Multiple Cox Proportional Hazards Models Research Articles

Mortality and complications after hip fracture among elderly patients undergoing hemodialysis.

BackgroundOsteoporotic hip fractures cause high mortality and morbidity in elderly adults. Compared to the general population, subjects with end-stage renal disease and hemodialysis often develop mineral bone disorders and have a higher risk for hip fractures.MethodsWe conducted a matched cohort study design and used competing risk analysis to estimate the cumulative incidence of the complication rate. Subjects aged greater than 60 years with hip fracture were selected from Taiwan’s National Health Insurance Research Database covering a period from 1997 to 2007, and these subjects were followed up until 2009. We used the Kaplan–Meier method to estimate the overall survival and used the log-rank test and multiple Cox proportional hazards model to explore the risk factors for survival. The cumulative incidence of the first complication was estimated using competing risk analysis.ResultsAmong hemodialysis subjects, the three-month, one-year, two-year and five-year mortality rates were 17.3 %, 37.2 %, 51.5 %, and 80.5 %, respectively; the one-year and five-year cumulative incidences of the first surgical complication were 14.2 % and 20.6 %, respectively; and the three-month cumulative incidence of the first medical complication was 24.1 %. Hemodialysis subjects presented a 2.32 times (95 % CI: 2.16–2.49) higher hazard ratio of overall death, 1.15 times (95 % CI: 1.01–1.30) higher sub-hazard ratio (sub-HR) of surgical complications, and 1.35 times (95 % CI: 1.21–1.52) higher sub-HR of the first medical complication than non-hemodialysis controls.ConclusionsThe overall mortality and complication rates of hemodialysis subjects after surgery for hip fracture were significantly higher than those of non-hemodialysis subjects. Further prospective studies which include important risk factors are necessary to more precisely quantify the adjusted effect of hemodialysis.Electronic supplementary materialThe online version of this article (doi:10.1186/s12882-015-0099-0) contains supplementary material, which is available to authorized users.

115 The Impact of Smoking on Mortality Following Percutaneous Coronary Intervention-an Observational Study

<h3>Aim</h3> To assess the impact of smoking on mortality following percutaneous coronary intervention (PCI). <h3>Methods</h3> A total of 13041 patients underwent PCI between 2008 and 2011 at two large tertiary centres. Data on smoking status were missing in 385 patients. The remaining 12656 patients were included in this analysis. Patients were divided into three groups according to their smoking status at the time of index PCI procedure: non- smokers (n = 4288, 33.9%), ex-smokers (n = 4806, 38.0%) and current smokers (n = 3562, 28.1%). <h3>Results</h3> Current smokers underwent PCI at a younger age (mean ± SD) and more frequently for acute coronary syndrome (ACS) (57 ± 11 years, 84.1% for ACS) compared to non-smokers (67 ± 12 years, 62.9%) and ex-smokers (67 ± 11 years, 57.0%), p &lt; 0.0001 for both. Overall 30-day mortality rate was (1.8%). In a logistic regression model adjusted for several confounders, the adjusted odds ratios (95% confidence intervals [CI]) for 30-day mortality were 0.98 (0.70–1.38) in ex-smokers and 1.60 (1.10–2.32) in current smokers compared to non-smokers. Overall mortality rate between 30 days and 6 months post PCI was 1.5%. In the multiple Cox proportional hazards model, the adjusted hazard ratios (95% CI) for mortality between 30 days and 6 months were 1.19 (0.84–1.67) in ex-smokers and 1.03 (0.65–1.65) in current smokers compared to non-smokers. <h3>Conclusions</h3> This large observational study of unselected, real-world patients demonstrates a younger age at presentation for PCI and increased 30-day post-PCI mortality in current smokers compared to ex-smokers and non-smokers. Ex-smokers have a similar age at presentation and post PCI mortality outcome as non-smokers. These findings further re-enforce the public health message regarding the harmful effect of smoking and the benefits of smoking cessation in patients undergoing PCI.

Impact of bariatric surgery on life expectancy in severely obese patients with diabetes: a decision analysis.

To create a decision analytic model to estimate the balance between treatment risks and benefits for severely obese patients with diabetes. Bariatric surgery leads to many desirable metabolic changes, but long-term impact of bariatric surgery on life expectancy in patients with diabetes has not yet been quantified. We developed a Markov state transition model with multiple Cox proportional hazards models and logistic regression models as inputs to compare bariatric surgery versus no surgical treatment for severely obese diabetic patients. The model is informed by data from 3 large cohorts: (1) 159,000 severely obese diabetic patients (4185 had bariatric surgery) from 3 HMO Research Network sites; (2) 23,000 subjects from the Nationwide Inpatient Sample; and (3) 18,000 subjects from the National Health Interview Survey linked to the National Death Index. In our main analyses, we found that a 45-year-old woman with diabetes and a body mass index (BMI) of 45 kg/m gained an additional 6.7 years of life expectancy with bariatric surgery (38.4 years with surgery vs 31.7 years without surgery). Sensitivity analyses revealed that the gain in life expectancy decreased with increasing BMI, until a BMI of 62 kg/m is reached, at which point nonsurgical treatment was associated with greater life expectancy. Similar results were seen for both men and women in all age groups. For most severely obese patients with diabetes, bariatric surgery seems to improve life expectancy; however, surgery may reduce life expectancy for the super obese with BMIs over 62 kg/m.

