Survival Analysis of Patients with Heart Failure: Implications of Time-Varying Regression Effects in Modeling Mortality

Background: Several models have been designed to predict survival of patients with heart failure. These, while available and widely used for both stratifying and deciding upon different treatment options on the individual level, have several limitations. Specifically, some clinical variables that may influence prognosis may have an influence that change over time. Statistical models that include such characteristic may help in evaluating prognosis. The aim of the present study was to analyze and quantify the impact of modeling heart failure survival allowing for covariates with time-varying effects known to be independent predictors of overall mortality in this clinical setting. Methodology: Survival data from an inception cohort of five hundred patients diagnosed with heart failure functional class III and IV between 2002 and 2004 and followed-up to 2006 were analyzed by using the proportional hazards Cox model and variations of the Cox's model and also of the Aalen's additive model. Principal Findings: One-hundred and eighty eight (188) patients died during follow-up. For patients under study, age, serum sodium, hemoglobin, serum creatinine, and left ventricular ejection fraction were significantly associated with mortality. Evidence of time-varying effect was suggested for the last three. Both high hemoglobin and high LV ejection fraction were associated with a reduced risk of dying with a stronger initial effect. High creatinine, associated with an increased risk of dying, also presented an initial stronger effect. The impact of age and sodium were constant over time. Conclusions: The current study points to the importance of evaluating covariates with time-varying effects in heart failure models. The analysis performed suggests that variations of Cox and Aalen models constitute a valuable tool for identifying these variables. The implementation of covariates with time-varying effects into heart failure prognostication models may reduce bias and increase the specificity of such models.

Assessment of candidemia-attributable mortality in critically ill patients using propensity score matching analysis

[EN] Introduction: Candidemia in critically ill patients is usually a severe and life-threatening condition with a high crude mortality. Very few studies have focused on the impact of candidemia on ICU patient outcome and attributable mortality still remains controversial. This study was carried out to determine the attributable mortality of ICU-acquired candidemia in critically ill patients using propensity score matching analysis. Methods: A prospective observational study was conducted of all consecutive non-neutropenic adult patients admitted for at least seven days to 36 ICUs in Spain, France, and Argentina between April 2006 and June 2007. The probability of developing candidemia was estimated using a multivariate logistic regression model. Each patient with ICU-acquired candidemia was matched with two control patients with the nearest available Mahalanobis metric matching within the calipers defined by the propensity score. Standardized differences tests (SDT) for each variable before and after matching were calculated. Attributable mortality was determined by a modified Poisson regression model adjusted by those variables that still presented certain misalignments defined as a SDT > 10%. Results: Thirty-eight candidemias were diagnosed in 1,107 patients (34.3 episodes/1,000 ICU patients). Patients with and without candidemia had an ICU crude mortality of 52.6% versus 20.6% (P < 0.001) and a crude hospital mortality of 55.3% versus 29.6% (P = 0.01), respectively. In the propensity matched analysis, the corresponding figures were 51.4% versus 37.1% (P = 0.222) and 54.3% versus 50% (P = 0.680). After controlling residual confusion by the Poisson regression model, the relative risk (RR) of ICU- and hospital-attributable mortality from candidemia was RR 1.298 (95% confidence interval (CI) 0.88 to 1.98) and RR 1.096 (95% CI 0.68 to 1.69), respectively. Conclusions: ICU-acquired candidemia in critically ill patients is not associated with an increase in either ICU or hospital mortality.

