agreement rate for staging by site visit at medical institutes. METHODS In this retrospective chart review study, 392 cancer patients in year 2013 were randomly selected from 14 hospitals; the senior cancer registrar reviewers had compared each original chart with data from the Taiwan Cancer Registry to assess agreement rate for staging. The hospitals were classified into two groups on the basis of the number of cancer patients. The kappa (κ) statistic method and multiple regression analysis were used to compare among the medical institutes and qualified cancer registrars. RESULTS The agreement rate was high in pharynx, esophageal, rectal, breast and prostate cancers, and low in ovarian and other cancers for clinical and pathological staging. After adjustment for the experience of the qualified cancer registrar, low-caseload hospitals had a significantly lower clinical staging agreement rate than that of high-caseload hospitals. After controlling the hospital cancer caseloads the cancer registrar background becomes one of significant factor. That is long duration between a basic license to an advanced license exceeded 5 years, having lower agreement rate. CONCLUSIONS The reliability of staging data in the Taiwan Cancer Registry is affected not only by the cancer type but also by the number of patients treated in hospital. Moreover, the experience of cancer registrar strongly influences agreement rate, especially in clinical staging.

Use of β-blockers (BBs) in patients with chronic obstructive pulmonary disease (COPD) and cardiovascular diseases is supported by increasing evidence. However, most of these studies focused on the survival outcome and used a non-active comparison, prevalent-user design. We aimed to examine the risk of overall death and cardiovascular outcomes associated with use of cardioselective BBs using an active comparison, incident cohort approach. We identified COPD patients initiating cardioselective BBs or non-dihydropyridine calcium channel blockers (CCBs) between 2007 and 2011 in the population-based Taiwan database. A Cox regression model was applied to estimate hazard ratios (HRs) for overall death, cardiovascular death, and cardiovascular events comparing cardioselective BBs and non-dihydropyridine CCBs after propensity score matching. We also conducted sensitivity analyses to quantify the unmeasured confounding effect from COPD severity. A total of 107,902 patients were included. Cardioselective BBs were associated with a modest, lower risk of overall death (HR, 0.85; 95 % CI, 0.81–0.88). The reduced risk of overall death, however, was vulnerable to distribution of COPD severity and was easily weakened with lower prevalence of severe COPD patients in the initiators of cardioselective BBs and higher prevalence of severe COPD patients in the initiators of non-dihydropyridine CCBs. No excess benefit for cardiovascular death (HR, 1.05; 95 % CI, 0.97–1.13) or cardiovascular events (HR, 0.98; 95 % CI, 0.94–1.03) was detected. The present study demonstrated a potential effect of confounding by COPD severity and therefore did not suggest an association between use of cardioselective BB and survival benefit in COPD patients.

Background Opioids are crucial in cancer pain management. We examined the nationwide prescription patterns of opioids in Taiwan cancer patients to find the potential concerns. Methods We reviewed the claims database of the National Health Insurance of Taiwan for patients diagnosed with cancer from 2003 to 2011. The use and cost of analgesics were analyzed. Opioids were classified into recommended strong opioids (morphine and transdermal fentanyl), recommended weak opioids (tramadol, buprenorphine, and codeine), and unrecommended opioids (propoxyphene, nalbuphine, and meperidine). Results We enrolled 1,424,048 patients with cancer, and ∼50% of them took analgesics. Among analgesic users, patients who used opioids increased from 48.2% in 2003 to 52.0% in 2010. Approximately 92% of the opioid use came from recommended opioids, either strong (51%) or weak opioids (41%). The ratio of the use of short-acting strong opioids to that of long-acting opioids increased from 0.41 in 2003 to 0.63 in 2011. Transdermal fentanyl accounted for > 50% of the use of strong opioids. Among weak opioids, the use of tramadol gradually increased to 71% in 2011. On average, opioids contributed to 0.79‰ of all medical expenditures and 2.94‰ of all medication costs. Conclusion The use of short-acting strong opioids increased during the study period. Instead of oral opioids, transdermal fentanyl was the most commonly used opioid among Taiwan cancer patients. The use of weak opioids, particularly tramadol, was high. These concerns should be the focus of pain management education.