The Risk Factors and the Impact of Hernia Development on Technique Survival in Peritoneal Dialysis Patients: A Population-Based Cohort Study

There is a lack of consensus on the risk factors for hernia formation, and the impact on peritoneal dialysis (PD) survival has seldom been studied. This was a population-based study and all collected data were retrieved from the National Health Insurance Research Database of Taiwan. Patients who commenced PD between January 1998 and December 2006 were screened for inclusion. Multiple logistic regression and Cox proportional hazards models were applied to estimate the predictors for hernia formation and determine the predictors of PD withdrawal. A total of 6,928 PD patients were enrolled and followed until December 2009, with 631 hernia events and 391 hernioplasties being registered in 530 patients (7.7%). The incidence rate was 0.04 hernias/patient/year. Longer PD duration (per 1 month increase, hazard ratio (HR) 1.019) and history of mitral valve prolapse (MVP) (HR 1.584) were independent risk factors for hernia formation during PD, and female gender (HR 0.617) was a protective factor. On the other hand, there were 4,468 PD withdrawals, with cumulative incidence rates of 41% at 1 year, 66% at 3 years, and 82% at 5 years. Independent determinants for cumulative PD withdrawal included hernia formation during PD (HR 1.154), age (per 1 year increase, HR 1.014), larger dialysate volume (per 1 liter increase, HR 0.496), female gender (HR 0.763), heart failure (HR 1.092), hypertension (HR 1.207), myocardial infarction (HR 1.292), chronic obstructive pulmonary disease (COPD) (HR 1.227), cerebrovascular accident (CVA) (HR 1.364), and history of MVP (HR 0.712) CONCLUSIONS: Prolonged PD duration was a risk factor for hernia formation, and female gender was protective. Hernia formation during PD therapy may increase the risk of PD withdrawal.

Preoperative glycemic control status as a significant predictor of biochemical recurrence in prostate cancer patients after radical prostatectomy.

BackgroundThe effect of diabetes mellitus (DM) on prostate cancer (PCa) outcome remains controversial. Thus, we investigated the association of DM history, glycemic control, and metformin use with oncologic outcomes after radical prostatectomy (RP).MethodsWe reviewed the records of 746 contemporary patients who had hemoglobin A1c (HbA1c) measured within the 6 months preceding RP. The associations between clinical variables and risk of adverse pathological features and biochemical recurrence (BCR) were tested using a multivariate logistic regression and multiple Cox-proportional hazards model, respectively. BCR was defined as prostatic specific antigen (PSA) > 0.2 ng/mL in 2 consecutive tests.ResultsThere were no significant differences in the rates of adverse pathologic features and BCR-free survival between patients with (n = 209) and without (n = 537) a history of DM diagnosis (all p > 0.05). In multivariate analyses, high HbA1c level (≥ 6.5%) was significantly related with high pathologic Gleason score (≥ 4+3; odds ratio [OR] 1.704, p = 0.019) and BCR-free survival (OR 1.853, p = 0.007). Metformin use was not associated with BCR-free survival (OR 0.662, p = 0.125).ConclusionsPoor glycemic control was significantly associated with BCR after RP. Meanwhile, metformin use was not associated with biochemical outcome after RP. Further investigation would be needed to identify exact mechanism underlying the impact of glycemic control on PCa treatment outcome.

A Diffuse Large B-Cell Lymphoma Classification System That Associates Normal B-Cell Subset Phenotypes with Prognosis

