Choosing between Cox proportional hazards and logistic models for interval-censored data via bootstrap

This work develops a new methodology in order to discriminate models for interval-censored data based on bootstrap residual simulation by observing the deviance difference from one model in relation to another, according to Hinde (1992). Generally, this sort of data can generate a large number of tied observations and, in this case, survival time can be regarded as discrete. Therefore, the Cox proportional hazards model for grouped data (Prentice & Gloeckler, 1978) and the logistic model (Lawless, 1982) can befitted by means of generalized linear models. Whitehead (1989) considered censoring to be an indicative variable with a binomial distribution and fitted the Cox proportional hazards model using complementary log-log as a link function. In addition, a logistic model can be fitted using logit as a link function. The proposed methodology arises as an alternative to the score tests developed by Colosimo et al. (2000), where such models can be obtained for discrete binary data as particular cases from the Aranda-Ordaz distribution asymmetric family. These tests are thus developed with a basis on link functions to generate such a fit. The example that motivates this study was the dataset from an experiment carried out on a flax cultivar planted on four substrata susceptible to the pathogen Fusarium oxysoprum. The response variable, which is the time until blighting, was observed in intervals during 52 days. The results were compared with the model fit and the AIC values.

Avaliação da longevidade de vacas da raça Nelore por diferentes critérios utilizando análise de sobrevivência

Pós-graduação em Genética e Melhoramento Animal - FCAV

Characterization of the variable cow's age at last calving as a measurement of longevity by using the Kaplan-Meier estimator and the Cox model

In most studies on beef cattle longevity, only the cows reaching a given number of calvings by a specific age are considered in the analyses. With the aim of evaluating all cows with productive life in herds, taking into consideration the different forms of management on each farm, it was proposed to measure cow longevity from age at last calving (ALC), that is, the most recent calving registered in the files. The objective was to characterize this trait in order to study the longevity of Nellore cattle, using the Kaplan-Meier estimators and the Cox model. The covariables and class effects considered in the models were age at first calving (AFC), year and season of birth of the cow and farm. The variable studied (ALC) was classified as presenting complete information (uncensored = 1) or incomplete information (censored = 0), using the criterion of the difference between the date of each cow's last calving and the date of the latest calving at each farm. If this difference was >36 months, the cow was considered to have failed. If not, this cow was censored, thus indicating that future calving remained possible for this cow. The records of 11 791 animals from 22 farms within the Nellore Breed Genetic Improvement Program ('Nellore Brazil') were used. In the estimation process using the Kaplan-Meier model, the variable of AFC was classified into three age groups. In individual analyses, the log-rank test and the Wilcoxon test in the Kaplan-Meier model showed that all covariables and class effects had significant effects (P < 0.05) on ALC. In the analysis considering all covariables and class effects, using the Wald test in the Cox model, only the season of birth of the cow was not significant for ALC (P > 0.05). This analysis indicated that each month added to AFC diminished the risk of the cow's failure in the herd by 2%. Nonetheless, this does not imply that animals with younger AFC had less profitability. Cows with greater numbers of calvings were more precocious than those with fewer calvings. Copyright © The Animal Consortium 2012.

Tests of proportional hazards and proportional odds models for grouped survival data

In this paper, we derive score test statistics to discriminate between proportional hazards and proportional odds models for grouped survival data. These models are embedded within a power family transformation in order to obtain the score tests. In simple cases, some small-sample results are obtained for the score statistics using Monte Carlo simulations. Score statistics have distributions well approximated by the chi-squared distribution. Real examples illustrate the proposed tests.

Likelihood approximations and discrete models for tied survival data

Ties among event times are often recorded in survival studies. For example, in a two week laboratory study where event times are measured in days, ties are very likely to occur. The proportional hazards model might be used in this setting using an approximated partial likelihood function. This approximation works well when the number of ties is small. on the other hand, discrete regression models are suggested when the data are heavily tied. However, in many situations it is not clear which approach should be used in practice. In this work, empirical guidelines based on Monte Carlo simulations are provided. These recommendations are based on a measure of the amount of tied data present and the mean square error. An example illustrates the proposed criterion.

