Cutpoint selection for discretizing a continuous covariate for generalized estimating equations

We consider consider the problem of dichotomizing a continuous covariate when performing a regression analysis based on a generalized estimation approach. The problem involves estimation of the cutpoint for the covariate and testing the hypothesis that the binary covariate constructed from the continuous covariate has a significant impact on the outcome. Due to the multiple testing used to find the optimal cutpoint, we need to make an adjustment to the usual significance test to preserve the type-I error rates. We illustrate the techniques on one data set of patients given unrelated hematopoietic stem cell transplantation. Here the question is whether the CD34 cell dose given to patient affects the outcome of the transplant and what is the smallest cell dose which is needed for good outcomes. (C) 2010 Elsevier BM. All rights reserved.

Local influence in estimating equations

Local influence diagnostics based on estimating equations as the role of a gradient vector derived from any fit function are developed for repeated measures regression analysis. Our proposal generalizes tools used in other studies (Cook, 1986: Cadigan and Farrell, 2002), considering herein local influence diagnostics for a statistical model where estimation involves an estimating equation in which all observations are not necessarily independent of each other. Moreover, the measures of local influence are illustrated with some simulated data sets to assess influential observations. Applications using real data are presented. (C) 2010 Elsevier B.V. All rights reserved.

Analysis of Melanoma Onset: Assessing Familial Aggregation by Using Estimating Equations and Fitting Variance Components via Bayesian Random Effects Models

We investigate whether relative contributions of genetic and shared environmental factors are associated with an increased risk in melanoma. Data from the Queensland Familial Melanoma Project comprising 15,907 subjects arising from 1912 families were analyzed to estimate the additive genetic, common and unique environmental contributions to variation in the age at onset of melanoma. Two complementary approaches for analyzing correlated time-to-onset family data were considered: the generalized estimating equations (GEE) method in which one can estimate relationship-specific dependence simultaneously with regression coefficients that describe the average population response to changing covariates; and a subject-specific Bayesian mixed model in which heterogeneity in regression parameters is explicitly modeled and the different components of variation may be estimated directly. The proportional hazards and Weibull models were utilized, as both produce natural frameworks for estimating relative risks while adjusting for simultaneous effects of other covariates. A simple Markov Chain Monte Carlo method for covariate imputation of missing data was used and the actual implementation of the Bayesian model was based on Gibbs sampling using the free ware package BUGS. In addition, we also used a Bayesian model to investigate the relative contribution of genetic and environmental effects on the expression of naevi and freckles, which are known risk factors for melanoma.

Effect of n-3 fatty acids on markers of brain injury and incidence of sepsis-associated delirium in septic patients.

BACKGROUND: Data regarding immunomodulatory effects of parenteral n-3 fatty acids in sepsis are conflicting. In this study, the effect of administration of parenteral n-3 fatty acids on markers of brain injury, incidence of sepsis-associated delirium, and inflammatory mediators in septic patients was investigated. METHODS: Fifty patients with sepsis were randomized to receive either 2 ml/kg/day of a lipid emulsion containing highly refined fish oil (equivalent to n-3 fatty acids 0.12 mg/kg/day) during 7 days after admission to the intensive care unit or standard treatment. Markers of brain injury and inflammatory mediators were measured on days 1, 2, 3 and 7. Assessment for sepsis-associated delirium was performed daily. The primary outcome was the difference in S-100β from baseline to peak level between both the intervention and the control group, compared by t-test. Changes of all markers over time were explored in both groups, fitting a generalized estimating equations model. RESULTS: Mean difference in change of S-100β from baseline to peak level was 0.34 (95% CI: -0.18-0.85) between the intervention and control group, respectively (P = 0.19). We found no difference in plasma levels of S-100β, neuron-specific enolase, interleukin (IL)-6, IL-8, IL-10, and C-reactive protein between groups over time. Incidence of sepsis-associated delirium was 75% in the intervention and 71% in the control groups (risk difference 4%, 95% CI -24-31%, P = 0.796). CONCLUSION: Administration of n-3 fatty acids did not affect markers of brain injury, incidence of sepsis-associated delirium, and inflammatory mediators in septic patients.