▪BackgroundRecent findings have introduced biological classification of non-Hodgkin lymphomas as exemplified by the “activated B-cell-like” (ABC) and “germinal-center B-cell–like” (GCB) subgroups of diffuse large B-cell lymphoma (DLBCL).AimsThe goal is to generate a refined cell of origin (COO) classification that includes B-cell subset associated gene signatures (BAGS) from the normal B-cell hierarchy.MethodsWe have combined fluorescence activated cell sorting and gene expression profiling by Gene Chip HG U133 Plus 2.0 to generate five BAGS for naïve, germinal centrocytes and centroblasts, post germinal memory B-cells, and plasmablasts of normal human tonsils. Clinical data sets are the Aalborg Project CHEPRETRO (N=89), the Lymphoma/Leukemia Molecular Profiling Project LLMPP (N=233), the International DLBCL Rituximab-CHOP Consortium MD Anderson Project IDRC (N=460), the Mayo Clinic, Brigham & Women Hospital, and Dana-Farber Cancer Institute Project MDFCI (N=88) available on the GEO website. All statistical analyses were done with R version 3.0.1 and full documentation is provided by a Sweave document.ResultsFirst, we verified the quality of the sampled B-cell subsets based on the expression patterns of differentiation molecules, transcription factors, and genes matching biological knowledge. Next, we constructed a BAGS-classifier provided by 77-115 gene probe sets, capable of assigning samples to each of the five COO subtypes.Second, we assigned individual lymphoma cases in 5 patient cohorts including a total of 1063 patient. BAGS identified COO subtypes with frequencies of 2-7 % naïve (N), 35-41 % centrocytes (CC), 18-21 % centroblasts (CB), 4-15 % memory (M), 12-18 % plasmablast (PB), and 15-16 % unclassified (UC) subtypes. The frequencies was not different between cohorts (p=0.41).Third, the BAGS subtypes was associated significantly with overall survival and time to progression for R-CHOP–treated patients in clinical cohorts from the LLMPPN (p=0.0441/0.0358) and the IDRC Program (p=0.002/8e-04).Fourth, we found a very high fraction of GCB samples to be of CC or CB subtypes. On the contrary, a major fraction of BAGS-unclassified subtypes were of the ABC class. In a multiple Cox proportional hazards model we identified the BAGS subtypes to be a prognostic variables independent of ABC/GCB subtypes but not of IPI and age. The most significant impact was observed within the GCB subclass, where the GCB-CC subtype had superior prognosis compared to the GCB-CB subtype, in accordance with individual assignments for drug specific sensitivity to hydroxydaunomycin and vincristine.Fifth, we performed genetic evaluation of the BAGS subtypes by mutation analysis within the CHEPRETRO cohort for EZH2, CD79B, and MYD88 identifying frequencies of 6.3%, 10.1% and 14.7%, respectively. The EZH2 mutation was only identified in the GCBN-CC and -CB subtypes. Mutations of CD79B and MYD88 were preferential in ABC, present in all subtypes. The CC subtype had high p53 mutation and indel frequencies, whereas the CB subtype had high Chr12q15 amplification frequencies and a complex genotype. Finally, the CC subtype expressed LMO2, NF-κB targets, CD58, Stroma1, and MHCII genes, known to have prognostic impact.SummaryIn summary, this study addresses an unmet medical need for a new diagnostic platform for individual DLBCL classification of “cell of origin” phenotyping attempting to design new strategies and more effective disease management. DisclosuresNo relevant conflicts of interest to declare.

Insurance Status Impacts Survival in Follicular Lymphoma

Introduction: Follicular lymphoma (FL) is the most common indolent non-Hodgkin lymphoma. Though no agreed upon standard of care has been established, the incorporation of immunotherapy has been widely adopted given a number of studies demonstrating improvement in response rates, progression-free survival (PFS), and overall survival (OS) with the addition of rituximab. The most common frontline management strategy in the United States (US) for FL is chemoimmunotherapy. However, there is significant heterogeneity in practice patterns, and the relationships between insurance status, chemoimmunotherapy use, and survival in patients (pts) with FL across the US remain unclear. We utilized a national cohort of adult pts with long-term follow-up of FL to test the hypotheses that uninsured patients are less likely to receive chemoimmunotherapy and more likely to experience poorer OS even after controlling for socieodemographic factors, clinical factors, and treatment.Methods: We used data from the National Cancer Data Base (NCDB), a nationwide oncology outcomes database jointly sponsored by the Commission on Cancer (CoC) of the American College of Surgeons and the American Cancer Society. Begun in 1989, NCDB currently contains approximately 29 million records from hospital cancer registries covered by more than 1,500 CoC-accredited cancer programs in US and Puerto Rico. Approximately 75% of all newly diagnosed cancer cases in the US are captured. Data collection by CoC-accredited cancer program uses national standardization, and vital status and date of death are reported to the NCDB on a yearly basis.Pts with FL were identified using the International Classification Oncology (ICD-O) histology codes 9690, 9691, 9695, 9698. The population was restricted to pts diagnosed in 2004, adults age 18-101, and those who received all/part of their first course of treatment at the reporting facility. Cases with missing data (age, gender, race, region, insurance data, treatment) and those with non-Medicare/Medicaid government insurance (VA, Indian Health Services) were removed. Pts with HIV, ≥65 years of age, or aged <65 years whose primary insurance was Medicare were also removed. All information was captured using standardized codes defined by the Facility Oncology Registry Data Standards (FORDS). Chi-square tests were used to compare clinical characteristics by insurance status. Log-binomial models were fitted to examine the association of insurance status with clinical factors including advanced stage (III/IV), presence of B-symptoms, comorbidities, and receipt of chemoimmunotherapy. Kaplan-Meier survival curves and log-rank tests were performed. Multiple variable Cox proportional hazards models examined the contribution of prognostic factors to survival disparity by insurance status.Results: 2,703 pts diagnosed in 2004 in the NCDB met inclusion criteria. Compared with privately insured pts, uninsured or Medicaid insured pts were more likely to have advanced stage disease and B symptoms, but no significant difference in the presence of comorbidities. Compared with privately insured pts, there were no differences in chemoimmunotherapy use for uninsured patients, but Medicaid insured pts were more likely to receive chemoimmunotherapy (Medicaid risk ratio [RR] 1.41; 95% confidence interval [CI] 1.09-1.82). Medicaid and uninsured pts had a significantly higher risk of death compared with privately insured pts even after controlling for socieodemographic, prognostic, and treatment factors (uninsured hazard ratio [HR] 1.78; 95% CI 1.24-2.54; Medicaid HR 2.50; 95% CI 1.84-3.39). Disease stage, age, sex, residing in an area with > 20% of residents not graduating from high school, the presence of B symptoms, and comorbidities, but not treatment were significant, independent predictors of survival in pts with FL and contributed to the disparities in survival by insurance status.Conclusions: Although uninsured and Medicaid insured pts with FL in the US did not have lower use of chemoimmunotherapy, both groups had inferior survival compared to privately insured pts. Despite heterogeneity in management strategies for FL, access to adequate health care has the potential to improve survival for pts with FL. [Display omitted] DisclosuresNastoupil:Genentech: Honoraria; Celgene: Honoraria; TG Therapeutics: Research Funding; Janssen: Research Funding.