Gender-related mortality trends among diabetic patients with ST-segment elevation myocardial infarction: insights from a nationwide registry 1997-2010

BACKGROUND Data on temporal trends in outcomes, gender differences, and adherence to evidence-based therapy (EBT) of diabetic patients with ST-segment elevation myocardial infarction (STEMI) are sparse. METHODS We performed a retrospective analysis of prospectively acquired data on 3565 diabetic (2412 males and 1153 females) STEMI patients enrolled in the Swiss AMIS Plus registry between 1997 and 2010 and compared in-hospital outcomes and adherence to EBT with the nondiabetic population (n=15,531). RESULTS In-hospital mortality dramatically decreased in diabetic patients, from 19.9% in 1997 to 9.0% in 2010 (p trend<0.001) with an age-adjusted decrease of 6% per year of admission. Similar trends were observed for age-adjusted reinfarction (OR 0.86, p<0.001), cardiogenic shock (OR 0.88, p<0.001), as well as death, reinfarction, or stroke (OR 0.92, p<0.001). However, the mortality benefit over time was observed in diabetic males (p trend=0.006) but not females (p trend=0.082). In addition, mortality remained twice as high in diabetic patients compared with nondiabetic ones (12.1 vs. 6.1%, p<0.001) and diabetes was identified as independent predictor of mortality (OR 1.23, p=0.022). Within the diabetic cohort, females had higher mortality than males (16.1 vs. 10.2%, p<0.001) and female gender independently predicted in-hospital mortality (OR 1.45, p=0.015). Adherence to EBT significantly improved over time in diabetic patients (p trend<0.001) but remained inferior - especially in women - to the one of nondiabetic individuals. CONCLUSIONS In-hospital mortality and morbidity of diabetic STEMI patients in Switzerland improved dramatically over time but, compared with nondiabetic counterparts, gaps in outcomes as well as EBT use persisted, especially in women.

The 2011-12 pilot European Sentinel Registry of Transcatheter Aortic Valve Implantation: in-hospital results in 4,571 patients

AIMS The aim of this prospective multinational registry is to assess and identify predictors of in-hospital outcome and complications of contemporary TAVI practice. METHODS AND RESULTS The Transcatheter Valve Treatment Sentinel Pilot Registry is a prospective independent consecutive collection of individual patient data entered into a web-based case record form (CRF) or transferred from compatible national registries. A total of 4,571 patients underwent TAVI between January 2011 and May 2012 in 137 centres of 10 European countries. Average age was 81.4±7.1 years with equal representation of the two sexes. Logistic EuroSCORE (20.2±13.3), access site (femoral approach: 74.2%), type of anaesthesia and duration of hospital stay (9.3±8.1 days) showed wide variations among the participating countries. In-hospital mortality (7.4%), stroke (1.8%), myocardial infarction (0.9%), major vascular complications (3.1%) were similar in the SAPIEN XT and CoreValve (p=0.15). Mortality was lower in transfemoral (5.9%) than in transapical (12.8%) and other access routes (9.7%; p<0.01). Advanced age, high logistic EuroSCORE, pre-procedural ≥grade 2 mitral regurgitation and deployment failure predicted higher mortality at multivariate analysis. CONCLUSIONS Increased operator experience and the refinement of valve types and delivery catheters may explain the lower rate of mortality, stroke and vascular complications than in historical studies and registries.

Percutaneous mitral valve edge-to-edge repair: in-hospital results and 1-year follow-up of 628 patients of the 2011-2012 Pilot European Sentinel Registry.

BACKGROUND The use of transcatheter mitral valve repair (TMVR) has gained widespread acceptance in Europe, but data on immediate success, safety, and long-term echocardiographic follow-up in real-world patients are still limited. OBJECTIVES The aim of this multinational registry is to present a real-world overview of TMVR use in Europe. METHODS The Transcatheter Valve Treatment Sentinel Pilot Registry is a prospective, independent, consecutive collection of individual patient data. RESULTS A total of 628 patients (mean age 74.2 ± 9.7 years, 63.1% men) underwent TMVR between January 2011 and December 2012 in 25 centers in 8 European countries. The prevalent pathogenesis was functional mitral regurgitation (FMR) (n = 452 [72.0%]). The majority of patients (85.5%) were highly symptomatic (New York Heart Association functional class III or higher), with a high logistic EuroSCORE (European System for Cardiac Operative Risk Evaluation) (20.4 ± 16.7%). Acute procedural success was high (95.4%) and similar in FMR and degenerative mitral regurgitation (p = 0.662). One clip was implanted in 61.4% of patients. In-hospital mortality was low (2.9%), without significant differences between groups. The estimated 1-year mortality was 15.3%, which was similar for FMR and degenerative mitral regurgitation. The estimated 1-year rate of rehospitalization because of heart failure was 22.8%, significantly higher in the FMR group (25.8% vs. 12.0%, p[log-rank] = 0.009). Paired echocardiographic data from the 1-year follow-up, available for 368 consecutive patients in 15 centers, showed a persistent reduction in the degree of mitral regurgitation at 1 year (6.0% of patients with severe mitral regurgitation). CONCLUSIONS This independent, contemporary registry shows that TMVR is associated with high immediate success, low complication rates, and sustained 1-year reduction of the severity of mitral regurgitation and improvement of clinical symptoms.

