Intérêt de la détection et de la prise en compte d’une surdispersion dans un modèle de Poisson : illustration à partir d’un exemple

BackgroundResearchers often use the Poisson regression model to analyze count data. Overdispersion can occur when a Poisson regression model is used, resulting in an underestimation of variance of the regression model parameters. Our objective was to take overdispersion into account and assess its impact with an illustration based on the data of a study investigating the relationship between use of the Internet to seek health information and number of primary care consultations.MethodsThree methods, overdispersed Poisson, a robust estimator, and negative binomial regression, were performed to take overdispersion into account in explaining variation in the number (Y) of primary care consultations. We tested overdispersion in the Poisson regression model using the ratio of the sum of Pearson residuals over the number of degrees of freedom (χ²/df). We then fitted the three models and compared parameter estimation to the estimations given by Poisson regression model.ResultsVariance of the number of primary care consultations (Var[Y] = 21.03) was greater than the mean (E[Y] = 5.93) and the χ²/df ratio was 3.26, which confirmed overdispersion. Standard errors of the parameters varied greatly between the Poisson regression model and the three other regression models. Interpretation of estimates from two variables (using the Internet to seek health information and single parent family) would have changed according to the model retained, with significant levels of 0.06 and 0.002 (Poisson), 0.29 and 0.09 (overdispersed Poisson), 0.29 and 0.13 (use of a robust estimator) and 0.45 and 0.13 (negative binomial) respectively.ConclusionDifferent methods exist to solve the problem of underestimating variance in the Poisson regression model when overdispersion is present. The negative binomial regression model seems to be particularly accurate because of its theorical distribution ; in addition this regression is easy to perform with ordinary statistical software packages.     

Évolution du statut nutritionnel d’une cohorte d’enfants tunisiens au cours de l’hospitalisation : facteurs de risque de la dénutrition hospitalière

ObjectivesStudy the evolution of the nutritional status of a cohort of hospitalized children and identify the risk factors of hospital malnutrition.MethodsProspective, cross-sectional study carried out in a pediatric department over a period of six months, including all children aged ≥ 30 days, hospitalized for a period ≥ six days. Anthropometric data were assessed on admission and discharge. Food consumption was assessed using the flower tool. We identified the risk factors for hospital undernutrition (HUN) by multivariate analysis.ResultsWe included 120 patients with a mean age 46.3 months. The prevalence of acute undernutrition at admission was 21.7% and that of chronic undernutrition was 10%. The prevalence of acute undernutrition at discharge rose to 34%. Weight loss during hospitalization was noted in 68.3% of cases. The prevalence of HUN was 55% considering a decrease in BMI or P/PAT z-score ≥ 25%. The risk factors for HUN were: age ≤ 24 months (P = 0.039; OR 95% CI = 2.67 [1.05–6.82]), the presence of undernutrition on admission (P = 0.002; OR 95% CI = 2.32 [0.93–6.51]) and average food consumption < 50% during hospitalization (P < 10^?3; OR 95% CI = 6.69 [2.57–17.40]).ConclusionScreening for undernutrition on admission to hospital as well as assessment of the nutritional risk in hospitalized children is essential so that preventive or curative nutritional care can be taken.     

Quels freins à l’établissement du certificat de décès à domicile ? Enquête auprès de médecins sur le territoire de Normandie

IntroductionDrafting a death certificate is an obligation for any physician called upon to note a death. It has a legal and epidemiological impact. The objective of this study is to highlight the obstacles encountered by professionals when this process takes place in a home setting.MethodA study was carried out by means of a questionnaire sent to doctors in the Normandy region.ResultsOne fifth (20%) of the general practitioners (GPs), 25% of the SAMU practitioners (intervening in mobile emergency throughout the country), 27.3% of SOS doctors (intervening 24 hours a day in the deceased person's home in the absence of a GP) stated that they had at least once refused to establish a death certificate. Only 36.4% of SOS doctors, 62.5% of emergency physicians and 41% of GPs considered the drafting of a death certificate to be one of their mandatory missions. No less than 17% of GPs, 25% of SAMU doctors and 9% of SOS doctors stated that due to lack of time, they were occasionally unable to travel to fill out a death certificate, a factor causing delays that were detrimental, especially for the surviving relatives. We highlighted several other obstacles to the completion of death certificates: the difficulty of completing them accurately without knowledge of the deceased person's history, possible inability of professionals on tight schedules to deal with unforeseen requests and, more rarely, the issues raised by an absence of remuneration.ConclusionOur study highlights major obstacles in the preparation of death certificates, obstacles that should be taken into consideration insofar as they affect the quality of the data collected, and entail ethical challenges.