The Model for End-stage Liver Disease score is potentially a useful predictor of hyperkalemia occurrence among hospitalized angiotensin receptor blocker users.

Angiotensin receptor blockers (ARBs) are medications commonly used for treating conditions such as hypertension. However, ARBs are frequently associated with hyperkalemia, a potentially critical adverse event, in high-risk patients. Although both the liver and the kidney are major elimination routes of ARBs, the relationship between hepatorenal function and ARB-related hyperkalemia has not yet been investigated. The purpose of this study was to evaluate the risk of hyperkalemia, in terms of various hepatorenal functions, for hospitalized patients newly initiated on ARB treatment. We evaluated ARB-related hyperkalemia in a cohort of 5530 hospitalized patients, who had not previously used ARBs, between 12 April 2004 and 31 May 2012. Hepatorenal function was assessed by the Model for End-stage Liver Disease (MELD) score. Hyperkalemia risk was assessed by hepatorenal function, risks were categorized into the four MELD scoring groups, and the groups were compared with one another. The MELD score was significantly different between the hyperkalemic and non-hyperkalemic groups (independent t-test, P < 0.001). The MELD score 10-14, 15-19 and ≥ 20 groups showed higher risks of hyperkalemia than the lowest MELD score group {log-rank test, P < 0.001; multiple Cox proportional hazard model, hazard ratios 1.478 (P = 0.003), 2.285 (P < 0.001) and 3.024 (P < 0.001), respectively}. The MELD score showed a stronger predictive performance for hyperkalemia than either serum creatinine or estimated glomerular filtration rate alone. Furthermore, the MELD score showed good predictive performance for ARB-related hyperkalemia among hospitalized patients. The clinical implications and reasons for these findings merit future investigation.

Trends and outcome predictors after traumatic brain injury surgery: a nationwide population-based study in Taiwan.

The authors sought to analyze trends in hospital resource utilization and mortality rates in a population of patients who had received traumatic brain injury (TBI) surgery. This nationwide population-based cohort study retrospectively analyzed 18,286 patients who had received surgical treatment for TBI between 1998 and 2010. The multiple linear regression model and Cox proportional hazards model were used for multivariate assessment of outcome predictors. The prevalence rate of surgical treatment for patients with TBI gradually but significantly (p < 0.001) increased by 47.6% from 5.0 per 100,000 persons in 1998 to 7.4 per 100,000 persons in 2010. Age, sex, Deyo-Charlson comorbidity index score, hospital volume, and surgeon volume were significantly associated with TBI surgery outcomes (p < 0.05). Over the 12-year period analyzed, the estimated mean hospital treatment cost increased 19.06%, whereas the in-hospital mortality rate decreased 10.9%. The estimated mean time of overall survival after TBI surgery (± SD) was 83.0 ± 4.2 months, and the overall in-hospital and 1-, 3-, and 5-year survival rates were 74.5%, 67.3%, 61.1%, and 57.8%, respectively. These data reveal an increased prevalence of TBI, especially in older patients, and an increased hospital treatment cost but a decreased in-hospital mortality rate. Health care providers and patients should recognize that attributes of the patient and of the hospital may affect hospital resource utilization and the mortality rate. These results are relevant not only to other countries with similar population sizes but also to countries with larger populations.