Calcium Disorders in the Emergency Department: Independent Risk Factors for Mortality.

BACKGROUND Calcium disorders are common in both intensive care units and in patients with chronic kidney disease and are associated with increased morbidity and mortality. It is unknown whether calcium abnormalities in unselected emergency department admissions have an impact on in-hospital mortality. METHODS This cross-sectional analysis included all admissions to the Emergency Department at the Inselspital Bern, Switzerland from 2010 to 2011. For hyper- and hypocalcaemic patients with a Mann-Whitney U-test, the differences between subgroups divided by age, length of hospital stay, creatinine, sodium, chloride, phosphate, potassium and magnesium were compared. Associations between calcium disorders and 28-day in-hospital mortality were assessed using the Cox proportional hazard regression model. RESULTS 8,270 patients with calcium measurements were included in our study. Overall 264 (3.2%) patients died. 150 patients (6.13%) with hypocalcaemia and 7 patients with hypercalcaemia (6.19%) died, in contrast to 104 normocalcaemic patients (1.82%). In univariate analysis, calcium serum levels were associated with sex, mortality and pre-existing diuretic therapy (all p<0.05). In multivariate Cox regression analysis, hypocalcaemia and hypercalcaemia were independent risk factors for mortality (HR 2.00 and HR 1.88, respectively; both p<0.01). CONCLUSION Both hypocalcaemia and hypercalcaemia are associated with increased 28-day in-hospital mortality in unselected emergency department admissions.

Hypermagnesemia is a strong independent risk factor for mortality in critically ill patients: results from a cross-sectional study.

BACKGROUND Patients with electrolyte imbalances or disorders have a high risk of mortality. It is unknown if this finding from sodium or potassium disorders extends to alterations of magnesium levels. METHODS AND PATIENTS In this cross-sectional analysis, all emergency room patients between 2010 and 2011 at the Inselspital Bern, Switzerland, were included. A multivariable logistic regression model was performed to assess the association between magnesium levels and in-hospital mortality up to 28days. RESULTS A total of 22,239 subjects were screened for the study. A total of 5339 patients had plasma magnesium concentrations measured at hospital admission and were included into the analysis. A total of 6.3% of the 352 patients with hypomagnesemia and 36.9% of the 151 patients with hypermagnesemia died. In a multivariate Cox regression model hypermagnesemia (HR 11.6, p<0.001) was a strong independent risk factor for mortality. In these patients diuretic therapy revealed to be protective (HR 0.5, p=0.007). Hypomagnesemia was not associated with mortality (p>0.05). Age was an independent risk factor for mortality (both p<0.001). CONCLUSION The study does demonstrate a possible association between hypermagnesemia measured upon admission in the emergency department, and early in-hospital mortality.

Hyperphosphatemia Is an Independent Risk Factor for Mortality in Critically Ill Patients: Results from a Cross-Sectional Study.