La salmonellose chez les bovins laitiers : présentation clinique et culture bactériologique

Huit cent trente et un troupeaux de vaches laitières répartis dans 5 états américains ont été enrôlés dans une étude de cohorte prospective. Un modèle d’équations d'estimation généralisées a été utilisé pour étudier l'association entre les signes cliniques et la détection de salmonelles dans les fèces des animaux soupçonnés de salmonellose clinique. La sensibilité et la spécificité de la culture bactériologique ont été estimées à l’aide d’un modèle de classes latentes. Dix-huit pour cent des 874 échantillons provenant de veaux et 29% des 1479 échantillons de vaches adultes étaient positifs pour Salmonella spp. Il n’a pas été possible d’établir une association claire entre les différents signes cliniques observés et la détection de salmonelles. Les 2 sérotypes les plus fréquemment isolés étaient Typhimurium et Newport. La probabilité de détecter des salmonelles était plus élevée chez les veaux où un autre agent entéropathogène était également détecté. La proportion d’échantillons positifs était plus élevée parmi les vaches ayant reçu des antibiotiques dans les jours précédant l’échantillonnage. La sensibilité de la culture a été estimée à 0,48 (intervalle de crédibilité à 95% [ICr95%]: 0,22-0,95) pour les veaux et 0,78 (ICr95%: 0,55-0,99) pour les vaches. La spécificité de la culture était de 0,94 (ICr95%: 0,87-1,00) pour les veaux et de 0,96 (ICr95%: 0,90-1,00) pour les vaches. Malgré une sensibilité imparfaite, la culture bactériologique demeure utile pour obtenir une meilleure estimation de la probabilité post-test de salmonellose clinique chez un bovin laitier, par rapport à la probabilité estimée suite au seul examen clinique.

Bayesian analysis for a skew extension of the multivariate null intercept measurement error model

Skew-normal distribution is a class of distributions that includes the normal distributions as a special case. In this paper, we explore the use of Markov Chain Monte Carlo (MCMC) methods to develop a Bayesian analysis in a multivariate, null intercept, measurement error model [R. Aoki, H. Bolfarine, J.A. Achcar, and D. Leao Pinto Jr, Bayesian analysis of a multivariate null intercept error-in -variables regression model, J. Biopharm. Stat. 13(4) (2003b), pp. 763-771] where the unobserved value of the covariate (latent variable) follows a skew-normal distribution. The results and methods are applied to a real dental clinical trial presented in [A. Hadgu and G. Koch, Application of generalized estimating equations to a dental randomized clinical trial, J. Biopharm. Stat. 9 (1999), pp. 161-178].

Inference and local influence assessment in skew-normal null intercept measurement error model

In this article, we discuss inferential aspects of the measurement error regression models with null intercepts when the unknown quantity x (latent variable) follows a skew normal distribution. We examine first the maximum-likelihood approach to estimation via the EM algorithm by exploring statistical properties of the model considered. Then, the marginal likelihood, the score function and the observed information matrix of the observed quantities are presented allowing direct inference implementation. In order to discuss some diagnostics techniques in this type of models, we derive the appropriate matrices to assessing the local influence on the parameter estimates under different perturbation schemes. The results and methods developed in this paper are illustrated considering part of a real data set used by Hadgu and Koch [1999, Application of generalized estimating equations to a dental randomized clinical trial. Journal of Biopharmaceutical Statistics, 9, 161-178].

A comparison of estimated glomerular filtration rates using Cockcroft−Gault and the Chronic Kidney Disease Epidemiology Collaboration estimating equations in HIV infection