A Generalized Log-Normal Model for Grouped Survival Data

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

CD147 overexpression allows an accurate discrimination of bladder cancer patients' prognosis

Background: Urothelial bladder carcinoma (UBC) is a chemo-sensitive tumour, but the response to treatment is heterogeneous. CD 147 has been associated with chemotherapy resistance. We aimed to define tumours with an aggressive phenotype by the combined analysis of clinicopathological and biological parameters.Methods: 77 patients with T1G3 or muscle-invasive UBC treated by radical cystectomy were studied. Immunohistochemistry was performed to detect CD147, heparanase, CD31 (blood vessels identification) and D2-40 (lymphatic vessels identification) expressions. The immunohistochemical reactions were correlated with the clinicopathological and the outcome parameters. 5-year disease-free survival (DFS) and overall survival (OS) rates were estimated using the Kaplan-Meier method. Multivariate analysis was performed by Cox proportional hazards analysis.Results: The 5-year DFS and OS rates were significantly influenced by the classical clinicopathological parameters, and by the occurrence of lymphovascular invasion. CD 147 and heparanase immunoexpression did not affect patients' outcome. However, patients with pT3/pT4 tumours had a median OS time of 14.7 months (95% CI 7.1-22.3, p = 0.003), which was reduced to 9.2 months (95% CI 1.5-17.0, p = 0.008) if the tumours were CD147 positive. We developed a model of tumour aggressiveness using parameters as stage, grade, lymphovascular invasion and CD147 immunoexpression, which separated a low aggressiveness from a high aggressiveness group, remaining as an independent prognostic factor of DFS (HR 3.746; 95% CI 1.244-11.285; p = 0.019) and OS (HR 3.247; 95% CI 1.015-10.388, p = 0.047).Conclusion: CD 147 overexpression, included in a model of UBC aggressiveness, may help surgeons to identify patients who could benefit from a personalized therapeutic regimen. Additional validation is needed. (C) 2011 Elsevier Ltd. All rights reserved.

Could albumin level explain the higher mortality in hemodialysis patients with pulmonary hypertension?

Background: The pathogenesis of pulmonary hypertension (PH) in hemodialysis is still unclear. The aim of thisstudy was to identify the risk factors associated with the presence of PH in chronic hemodialysis patients and toverify whether these factors might explain the highest mortality among them.Methods: We conducted a retrospective study of hemodialysis patients who started treatment from August 2001to October 2007 and were followed up until April 2011 in a Brazilian referral medical school. According to theresults of echocardiography examination, patients were allocated in two groups: those with PH and those withoutPH. Clinical parameters, site and type of vascular access, bioimpedance, and laboratorial findings were comparedbetween the groups and a logistic regression model was elaborated. Actuarial survival curves were constructed andhazard risk to death was evaluated by Cox regression analysis.Results: PH > 35 mmHg was found in 23 (30.6%) of the 75 patients studied. The groups differed in extracellularwater, ventricular thickness, left atrium diameter, and ventricular filling. In a univariate analysis, extracellular waterwas associated with PH (relative risk = 1.194; 95% CI of 1.006 1.416; p = 0.042); nevertheless, in a multiple model,only left atrium enlargement was independently associated with PH (relative risk =1.172; 95% CI of 1.010 1.359;p = 0.036). PH (hazard risk = 3.008; 95% CI of 1.285 7.043; p = 0.011) and age (hazard risk of 1.034 per year of age;95% CI of 1.000 7.068; p = 0.047) were significantly associated with mortality in a multiple Cox regression analysis.However, when albumin was taken in account the only statistically significant association was between albuminlevel and mortality (hazard risk = 0.342 per g/dL; 95% CI of 0.119 0.984; p = 0.047) while the presence of PH lost itsstatistical significance (p = 0.184). Mortality was higher in patients with PH (47.8% vs 25%) who also had astatistically worse survival after the sixth year of follow up.Conclusions: PH in hemodialysis patients is associated with parameters of volume overload that sheds light on itspathophysiology. Mortality is higher in hemodialysis patients with PH and the low albumin level can explain thisassociation.© 2012 Greenfield et al.; licensee BioMed Central Ltd.