BACKGROUND Phosphate imbalances or disorders have a high risk of morbidity and mortality in patients with chronic kidney disease. It is unknown if this finding extends to mortality in patients presenting at an emergency room with or without normal kidney function. METHODS AND PATIENTS This cross sectional analysis included all emergency room patients between 2010 and 2011 at the Inselspital Bern, Switzerland. A multivariable cox regression model was applied to assess the association between phosphate levels and in-hospital mortality up to 28 days. RESULTS 22,239 subjects were screened for the study. Plasma phosphate concentrations were measured in 2,390 patients on hospital admission and were included in the analysis. 3.5% of the 480 patients with hypophosphatemia and 10.7% of the 215 patients with hyperphosphatemia died. In univariate analysis, phosphate levels were associated with mortality, age, diuretic therapy and kidney function (all p<0.001). In a multivariate Cox regression model, hyperphosphatemia (OR 3.29, p<0.001) was a strong independent risk factor for mortality. Hypophosphatemia was not associated with mortality (p>0.05). CONCLUSION Hyperphosphatemia is associated with 28-day in-hospital mortality in an unselected cohort of patients presenting in an emergency room.

Exploring race/ethnicity as a risk factor for mortality in newborns following congenital heart surgery at a center of excellence

The purpose of this study was to determine if race/ethnicity was a significant risk factor for hospital mortality in children following congenital heart surgery in a contemporary sample of newborns with congenital heart disease. Unlike previous studies that utilized administrative databases, this study utilized clinical data collected at the point of care to examine racial/ethnic outcome differences in the context of the patients' clinical condition and their overall perioperative experience. A retrospective cohort design was used. The study sample consisted of 316 newborns (<31 days of age) who underwent congenital heart surgery between January 2007 through December 2009. A multivariate logistic regression model was used to determine the impact of race/ethnicity, insurance status, presence of a spatial anomaly, prenatal diagnosis, postoperative sepsis, cardiac arrest, respiratory failure, unplanned reoperation, and total length of stay in the intensive care unit on outcomes following congenital heart surgery in newborns. The study findings showed that the strongest predictors of hospital mortality following congenital heart surgery in this cohort were postoperative cardiac arrest, postoperative respiratory failure, having a spatial anomaly, and total ICU LOS. Race/ethnicity and insurance status were not significant risk factors. The institution where this study was conducted is designated as a center of excellence for congenital heart disease. These centers have state-of-the-art facilities, extensive experience in caring for children with congenital heart disease, and superior outcomes. This study suggests that optimal care delivery for newborns requiring congenital heart surgery at a center of excellence portends exceptional outcomes and this benefit is conferred upon the entire patient population despite the race/ethnicity of the patients. From a public health and health services view, this study also contributes to the overall body of knowledge on racial/ethnic disparities in children with congenital heart defects and puts forward the possibility of a relationship between quality of care and racial/ethnic disparities. Further study is required to examine the impact of race/ethnicity on the long-term outcomes of these children as they encounter the disparate components of the health care delivery system. There is also opportunity to study the role of race/ethnicity on the hospital morbidity in these patients considering current expectations for hospital survival are very high, and much of the current focus for quality improvement rests in minimizing the development of patient morbidities.^

Comparison of two strategies for the treatment of acute myocardial infarction: In-hospital outcomes analysis