OBJECTIVES: The aim of this study was to determine whether the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI)- or Cockcroft-Gault (CG)-based estimated glomerular filtration rates (eGFRs) performs better in the cohort setting for predicting moderate/advanced chronic kidney disease (CKD) or end-stage renal disease (ESRD). METHODS: A total of 9521 persons in the EuroSIDA study contributed 133 873 eGFRs. Poisson regression was used to model the incidence of moderate and advanced CKD (confirmed eGFR < 60 and < 30 mL/min/1.73 m(2) , respectively) or ESRD (fatal/nonfatal) using CG and CKD-EPI eGFRs. RESULTS: Of 133 873 eGFR values, the ratio of CG to CKD-EPI was ≥ 1.1 in 22 092 (16.5%) and the difference between them (CG minus CKD-EPI) was ≥ 10 mL/min/1.73 m(2) in 20 867 (15.6%). Differences between CKD-EPI and CG were much greater when CG was not standardized for body surface area (BSA). A total of 403 persons developed moderate CKD using CG [incidence 8.9/1000 person-years of follow-up (PYFU); 95% confidence interval (CI) 8.0-9.8] and 364 using CKD-EPI (incidence 7.3/1000 PYFU; 95% CI 6.5-8.0). CG-derived eGFRs were equal to CKD-EPI-derived eGFRs at predicting ESRD (n = 36) and death (n = 565), as measured by the Akaike information criterion. CG-based moderate and advanced CKDs were associated with ESRD [adjusted incidence rate ratio (aIRR) 7.17; 95% CI 2.65-19.36 and aIRR 23.46; 95% CI 8.54-64.48, respectively], as were CKD-EPI-based moderate and advanced CKDs (aIRR 12.41; 95% CI 4.74-32.51 and aIRR 12.44; 95% CI 4.83-32.03, respectively). CONCLUSIONS: Differences between eGFRs using CG adjusted for BSA or CKD-EPI were modest. In the absence of a gold standard, the two formulae predicted clinical outcomes with equal precision and can be used to estimate GFR in HIV-positive persons.

A comparison of Markov and generalized linear models of hospital mortality

This paper reports a comparison of three modeling strategies for the analysis of hospital mortality in a sample of general medicine inpatients in a Department of Veterans Affairs medical center. Logistic regression, a Markov chain model, and longitudinal logistic regression were evaluated on predictive performance as measured by the c-index and on accuracy of expected numbers of deaths compared to observed. The logistic regression used patient information collected at admission; the Markov model was comprised of two absorbing states for discharge and death and three transient states reflecting increasing severity of illness as measured by laboratory data collected during the hospital stay; longitudinal regression employed Generalized Estimating Equations (GEE) to model covariance structure for the repeated binary outcome. Results showed that the logistic regression predicted hospital mortality as well as the alternative methods but was limited in scope of application. The Markov chain provides insights into how day to day changes of illness severity lead to discharge or death. The longitudinal logistic regression showed that increasing illness trajectory is associated with hospital mortality. The conclusion is reached that for standard applications in modeling hospital mortality, logistic regression is adequate, but for new challenges facing health services research today, alternative methods are equally predictive, practical, and can provide new insights. ^

Dependence of the crossover exponent with the diffusion rate in the generalized contact process model

We study how the crossover exponent, phi, between the directed percolation (DP) and compact directed percolation (CDP) behaves as a function of the diffusion rate in a model that generalizes the contact process. Our conclusions are based in results pointed by perturbative series expansions and numerical simulations, and are consistent with a value phi = 2 for finite diffusion rates and phi = 1 in the limit of infinite diffusion rate.

Fatores associados à recuperação nutricional de crianças em programa de suplementação alimentar

OBJETIVOS: Identificar os fatores associados à recuperação nutricional de crianças inscritas no Programa de Incentivo ao Combate às Carências Nutricionais (ICCN) no Município de Mogi das Cruzes. MÉTODOS: Foi realizado um ensaio institucional não controlado com 570 crianças inscritas no ICCN, que foram seguidas de julho de 1999 a julho de 2001. O estado nutricional foi avaliado segundo índice altura/idade, sendo consideradas eutróficas as crianças com escore z > - 1; de risco as que apresentaram z > - 2 e < - 1; desnutridas moderadas aquelas com z < - 2 e > - 3; e desnutridas graves as que apresentavam z < - 3. O impacto do ICCN foi analisado através de modelo multivariado com o uso de equações de estimação (GEE - Generalized Estimating Equations), sendo considerado significativo p<0,05. RESULTADOS/CONCLUSÃO: Ao final do seguimento, houve a melhora nutricional das crianças, com um gradiente, sendo maior a recuperação quanto maior a deficiência nutricional inicial. Observaram-se ganhos em altura de 1,12, 0,82, 0,57 e 0,45 desvios-padrão para as desnutridas graves, moderadas, em risco nutricional e eutróficas, respectivamente. Os fatores associados à evolução nutricional das crianças desnutridas foram a idade de 12 a 24 meses ao ingressar no ICCN, o peso ao nascer igual ou superior a 3 kg e o aleitamento materno. Os fatores associados negativamente à evolução nutricional neste grupo foram idade da mãe entre 20 a 40 anos e a ausência de trabalho remunerado materno. Para as crianças em risco, a renda familiar também se mostrou associada à melhor evolução nutricional. A experiência do ICCN em Mogi das Cruzes sugere que os programas de suplementação alimentar têm papel relevante na recuperação nutricional de desnutridos