Three-year follow-up of Interleukin 6 and C-reactive protein in chronic obstructive pulmonary disease

Background: Past studies have shown that mean values of Interleukin-6 (IL-6) and C-reactive protein (CRP) do not change significantly in COPD patients over a one-year period. However, longer period follow-up studies are still lacking. Thus, the aim of this study is to evaluate plasma CRP and IL-6 concentration over three years in COPD patients and to test the association between these inflammatory mediators and disease outcome markers. Methods: A cohort of 77 outpatients with stable COPD was evaluated at baseline, and 53 (mean FEV1, 56% predicted) were included in the prospective study. We evaluated Interleukin-6 (IL-6), C-reactive protein (CRP), six-minute walking distance (6MWD), and body mass index (BMI) at baseline and after three years. Plasma concentration of IL-6 was measured by high sensitivity ELISA, and CRP was obtained by high sensitivity particle-enhanced immunonephelometry. Results: IL-6 increased significantly after 3 years compared to baseline measurements [0.8 (0.5-1.3) vs 2.4 (1.3-4.4) pg/ml; p < 0.001] and was associated with worse 6MWD performance. In the Cox regression, increased IL-6 at baseline was associated with mortality [Hazard Ratio (95% CI) = 2.68 (0.13, 1.84); p = 0.02]. CRP mean values did not change [5 (1.6-7.9) vs 4.7 (1.7-10) pg/L; p = 0.84], although eleven patients (21%) presented with changes >3 mg/L in CRP after 3 years. Conclusions: The systemic inflammatory process, evaluated by IL-6, seems to be persistent, progressive and associated with mortality and worse physical performance in COPD patients. Trial registration: No.:NCT00605540. © 2013 Ferrari et al; licensee BioMed Central Ltd.

A Cohort Study of Caries Incidence and Baseline Socioeconomic, Clinical and Demographic Variables: A Kaplan-Meier Survival Analysis

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Predictors of first-year survival in patients with advanced COPD treated using long-term oxygen therapy

Little evidence-based guidance is available to aid clinicians in determining short-term prognoses in very severe COPD patients. Therefore, the present study was designed to provide a prospective assessment (1) of the mortality rates and (2) whether the baseline measurements may be determinants of 1-year mortality in hypoxemic COPD patients receiving long-term oxygen therapy (LTOT).Seventy-eight clinically stable patients with advanced COPD treated using LTOT were enrolled in a prospective cohort study. Outcome variable: first-year mortality. Baseline measurements: categorical variables: age (<60 or >= 60 years); gender; body mass index (<20 or >= 20 kg/m(2)); fat-free mass (FFM) index (<16 [men] and <15kg/m(2) [women]; baseline dyspnea index (BDI) (<= 3 or >3); and corticosteroid use. Continuous variables: smoking history; lung function; FFM; fat mass; hemoglobin; hematocrit; arterial blood gases; forearm muscle strength; St. George's Respiratory Questionnaire (SGRQ); and comorbidity score. By the end of 1-year of follow-up, 12 patients (15.4%) had died. Kaplan-Meier curves showed that BDI <= 3 was the only variable associated with higher mortality. Cox proportional hazards analysis revealed that tower PaO2 and SPO2, higher PaCO2 and SGRQ scores were associated with reduced survival. In the multivariate analysis, BDI remained predictive of mortality (hazard ratio [HR], 0.50; 95% confidence interval [CI], 0.31-0.81), as did PaO2 (HR, 0.49; 95% CI, 0.26-0.95). These data suggest that readily available parameters as dyspnea intensity and hypoxemia severity may be useful in predicting first-year survival rates in advanced COPD patients receiving LTOT (C) 2007 Elsevier Ltd. All rights reserved.