The effect of national hospital accreditation in bariatric surgery on perioperative outcomes and long-term mortality

BackgroundControversy exists regarding the relevance of Center of Excellence accreditation to bariatric surgery outcomes. The objective of this study was to evaluate the impact of national hospital accreditation on perioperative and long-term outcomes following bariatric surgery. MethodsRetrospective, longitudinal study using 2004–2010 data from the New York Statewide Planning and Research Cooperative longitudinal administrative database (n = 47,342). Multivariable logistic regression analyzed outcomes following laparoscopic bariatric surgery. Accredited hospitals and accreditation year were identified from the Centers for Medicaid and Medicare website. Outcomes were analyzed with and without temporal correlation to accreditation year.>30-day mortality was determined from social security death records. ResultsRisk of perioperative morbidity OR 1.4 (range 1.2–1.6, P<.001), mortality OR 2.6 (range 1.3–5.4, P = .01) and all-cause long-term mortality OR 1.4 (range 1.2–1.7, P = .0002) were significantly increased in unaccredited versus accredited hospitals on univariate analysis. In accredited hospitals, significant changes in payor and patient mix, operation, perioperative, and long-term outcomes were demonstrated following accreditation. A significant decrease in operations performed on black patients, Hispanic patients, and Medicare patients was also identified. Controlling for patient demographics, co-morbidity, insurance, and operative procedure, multivariable logistic regression demonstrated accreditation as independently associated with fewer major complications versus unaccredited hospitals OR 0.72 (range .63–.83, P<.001) and within the same hospital following accreditation OR .86 (range 0.77–0.96, P = .01). Following multiple cox proportional hazard model analysis, long-term mortality differences were not significant. ConclusionIn New York State, bariatric hospital accreditation improved patient outcomes as compared to unaccredited hospitals and within the same hospital compared to preaccreditation. Significant changes were identified for some underserved at-risk populations. Measures to ensure equitable health care for at-risk populations following institutional accreditation are imperative.

Impact of hypertension on the association of BMI with risk and age at onset of type 2 diabetes mellitus: age- and gender-mediated modifications.

AimsGiven that BMI correlates with risk of Type 2 diabetes mellitus (T2DM), and that hypertension is a common comorbid condition, we hypothesize that hypertension augments significantly the impact of obesity on T2DM onset.MethodsWe obtained data on T2DM in Kuwaiti natives from Kuwait Health Network Registry. We considered 1339 comorbid individuals with onset of hypertension preceding that of T2DM, and 3496 non-hypertensive individuals but with T2DM. Multiple linear regressions, ANOVA tests, and Cox proportional hazards models were used to quantify the impact of hypertension on correlation of BMI with age at onset and risk of T2DM.ResultsImpact of increasing levels of BMI on age at onset ot T2DM is seen augmented in patients diagnosed with hypertension. We find that the slope of the inverse linear relationship between BMI and onset age of T2DM is much steep in hypertensive patients (−0.69, males and −0.39, females) than in non-hypertensive patients (−0.36, males and −0.17, females). The decline in onset age for an unit increase of BMI is two-fold in males than in females. Upon considering BMI as a categorical variable, we find that while the mean onset age of T2DM in hypertensive patients decreases by as much as 5–12 years in every higher BMI categories, significant decrease in non-hypertensive patients exists only when severely obese. Hazard due to hypertension (against the baseline of non-hypertension and normal weight) increases at least two-fold in every obese category. While males have higher hazard due to hypertension in early adulthood, females have higher hazard in late adulthood.ConclusionPre-existing condition of hypertension augments the association of BMI with Type 2 diabetes onset in both males and females. The presented results provide health professionals directives on the extent of weight-loss required to delay onset of Type 2 diabetes in hypertensive versus non-hypertensive patients.

Predicting Coronary Heart Disease Using Risk Factor Categories for a Japanese Urban Population, and Comparison with the Framingham Risk Score: The Suita Study

The Framingham risk score (FRS) is one of the standard tools used to predict the incidence of coronary heart disease (CHD). No previous study has investigated its efficacy for a Japanese population cohort. The purpose of this study was to develop new coronary prediction algorithms for the Japanese population in the manner of the FRS, and to compare them with the original FRS. Our coronary prediction algorithms for Japanese were based on a large population-based cohort study (Suita study). The study population comprised 5,521 healthy Japanese. They were followed-up for 11.8 years on average, and 213 cases of CHD were observed. Multiple Cox proportional hazard model by stepwise selection was used to construct the prediction model. Our coronary prediction algorithms for Japanese patients were based on a large populationbased cohort study (the Suita study). A multiple Cox proportional hazard model by stepwise selection was used to construct the prediction model. The C-statistics showed that the new model had better accuracy than the original and recalibrated Framingham scores. The net reclassification improvement (NRI) by the Suita score with the inclusion of CKD was 41.2% (P＜0.001) compared with the original FRS. The recalibration of the FRS slightly improved the efficiency of the prediction, but it was still worse than the Suita score with the CKD model. The calibration analysis suggested that the original FRS and the recalibrated FRS overestimated the risk of CHD in the Japanese population. The Suita score with CKD more accurately predicted the risk of CHD. The FRS and recalibrated FRS overestimated the 10-year risk of CHD for the Japanese population. A predictive score including CKD as a coronary risk factor for the Japanese population was more accurate for predicting CHD than the original Framingham risk scores in terms of the C-statics and NRI.