Exhaled nitric oxide as a predictor of exacerbation in children with moderate-to-severe asthma: a prospective, 5-month study

Background: Inhaled corticosteroids (ICSs) are recommended as the first line of treatment in children with moderate-to-severe asthma. Exhaled nitric oxide (ENO) has been proposed as a clinically useful marker of control that might help identify patients in whom ICS dose may be safely reduced. Objective: To evaluate the ability of ENO to predict future asthma exacerbations in children with moderate-to-severe asthma undergoing ICS tapering. Methods: This is an observational study with no control group. ENO was measured biweekly for 14 weeks in 32 children with moderate-to-severe asthma who were undergoing ICS tapering. Clinical evaluations and spirometry were performed concomitantly, and families kept daily diaries to record symptoms between visits. We used generalized estimating equations to model the In (odds) of an asthma exacerbation in the subsequent 2-week interval as a function of ENO level at the start of the interval while adjusting for age, sex, asthma severity, and current medication use. Results: We were able to successfully lower ICS doses in 10 (56%) of the 18 children with moderate asthma and in 3 (21%) of the 14 children with severe asthma. In 83 of the 187 follow-up clinical evaluations, children were determined to have had an exacerbation during the preceding 2 weeks. ENO levels, whether expressed as a continuous variable or dichotomized, were not associated with future risk for exacerbations in either unadjusted or adjusted models. Conclusion: ENO was not a useful clinical predictor of future asthma exacerbations for children with moderate-to-severe asthma undergoing ICS tapering. Ann Allergy Asthma Immunol. 2009; 103:206-211.

Long-term nutritional assessment of patients with severe short bowel syndrome managed with home enteral nutrition and oral intake

Background: Parenteral nutrition (PN) is used to control the nutritional state after severe intestinal resections. Whenever possible, enteral nutrition (EN) is used to promote intestinal rehabilitation and reduce PN dependency. Our aim is to verify whether EN + oral intake (01) in severe short bowel syndrome (SBS) surgical adult patients can maintain adequate nutritional status in the long term. Methods: This longitudinal retrospective study included 10 patients followed for 7 post-operative years. Body mass index (BMI), percentage of involuntary loss of usual body weight (UWL), free fat mass (FFM), and fat mass (FM) composition assessed by bioelectric impedance, and laboratory tests were evaluated at 6, 12, 24, 36, 48, 60, 72, and 84 months after surgery. Energy and protein offered in HPN and at long term by HEN+ oral intake (01), was evaluated at the same periods. The statistical model of generalized estimating equations with p <0,05 was used. Results: With long term EN + 01 there was a progressive increase in the UWL, a decrease in BMI, FFM, and FM (p < 0,05). PN weaning was possible in eight patients. Infection due to central venous catheter (CVC) contamination was the most common complication (1.2 episodes CVC/patient/year). There was an increase in energy and protein intake supply provided by HEN+OI (p <0.05). All patients survived for at least 2 years, seven for 5 years and six for 7 years of follow-up. Conclusions: In the long term SBS surgical adult patients fed with HEN+OI couldn`t maintain adequate nutritional status with loss of FM and FFM. (Nutr Hosp. 2011;26:834-842) DOI:10.3305/nh.2011.26.4.5153