Background Saxagliptin has been reported to be associated with an increased risk of hospitalization for heart failure (HF). The objective of this study was to test whether the increased risk is drug specific or a class effect for dipeptidyl peptidase-4 (DPP-4) inhibitors. Methods Diabetic patients prescribed sitagliptin, saxagliptin, and vildagliptin between 2011 and 2013 were identified from Taiwan's National Health Insurance (NHI) claims database. The outcome of interest was the first hospitalization for HF. The patients were followed for one year from drug initiation to outcome occurrence, death, or study termination (December 31, 2013). A Cox proportional hazards regression model was used to calculate the hazard ratios (HR) and their 95% confidence intervals, using sitagliptin as the reference group. Results A total of 239,669 patients, including 159,330 sitagliptin, 38,561 saxagliptin, and 41,778 vildagliptin initiators, were included in the analysis. With a follow-up period ranging from 269days (vildagliptin) to 313days (sitagliptin), the crude incidence rate of HF was 2.77, 2.63, and 1.91 per 100 person-years for sitagliptin, saxagliptin, and vildagliptin, respectively. Saxagliptin had a similar risk (HR: 0.98, 95% CI: 0.91–1.06) to sitagliptin, while vildagliptin was associated with a lower risk of HF (HR: 0.85, 95% CI: 0.78–0.93). Auxiliary analyses using acarbose (n=130,800) as a reference group consistently showed no increased risk of HF associated with DDP-4 inhibitors. Conclusion Three DPP-4 inhibitors studied seem to be safe regarding the risk of HF, while the reduced risk of vildagliptin might be a spurious association or a chance finding.

Supplemental Digital Content is available in the text, To analyze the association between use of DPP-4 inhibitors and acute pancreatitis in high-risk type 2 diabetic patients. A retrospective nationwide cohort study was conducted using the Taiwan National Health Insurance claim database. The risk associated with sitagliptin was compared to that with acarbose, a second-line antidiabetic drug prescribed for patients with similar diabetes severity and with a known neutral effect on pancreatitis. Between January 1, 2009 and December 31, 2010, a total of 8526 sitagliptin initiators and 8055 acarbose initiators who had hypertriglyceridemia or prior hospitalization history for acute pancreatitis were analyzed for the risk of hospitalization due to acute pancreatitis stratified for baseline propensity score. In the crude analysis, sitagliptin was associated with a decreased risk of acute pancreatitis (hazard ratio [HR] 0.74; 95% confidence interval [CI]: 0.62–0.88) compared to acarbose in diabetic patients with prior history of hospitalization for pancreatitis or hypertriglyceridemia. The association was abolished after stratification for propensity score quintiles (adjusted HR 0.95; 95% CI: 0.79–1.16). Similar results were found separately in both patients’ histories of prior hospitalization of acute pancreatitis (adjusted HR 0.97; 95% CI: 0.76–1.24) and those with hypertriglyceridemia (adjusted HR 0.86; 95% CI: 0.65–1.13). No significant association was found for different durations or accumulative doses of sitagliptin. In the stratified analysis, no significant effect modification was found in relation to patients’ characteristics. Use of sitagliptin was not associated with an increased risk of acute pancreatitis in high-risk diabetic patients with hypertriglyceridemia or with history of acute pancreatitis.

The study aimed to investigate the association of long-term use of different antihypertensive agents with incident breast cancer.A total of 794 ,533 women aged at least 55 years were identified from Taiwan National Health Insurance claims database during 2001-2011. As of 31 December 2011, incident breast cancer patients were included as cases, and 1 : 4 age-matched controls were selected by risk-set sampling scheme. Logistic regression models were applied to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) of breast cancer incidence associated with different durations of use of angiotensin-converting enzyme inhibitors/angiotensin receptor blockers, β-blockers, and dihydropyridine calcium channel blockers (DHP CCBs). Different restriction rules were applied to reveal the potential effects of confounding by indication.Among the 9397 incident breast cancer patients and 37 ,588 controls, a significantly elevated risk was found for relatively short-term use of DHP CCBs (6 years) but not in those observed for more than 6 years. There was no association between either angiotensin-converting enzyme inhibitors/angiotensin receptor blockers or β-blockers use and breast cancer. Although restricting our analyses to those with any prescription of antihypertensive medications in 2001 or those with diagnosis of hypertension, there was no longer a statistically significant association between any use of DHP CCBs and breast cancer (OR: 1.21, 95% CI: 0.88-1.67 for the former, and OR: 1.71, 95% CI: 0.99-2.95 for the latter).The results demonstrated the potential effect of confounding by indication, and thus, did not suggest any association of the use of antihypertensive medication and breast cancer risk.