Clinical effects and economical impact of dutasteride and finasteride therapy in Italian men with LUTS

To investigate differences in the risk of benign prostatic hyperplasia (BPH)- related hospitalization, for surgical and non-surgical reasons, and of new prostate cancer (PCa) diagnosis between patients under dutasteride or finasteride treatment. A retrospective cohort study was conducted using data from record-linkage of administrative databases. Men aged ≥ 40 years old who had received a prescription for at least 10 boxes/year (index years: 2004-06) were included. The association of the outcomes was assessed using a multiple Cox proportional hazard model. Propensity score matched analysis and a 5-to-1, greedy 1:1 matching algorithm were performed. The budget impact analysis of dutasteride vs finasteride in BPH-treated patient was performed. From an initial cohort of about 1.5 million of Italian men, 19620 were selected. The overall hospitalization for BPH-non surgical reasons, for BPH-related surgery and for new detection of PCa incidence rates (IRs) were 8.20 (95% CI, 7.62-8.23), 18.0 (95% CI, 17.12-18.93) and 8.62 (95% CI, 8.03-9.26) per 1000 person-years, respectively. The multivariate analysis after the propensity score-matching showed that dutasteride was associated with an independent reduced likelihood of hospitalization for BPH-related surgery (HR 0.82; 95% CI 0.73-0.93; p = 0.0025) and of newly detected PCa (HR: 0.76,95% CI, 0.65-0.85; p = 0.0116). The IR for BPH-non surgical reasons was 8.07 (95% CI, 7.10-9.17) and 9.25 (95% CI, 8.19-10.44) per 1000 person-years, respectively. The IR for BPH-related surgery was 18.28 (95% CI, 17.17-20.32) and 21.28 (95% CI, 19.24-23.06) per 1000 person-years among patients under dutasteride compared with those under finasteride, respectively. For new-onset PCa, the IR was 8.01 (95% CI, 7.07-9.08) and 9.38 (95% CI, 8.32-10.58) per 1000 person-years The pharmacoeconomical evaluation showed that the net budget impact of the use of dutasteride vs. finasteride in 1000 BPH-treated patient for 1 year induces a saving of 3933 €. The clinical effects of dutasteride and finasteride are slightly different. The likelihood of hospitalization for BPH-related surgery and of newly detected PCa seems to be in favor of dutasteride. The budget impact analyses showed a slightly benefit for dutasteride. Comparative prospective studies are necessary to confirm these results.

Abstract P5-12-04: Ethnicity disparities of breast cancer in an insured population

Abstract Introduction: Hispanic women have lower breast cancer incidence, but a worse five-year survival rate, compared to non-Hispanic white women. As seen in other cancers, this disparity in outcomes may reflect being diagnosed at an advanced stage. Such disparities have been explained by nationwide studies as due to socioeconomic or cultural differences impeding access to quality health care. We reviewed data from the Florida Hospital System (FHS) to discover clinical variables that may influence the survival disparity. Study Population: The eight Central Florida campuses of FHS are major health care providers serving an ethnically diverse population, including 22.9% Hispanics. Over 95% of cases in the Florida Hospital tumor registry are insured patients with one or a combination of Medicaid, Medicare and various types of commercial insurance. With a data set limited to patients with insurance at diagnosis, the FHS patient population provides a uniquely appropriate patient population to differentiate the cause/effect of disparities in clinical outcome of breast cancer from other socioeconomic mitigating factors. Methods: We compared the diagnostic stage, survival and tumor marker tests of Hispanic and non-Hispanic white breast cancer patients diagnosed at the FHS and explored factors that may influence diagnosis, survival and tumor marker tests. Hispanic and non-Hispanic white female patients (≥ 18 yr) who were first diagnosed with a single breast cancer tumor at the FHS between 2001 and 2006 were included. Multiple Logistic regression analyses were used to estimate odds ratios (OR) of late to early stage in relation to ethnicity, age and insurance type. Multiple Cox proportional hazard models and cumulative incidence functions were used to compare death rates between the two ethnicities. Results: A total of 1131 patients were included: 134 identified as Hispanic and 997 as non-Hispanic white. Hispanics were found to be over 10 times as likely as non-Hispanic whites to have Medicaid insurance (OR = 10.15, 95% C.I.: 4.99-20.70). Medicaid patients were more likely to be diagnosed at a late stage than non-Medicaid patients (OR = 4.73; 95% C.I.: 2.16-10.38; p&amp;lt;0.001). As age increased, patients were less likely to be diagnosed at a late stage (OR = 0.86; 95% C.I.: 0.74-0.99; p = 0.038). The death rate was associated with age, insurance type, and stage rather than ethnicity. Hispanic patients had significantly higher percentage of PR positive than non-Hispanic patients (83.1% vs. 72.5%, p = 0.048). No association between ER/HER2 positive and ethnicity was detected. Conclusions: In this study population, there was no significant difference in survival based solely on ethnicity. Our findings support that interventions are needed to enhance the awareness of young people of early screening, and to improve equal access to early screening for people of low socioeconomic status or Medicaid insured patients. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P5-12-04.