A strategy of pre-hospital reduced dose fibrinolytic administration coupled with urgent coronary intervention (PCI) for patients with STEMI (FAST-PCI) has been found to be superior to primary PCI (PPCI) alone. A coordinated STEMI system-of-care that includes FAST-PCI might offer better outcomes than pre-hospital diagnosis and STEMI team activation followed by PPCI alone. We compared the in-hospital outcomes for patients treated with the FAST-PCI approach with outcomes for patients treated with the PPCI approach during a pause in the FAST-PCI protocol. In-hospital data for 253 STEMI patients (03/2003–12/2009), treated with FAST-PCI protocol were compared to 124 patients (12/2009–08/2011), treated with PPCI strategy alone. In-hospital mortality was the primary endpoint. Stroke, major bleeding, and reinfarction during index hospitalization were secondary endpoints. Comparing the strategies used during the two time intervals, in-hospital mortality was significantly lower with FAST-PCI than with PPCI (2.77% vs. 10.48%, p = 0.0017). Rates of stroke, reinfarction and major bleeding were similar between the two groups. There was a lower frequency of pre- PCI TIMI 0 flow (no patency) seen in patients treated with FAST-PCI compared to the PPCI patients (26.7% vs. 62.7%, p<0.0001). Earlier infarct related artery patency in the FAST-PCI group had a favorable impact on the incidence of cardiogenic shock at hospital admission (FAST-PCI- 3.1% vs. PPCI- 20.9%, p<0.0001). The FAST-PCI strategy was associated with earlier infarct related artery patency and the lower incidence of cardiogenic shock on hospital arrival, as well as with reduced in-hospital mortality among STEMI patients.^

Ten years of hospital admissions for liver cirrhosis in Portugal

Background and aims More data on epidemiology of liver diseases in Europe are needed. We aimed to characterize hospital admissions for liver cirrhosis in Portugal during the past decade. Patients and methods We analyzed all hospital admissions for cirrhosis in Portugal Mainland between 2003 and 2012 registered in the national Diagnosis-Related Group database. Cirrhosis was classified according to etiology considering alcohol, hepatitis B, and hepatitis C. Results Between 2003 and 2012, there were 63 910 admissions for cirrhosis in Portugal Mainland; 74.4% involved male patients. Etiologies of admitted cirrhosis were as follows: 76.0% alcoholic, 1.1% hepatitis B, 1.4% hepatitis B plus alcohol, 3.6% hepatitis C, and 4.0% hepatitis C plus alcohol. There was a significant decline (P <0.001) in admissions for alcoholic cirrhosis, whereas hospitalizations for cirrhosis caused by hepatitis C or hepatitis C plus alcohol increased by almost 50% (P <0.001). Patients admitted with alcoholic plus hepatitis B or C cirrhosis were significantly younger than those with either alcoholic or viral cirrhosis (53.1 vs. 59.4 years, respectively, P <0.001). Hospitalization rates for cirrhosis were 124.4/100 000 in men and 32.6/100 000 in women. Hepatocellular carcinoma and fluid retention were more common in viral cirrhosis, whereas encephalopathy and variceal bleeding were more frequent in alcoholic cirrhosis. Hepatorenal syndrome was the strongest predictor of mortality among cirrhosis complications (odds ratio 12.97; 95% confidence interval 11.95–14.09). In-hospital mortality was 15.2%. Conclusion Despite the decline in admissions for alcoholic cirrhosis and the increase in those related to hepatitis C, the observed burden of hospitalized liver cirrhosis in Portugal was essentially attributable to alcoholic liver disease.

Medicare hospital information.

Issued in numerous parts identified by no. and geographic area (e.g.: v. 46, Texas. Part 1) and including as its final no. a technical suppl. not additionally identified.

Usefulness of the Hepatocyte Growth Factor as a Predictor of Mortality in Patients Hospitalized With Acute Heart Failure Regardless of Ejection Fraction