Background. Recent evidence has shown that community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) is less virulent than traditional hospital-associated MRSA. We explored whether the antimicrobial susceptibilities of the different strains account for their disparity in clinical virulence. Methods. This 10-year retrospective cohort study enrolled 291 patients with community-onset, healthcareassociated MRSA bacteremia. The vancomycin minimum inhibitory concentration (MIC) and staphylococcal cassette chromosome mec (SCCmec) type were determined for all isolates. CA-MRSA was defined as an isolate possessing the SCCmec type IV or V genes, and hospital-associated MRSA (HA-MRSA) was defined as an isolate possessing SCCmec type I, II, or III genes. Low and high vancomycin MICs were defined as MICs of ≤1 and ≥2 μg/mL, respectively. Patients with bacteremia due to CA-MRSA with a low vancomycin MIC (n = 111), due to HA-MRSA with a low vancomycin MIC (n = 127), or due to HA-MRSA with a high vancomycin MIC (n = 47) entered the outcome analysis. The outcomes of the 2 HA-MRSA bacteremia groups were compared to those of the CA-MRSA bacteremia group. Results. Treatment failure was observed in 35 (31.5%), 59 (46.5%), and 27 (57.4%) of patients with low-vancomycin-MIC CA-MRSA, low-vancomycin-MIC HA-MRSA, and high-vancomycin-MIC HA-MRSA bacteremia, respectively. After adjustment for potential confounding factors, the risk of treatment failure was significantly higher among patients with low-vancomycin-MIC HA-MRSA (adjusted odds ratio [aOR], 1.853; 95% confidence interval [CI], 1.006–3.413) and high-vancomycin-MIC HA-MRSA (aOR, 2.393; 95% CI, 1.079–5.309), compared with patients with low-vancomycin-MIC CA-MRSA. Conclusions. The higher risk for treatment failure among patients with traditional hospital-associated MRSA infections, compared with patients with CA-MRSA infections, is independent of the vancomycin MIC, suggesting a potential intrinsic strain-specific virulence effect.

Background. Asian ethnicity is associated with a distinct molecular etiology, treatment response, and survival outcome among patients with non-small cell lung cancer (NSCLC). This study examines the survival impact of platinum-based adjuvant chemotherapy for Asian patients with stage I–IIIA NSCLC. Methods. This study recruited patients aged ≥18 years with histologically proven stage IA–IIIA NSCLC registered in the Taiwan Cancer Registry database in January 2004 to December 2007. Platinum-containing adjuvant chemotherapy had to be started within 90 days of the primary surgery. Kaplan–Meier survival curves, log-rank tests, and the Cox proportional hazards regression model were used to assess the influence of various risk factors on survival time. Results. This study included 2,231 patients with stage IA–IIIA NSCLC who underwent primary surgery with a clear surgical margin. The percentages of all causes of death were significantly lower for the chemotherapy group for both stage II and stage IIIA patients. Multivariate analysis identified platinum-based adjuvant chemotherapy as an independent prognostic factor for the overall survival outcome of stage II (hazard ratio [HR], 0.61; 95% confidence interval [CI], 0.39–0.94; p = .024) and IIIA (HR, 0.71; 95% CI, 0.52–0.96; p = .029) patients. Among these patients, those who received adjuvant chemotherapy had a superior overall survival outcome for both genders, for the subgroup of patients aged ≥70 years, and for those with adenocarcinoma. Conclusion. Platinum-based adjuvant chemotherapy should be considered in the treatment plan for Asian patients with resected stage II and stage IIIA NSCLC.