The Determinants of Dementia after Stroke (DEDEMAS) Study: Protocol and Pilot Data

About 20% of stroke patients develop dementia within a few months after their event, but the determinants and mechanisms of poststroke dementia are insufficiently understood. To identify and characterize the determinants of cognitive impairment poststroke. Observational prospective study in patients with acute stroke and no prior dementia. Six hundred subjects will be characterized by detailed interview, standardized clinical examinations, biometric measures (intima-media thickness, waist-hip ratio, and ankle-brachial index), multimodal imaging (magnetic resonance imaging, fluorodeoxyglucose-positron emission tomography (FDG-PET), amyloid-positron emission tomography (amyloid-PET), and retinal imaging), analysis of biomarkers derived from blood and cerebrospinal fluid, and detailed cognitive testing at repeat time points. Patients will be followed for five-years with a total of five personal visits and three telephone interviews. Primary end-point is the occurrence of poststroke dementia. Secondary end-points include poststroke cognitive impairment-no dementia, stroke recurrence, and death. Predictive factors for poststroke dementia will be identified by multiple Cox proportional-hazards model. Baseline characteristics of the first 71 patients (study inclusion between May 2011 and August 2012) are as follows: median age, 70 years (interquartile range, 65-75); female gender, 25 (35%); median National Institutes of Health Stroke Scale at admission, 2 (1-4); and etiological stroke subtypes according to TOAST classification, 15% large artery disease, 18% small vessel disease, 35% cardioembolic, and 32% undetermined or multiple competing etiologies. This study will provide insights into the mechanisms of poststroke dementia and hold the potential to identify novel diagnostic markers and targets for preventive therapies. The study is registered at http://www.clinicaltrials.gov (NCT01334749) and will be extended as a multicenter study starting 2013.

Factors Associated With Converting Scientific Abstracts to Published Manuscripts

The objective of this study was to identify factors associated with the conversion of scientific abstracts to publications in peer-reviewed journals. This was a retrospective study of abstracts presented by members of the Massachusetts General Hospital Department of Oral and Maxillofacial Surgery between 2000 and 2010. Predictor variables were categorized as abstract- or author-specific. The outcome variable was conversion of an abstract presented at a meeting to a published manuscript. Descriptive, bivariate, and multiple regression statistics were computed. P < 0.05 was significant. The sample was composed of 122 abstracts presented at meetings. Ninety abstracts (73.8%) were published in a peer-reviewed literature within 5 years of presentation. The mean time between presentation to publication was 21.9 ± 17.3 months (median, 19.0 months; range, 0-99 months). In bivariate analyses, study design, number of prior publications by the presenting author, and number of prior publications by the senior author were associated with time to publication (P < 0.06). In a multiple Cox proportional hazards model, higher levels of evidence (hazard ratio, 1.2; 95% confidence interval, 1.04-1.3; P = 0.006) and volume of prior publications by the senior author (hazard ratio, 1.007; 95% confidence interval, 1.003-1.011; P < 0.001) were associated with shorter publication times. Time to publication of scientific abstracts is associated with study quality, prior research experience by the presenting author, and senior author identity and experience.

The effects of dutasteride and finasteride on BPH-related hospitalization, surgery and prostate cancer diagnosis: a record-linkage analysis

To investigate differences in the risk of benign prostatic hyperplasia (BPH)-related hospitalization, for surgical and non-surgical reasons, and of new prostate cancer (PCa) diagnosis between patients using finasteride or dutasteride. A retrospective cohort study was conducted using data from record linkage of administrative databases (pharmaceutical prescription data, hospital discharge records, Italian population registry). Men aged ≥ 40 years old who had received a prescription for at least 10 packs/year between January 1, 2004 and December 31, 2004 were included and followed for 5 years. The association of the outcomes was assessed using a multiple Cox proportional hazard model. Propensity score-matched analysis and a 5-1, greedy 1:1 matching algorithm were performed. 8,132 patients were identified. Overall incidence rates of BPH hospitalization and BPH-related surgery were 21.05 (95 % CI 19.52-22.71) and 20.97 (95 % CI 19.45-22.61) per 1,000 person-years, respectively. In the dutasteride group compared with finasteride group, the incidence rate of both events was statistically significant lower: 16.07 versus 21.76 for BPH hospitalization and 15.91 versus 21.69 for BPH-related surgery. The incidence rate of new PCa was also lower for the dutasteride group [8.34 (95 % CI 5.96-11.68) vs. 10.25 (95 % CI 9.15-11.49)]. Dutasteride was associated with a reduction in BPH-related hospitalizations (HR 0.75, 95 % CI 0.58-0.98 and 0.58-0.98 for surgical and non-surgical reasons). The matched analysis confirmed the risk reduction with dutasteride for BPH-related surgery. These findings suggest that the clinical effects of dutasteride and finasteride might be different. Patients treated with dutasteride seem to be less likely to experience BPH-related hospitalization. Comparative studies are needed to confirm these results.