Hepatocyte growth factor (HGF) plays a role in the improvement of cardiac function and remodeling. Their serum levels are strongly related with mortality in chronic systolic heart failure (HF). The aim of this study was to study prognostic value of HGF in acute HF, interaction with ejection fraction, renal function, and natriuretic peptides. We included 373 patients (age 76 ± 10 years, left ventricular ejection fraction [LVEF] 46 ± 14%, 48% men) consecutively admitted for acute HF. Blood samples were obtained at admission. All patients were followed up until death or close of study (>1 year, median 371 days). HGF concentrations were determined using a commercial enzyme-linked immunosorbent assay (human HGF immunoassay). The predictive power of HGF was estimated by Cox regression with calculation of Harrell C-statistic. HGF had a median of 1,942 pg/ml (interquartile rank 1,354). According to HGF quartiles, mortality rates (per 1,000 patients/year) were 98, 183, 375, and 393, respectively (p <0.001). In Cox regression analysis, HGF (hazard ratio1SD = 1.5, 95% confidence interval 1.1 to 2.1, p = 0.002) and N-terminal pro b-type natriuretic peptide (NT-proBNP; hazard ratio1SD = 1.8, 95% confidence interval 1.2 to 2.6, p = 0.002) were independent predictors of mortality. Interaction between HGF and LVEF, origin, and renal function was nonsignificant. The addition of HGF improved the predictive ability of the models (C-statistic 0.768 vs 0.741, p = 0.016). HGF showed a complementary value over NT-proBNP (p = 0.001): mortality rate was 490 with both above the median versus 72 with both below. In conclusion, in patients with acute HF, serum HGF concentrations are elevated and identify patients at higher risk of mortality, regardless of LVEF, ischemic origin, or renal function. HGF had independent and additive information over NT-proBNP.

Patient physiological status during emergency care and rapid response team or cardiac arrest team activation during early hospital admission

OBJECTIVES: The objective of this study was to examine the relationship between rapid response team (RRT) or cardiac arrest team (CAT) activation within 72 h of emergency admission and (i) physiological status in the emergency department (ED) and (ii) risk for ICU admission and in-hospital mortality.

METHODS: A retrospective matched cohort study was conducted in three hospitals in Melbourne, Australia. The exposed cohort (n=660) included randomly selected adults admitted to the medical or surgical ward through the ED who had RRT or CAT activation within 72 h of admission. Unexposed matched controls (n=1320) did not have RRT or CAT activation.

RESULTS: The exposed cohort was more likely to have physiological abnormalities fulfilling hospital RRT activation criteria during ED care (36.7 vs. 23.8%, P<0.001). After adjusting for confounders, tachypnoea (adjusted odds ratio=1.92, 95% confidence interval: 1.38-2.67) or hypotension (AOR=1.43, 95% confidence interval: 1.00-2.03), fulfilling RRT activation criteria during ED care, was associated with RRT or CAT activation within 72 h of admission. The exposed cohort had more in-hospital deaths (16.5 vs. 3.6%, P<0.001), more unexpected in-hospital deaths (2.05 vs. 0.2%, P<0.001), more ICU admissions (11.8 vs. 0.7%, P<0.001) and longer lengths of hospital stay (median=8 vs. 5 days, P<0.001).

CONCLUSION: CAT/RRT activations within 72 h of emergency admission are associated with higher mortality and increased length of stay. Factors associated with CAT/RRT activation in the wards are often identifiable when patients are in the ED. Further studies are required to determine whether early identification and intervention in patients at risk for RRT or CAT activation can improve their eventual outcomes.

Multi-task transfer learning for in hospital-death prediction for ICU patients

Multi-Task Transfer Learning (MTTL) is an efficient approach for learning from inter-related tasks with small sample size and imbalanced class distribution. Since the intensive care unit (ICU) data set (publicly available in Physionet) has subjects from four different ICU types, we hypothesizethat there is an underlying relatedness amongst various ICU types. Therefore, this study aims to explore MTTL model for in-hospital mortality prediction of ICU patients. We used singletask learning (STL) approach on the augmented data as well as individual ICU data and compared the performance with the proposed MTTL model. As a performance measurement metrics, we used sensitivity (Sens), positive predictivity (+Pred), and Score. MTTL with class balancing showed the best performance with score of 0.78, 0.73, o.52 and 0.63 for ICU type 1(Coronary care unit), 2 (Cardiac surgery unit), 3 (Medical ICU) and 4 (Surgical ICU) respectively. In contrast the maximum score obtained using STL approach was 0.40 for ICU type 1 & 2. These results indicates that the performance of in-hospital mortality can be improved using ICU type information and by balancing the ’non-survivor’ class. The findings of the study may be useful for quantifying the quality of ICU care, managing ICU resources and selecting appropriate interventions.