Background. Diabetes mellitus (DM) has been implicated in influencing the survival duration of patients with breast cancer. However, less is known about the impact of DM and other comorbidities on the breast cancer–specific survival (BCS) and overall survival (OS) outcomes of Asian patients with early-stage breast cancer. Patients and Methods. The characteristics of female patients with newly diagnosed, early-stage breast cancer were collected from the Taiwan Cancer Registry database for 2003–2004. DM status and other comorbidities were retrieved from Taiwan's National Health Insurance database. The BCS and OS times of patients according to DM status were estimated via the Kaplan–Meier method. Cox's proportional hazard model was used to estimate adjusted hazard ratios (HRs) for the effects of DM, comorbidities, and other risk factors on mortality. Results. In total, 4,390 patients were identified and 341 (7.7%) presented with DM. The 5-year BCS and OS rates were significantly greater in DM patients than in non-DM patients (BCS, 85% versus 91%; OS, 79% versus 90%). Furthermore, after adjusting for clinicopathologic variables and comorbidities, DM remained an independent predictor of shorter BCS (adjusted HR, 1.53) and OS (adjusted HR, 1.71) times. Subgroup analyses also demonstrated a consistent prognostic influence of DM across different groups. Conclusion. In Asian patients with early-stage breast cancer, DM is an independent predictor of lower BCS and OS rates, even after adjusting for other comorbidities. The integration of DM care as part of the continuum of care for early-stage breast cancer should be emphasized.

Background/Aims The upper gastrointestinal (GI) toxicity associated with non-steroidal anti-inflammatory drugs (NSAID) use among cirrhotic patients remains unclear. The objective of this study was to evaluate the risk of upper GI adverse events associated with celecoxib and oral and parenteral non-selective NSAIDs in cirrhotic patients. Methods All the patients aged ≥ 20 years with a diagnosis of cirrhosis hospitalized for variceal bleeding and non-variceal upper GI adverse events (oesophageal, gastric, duodenal ulcer, bleeding; gastritis and duodenitis) in 2006 were identified using ICD-9-CM diagnosis codes from inpatient claims from the Taiwan National Health Insurance Database. In the case-crossover study design, the case period was defined as 1–30 days and the control period as 31–60 days before the date of hospitalization. The information for individual NSAID use was obtained from the outpatient pharmacy prescription database. Adjusted self-matched odds ratios (OR) and their 95% confidence intervals (CI) were estimated with a conditional logistic regression model. Results A total of 4876 cirrhotic patients were included. The adjusted OR (95% CI) was 1.44 (0.89–2.31) for celecoxib, 1.87 (1.66–2.11) for oral non-selective NSAIDs and 1.90 (1.55–2.32) for parenteral NSAIDs overall. Risks were similar for variceal and non-variceal events. Concomitant use of proton pump inhibitors and histamine-2 receptor antagonists tended to decrease the upper GI toxicity associated with non-selective NSAIDs and celecoxib. Conclusion The use of nsNSAIDs but not celecoxib was associated with a two-fold increased risk of variceal and non-variceal upper GI events among cirrhotic patients.

A concise and reliable composite quality score would be helpful in judging the quality of a hospital’s services, especially for pay-for-performance (P4P) initiatives. This study compared several nonlatent and latent composite quality scores to evaluate the quality of care using diabetes mellitus (DM) P4P data and discusses their characteristics and implications for P4P policy. The authors describe a cross-sectional study of the DM P4P data collected from the claims data of the Bureau of National Health Insurance (NHI) in Taiwan from January 2007 to December 2007. The DM patient outcome data, such as hemoglobin A1C values, were retrieved from the P4P database sponsored by the Bureau of NHI in Taiwan. The composite scores were derived from the following methods: 1) nonlatent scores methods (e.g., the raw sum score and the all-or-none score methods)and 2) latent scores methods (e.g., item-response theory-based Models I and II and the PRIDIT model). These scores are compared in terms of 2 aspects—agreement of hospital rankings (using Spearman’s rank correlation) and reliability (using bootstrap methods). The latent methods were superior to the nonlatent methods because they were more reliable and had specific weighting themes. The correlations among the 3 latent methods were moderately high. The use of the PRIDIT approach, which is moderately difficult compared with item response theory–based model, is recommended if the insurer wants to balance convenience and precision.