Desigualdade da sobrevivência de trabalhadores de Botucatu, São Paulo, Brasil

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Shorter CAG repeat length in the AR gene is associated with poor outcome in head and neck cancer

Objective: Alterations in the size of the [CAG](n) repeats of the AR gene have been described in several types tumors. The purpose of this study was to evaluate if there is an association between the AR [CAG](n) repeat alleles and the relative risk for head and neck cancer and to analyse microsatellite instability (MSI) and loss of heterozygosity (LOH) in these tumors.Design: Matched samples of blood and head and neck tumors were evaluated using two methodologies, silver-stained gels to perform the analyses of MSI and LOH, and automated analysis to confirm these results and for genotyping of the AR [CAG](n), repeat length. Sixty-nine individuals without cancer were used as a control group for both procedures. The Log-rank test was used to compare overall survival and disease-free survival curves. The Cox proportional hazards regression models were performed to determine the [CAG], repeats as an independent prognostic factor.Results: Patients with alleles <= 20 in the male group showed a correlation with lower disease-free survival (P = 0.0325) and with recurrence or metastasis (RR 2.52, CI 95%). in the female group, the allele 2 (longer allele) showed a significant lower mean of [CAG](n), repeat when compared to the control group. Microsatellite instability was detected in nine cases in both procedures. In six out of these nine cases, we observed a reduction of the AR [CAG](n) repeat length. LOH was detected in one out of 17 women informative for oral cancer in both procedures.Conclusion: These results suggest that short [CAG](n) repeat length (: 20) polymorphism is associated with poor prognosis in a subset of male patients with head and neck cancer and that AR gene microsatellite instability is uncommon in these tumors. (C) 2007 Elsevier Ltd. All rights reserved.

Modeling grouped survival data with time-dependent covariates

In this article, proportional hazards and logistic models for grouped survival data were extended to incorporate time-dependent covariates. The extension was motivated by a forestry experiment designed to compare five different water stresses in Eucalyptus grandis seedlings. The response was the seedling lifetime. The data set was grouped since there were just three occasions in which the seedlings was visited by the researcher. In each of these occasions also the shoot height was measured and therefore it is a time-dependent covariate. Both extended models were used in this example, and the results were very similar.

Estudo de sobrevivência de uma coorte de pessoas de 60 anos e mais no município de Botucatu (SP) - Brasil

O aumento proporcional do número de idosos na população tem motivado estudos no sentido de melhorar a qualidade de vida desta faixa etária através de políticas sociais e, entre elas, o planejamento em saúde. Com o objetivo de conhecer riscos de mortalidade para a população de sessenta anos e mais, um estudo de sobrevida foi realizado rastreando, no ano de 1992, os idosos participantes de um inquérito de morbidade referida realizado na cidade de Botucatu em 1983/84. Foram localizados 89,6% destes idosos. Curvas de sobrevivência foram calculadas com o método de Kaplan-Meier e a análise de riscos, utilizando-se a Regressão Múltipla de Cox ajustando-se o modelo agregando as variáveis por blocos. Para o sexo masculino foram encontradas associadas, independentemente, ao aumento da mortalidade as seguintes categorias de variáveis: idade de 70 anos e mais: Hazard Ratio (HR)=2,4 (1,6 - 3,7); salário menor que um salário mínimo: HR=2,2 (1,3 - 3,8); ter outras rendas: HR=2,2 (1,3 - 3,9); ser o chefe da família ou seu cônjuge: HR=2,3 (1,2 - 2,4); referência de doenças do aparelho circulatório: HR=1,6 (1,1 - 2,4); referência de diabetes mellitus: HR=3,0 (1,3 - 7,0). Para o sexo feminino, foram encontradas associadas a idade de 70 anos e mais: HR=4,6 (3,0 - 7,1); referência de diabetes mellitus: HR=3,0 (1,7-5,3) e ter outras rendas: HR=2,0 (1,1 - 4,0).