Mortality outcome of out-of-hours primary percutaneous coronary intervention in the current era

To assess the impact of the time of primary percutaneous coronary intervention (PPCI) on in-hospital and long-term all-cause mortality in ST-segment elevation myocardial infarction (STEMI). The study retrospectively analyses the prospectively collected data on 2571 consecutive PPCI-treated STEMI patients between March 2008 and June 2011. Of these, 1036 patients (40.3%) underwent PPCI during a weekday between 08:00 and 18:00 (routine-hours group) and 1535 patients (59.7%) underwent PPCI on a weekday between 18:00 and 08:00 or a weekend (out-of-hours group). Compared with the routine-hours group, the out-of-hours group had a lower mean age, fewer patients with previous angina, longer call-to-hospital time, and fewer multivessel PCI. The overall in-hospital mortality rate was 4.5% with no significant difference [0.2%, 95% confidence interval (CI): -1.4 to 1.9%] between the routine-hours group (4.3%) and the out-of-hours group (4.6%) (adjusted odds ratio: 1.33, 95% CI: 0.73-2.40, P = 0.35). During a mean follow-up period of 560 days, 295 patients (11.5%) died, 12.2% in the routine-hours group and 11.0% in the out-of-hours group (difference of -0.1%, 95% CI: -0.4 to 0.2%). In the multiple Cox proportional hazards model, there was no difference in mortality between the two groups (adjusted hazard ratio: 1.09, 95% CI: 0.82-1.46, P = 0.57). Similarly, no increase in mortality was seen in patients who underwent PPCI later at night (22:00-06:00). This study of real-world, unselected STEMI patients demonstrates that in a large, well-staffed centre, PPCI outside routine-working hours is safe with no difference in outcome of in-hospital and long-term mortality compared with PPCI during routine-working hours.

Comparative effectiveness of different angiotensin-converting enzyme inhibitors on the risk of hospitalization in patients with heart failure

Existing randomized controlled trials do not address the comparative effectiveness of different angiotensin-converting enzyme inhibitors (ACEIs) on hospitalization due to heart failure (HF)-hospitalization in patients with HF. We sought to examine the effect of four ACEIs on HF-hospitalization in a large real-world HF population. The study was a retrospective analysis of a national cohort of patients with HF identified from the Department of Veterans Affairs (TX, USA). A multiple propensity score analysis was used to balance 47 baseline patient characteristics between the different ACEIs. The effect of different ACEIs on time to HF-hospitalization was assessed using the multiple propensity score-weighted multivariable Cox proportional hazard model. The study included 139,994 patients with 69.50% (97,293) on lisinopril, 21.79% (30,503) on fosinopril, 8.41% (11,775) on captopril and 0.30% (423) on enalapril. Propensity scores balanced nearly all differences between different ACEIs groups. Enalapril (hazard ratio [HR]: 0.800; 95% CI: 0.492-1.297), fosinopril (HR: 0.971; 95% CI: 0.877-1.074), and lisinopril (HR: 1.005; 95% CI: 0.918-1.101) when compared with captopril were found to have similar effectiveness in reducing HF-hospitalizations. In patients with HF, we found that the four ACEIs are equally effective in reducing HF-hospitalization in day-to-day practice.

Expression and prognostic role of pan-Ras, Raf-1, pMEK1 and pERK1/2 in patients with hepatocellular carcinoma

Aims Ras/Raf/mitogen-activated protein kinase (MEK)/extracellular signal-regulated kinase (ERK) signaling cascades play important roles in the transmission of signals involved in apoptosis. Importantly, components of these pathways are aberrantly expressed in human cancer. However, there is limited data linking clinical outcomes with the aberrant expression of this pathway. The present study analyzed the prognostic values of pan-Ras, Raf-1, phosphorylated MEK1 (pMEK1) and phosphorylated ERK1/2 (pERK1/2) in hepatocellular carcinoma (HCC). Methods Expression of pan-Ras, Raf-1, pMEK1 and pERK1/2 in 81 HCC patients who underwent curative resection was examined by immunohistochemical staining. Long-term survival after resection of patients according to the expression of pan-Ras, Raf-1, pMEK1 and pERK1/2 was assessed using univariate analysis and multiple Cox proportional hazards model. Results In univariate analysis, patients with Raf-1 or pMEK1 overexpression had shorter disease-free survival (DFS) ( P＜ 0.05) and poorer overall survival (OS) ( P＜ 0.05) than groups with weak-expression of Raf-1 or pMEK1, respectively. Patients with pan-Ras overexpression had poorer overall survival (OS) ( P＜ 0.05) than the group with weak-expression of pan-Ras. Importantly, Raf-1 overexpression was a promising prognostic marker for poor survival according to multivariate Cox regression analysis (DFS, Hazard Ratio 1.807, P = 0.035; OS, Hazard Ratio 1.959, P = 0.044). Conclusions Raf-1 overexpression could be considered as an independent prognostic biomarker in HCC and may predict early tumor recurrence and death for HCC patients. It can be used to stratify patients at higher risk for poor prognosis and help to select the appropriate therapeutic regime of HCC.