The incidence of and risk factors for acquiring community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) among patients staying in intensive care units (ICUs) remain unclear. We enrolled patients staying in two ICUs at the Far Eastern Memorial Hospital during the period of 1 September 2008 to 30 September 2009 to clarify this issue. Surveillance cultures for MRSA were taken from nostril, sputum or throat, axillae, and the inguinal area in all enrolled patients upon admission to the ICU, every 3 days thereafter, and on the day of discharge from the ICU. For each MRSA isolate, we performed multilocus sequence typing, identified the type of staphylococcal cassette chromosome mec , detected the presence of the Panton-Valentine leukocidin gene, and conducted drug susceptibility tests. Among the 1,906 patients who were screened, 203 patients were carriers of MRSA before their admission to the ICU; 81 patients acquired MRSA during their stay in the ICU, including 31 who acquired CA-MRSA. The incidence rates of newly acquired MRSA and CA-MRSA during the ICU stay were 7.9 and 3.0 per 1,000 patient-days, respectively. Prior usage of antipseudomonal penicillins and antifungals and the presence of a nasogastric tube were found to be independent risk factors for acquiring CA-MRSA during the ICU stay when data for CA-MRSA carriers and patients without carriage of MRSA were compared ( P = 0.0035, 0.0330, and 0.0262, respectively). Prior usage of carbapenems was found to be a protective factor against acquiring CA-MRSA when data for patients with CA-MRSA and those with health care-associated MRSA acquired during ICU stay were compared ( P = 0.0240).

The difference in the outcomes of nosocomial bloodstream infection (BSI) caused by community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) strains and healthcare-associated MRSA (HA-MRSA) strains remains unclear.From January 1, 2006 to December 31, 2008, all adult patients hospitalized at National Taiwan University Hospital with nosocomial MRSA BSI were analyzed. Available MRSA isolates were submitted for subsequent microbiologic studies to determine whether they belonged to CA-MRSA strains.In total, 308 patients were enrolled and 253 MRSA isolates were available. Forty-seven isolates belonged to CA-MRSA strains. The all-cause mortality rates on Day 14 and Day 30 were 19.8% and 30.5%, respectively, and were not different between those caused by CA-MRSA and HA-MRSA strains. The independent risk factors for Day 14 mortality were septic shock, thrombocytopenia, and an inadequate serum trough level of vancomycin (p =0.0001, 0.0003, and 0.0381, respectively). Those for Day 30 mortality were septic shock, anemia, thrombocytopenia, presence of underlying malignancies, and MRSA isolates with a vancomycin minimum inhibitory concentration of 2 mg/L (p =0.0001, 0.0425, 0.0007, 0.0098, and 0.0012, respectively).The mortality rates of nosocomial MRSA BSI were not different between that caused by CA-MRSA and HA-MRSA strains.

OBJECTIVE To compare the efficacy of risperidone and olanzapine in schizophrenic patients with tardive dyskinesia on treatment with first-generation antipsychotics. METHOD We conducted a 24-week, rater-blinded, flexible-dose study. Sixty patients with DSM-IV schizophrenia (n = 58) or schizoaffective disorder (n = 2) met the DSM-IV research criteria for neuroleptic-induced tardive dyskinesia and were randomly assigned to a risperidone or olanzapine group. The primary outcome was a comparison of the change in the total scores on the Abnormal Involuntary Movement Scale (AIMS) from baseline to study end point between the groups. The study was conducted from July 2000 to June 2004. RESULTS The mean ± SD doses of risperidone and olanzapine from baseline to study end point were 1.9 ± 0.7 to 4.1 ± 1.4 mg/d and 8.1 ± 2.0 to 12.6 ± 5.4 mg/d, respectively. There were no statistically significant differences in demographic data, severity of tardive dyskinesia, or psychotic symptoms between risperidone and olanzapine groups at baseline assessment. Both groups showed significant improvement in mean ± SD AIMS total scores (risperidone: −7.4 ± 6.9, P