L’imputation multiple des données manquantes aléatoirement : concepts généraux et présentation d’une méthode Monte-Carlo

BackgroundStatistical analysis of a data set with missing data is a frequent problem to deal with in epidemiology. Methods are available to manage incomplete observations, avoiding biased estimates and improving their precision, compared to more traditional methods, such as the analysis of the sub-sample of complete observations.MethodsOne of these approaches is multiple imputation, which consists in imputing successively several values for each missing data item. Several completed data sets having the same distribution characteristics as the observed data (variability and correlations) are thus generated. Standard analyses are done separately on each completed dataset then combined to obtain a global result. In this paper, we discuss the various assumptions made on the origin of missing data (at random or not), and we present in a pragmatic way the process of multiple imputation. A recent method, Multiple Imputation by Chained Equations (MICE), based on a Monte-Carlo Markov Chain algorithm under missing at random data (MAR) hypothesis, is described. An illustrative example of the MICE method is detailed for the analysis of the relation between a dichotomous variable and two covariates presenting MAR data with no particular structure, through multivariate logistic regression.ResultsCompared with the original dataset without missing data, the results show a substantial improvement of the regression coefficient estimates with the MICE method, relatively to those obtained on the dataset with complete observations.ConclusionThis method does not require any direct assumption on joint distribution of the variables and it is presently implemented in standard statistical software (Splus, Stata). It can be used for multiple imputation of missing data of several variables with no particular structure.     

In vitro cytotoxicity of a remineralizing resin-based calcium phosphate cement.

Recently, a resin-based calcium phosphate cement (RCPC) has been reported as a remineralizing pulp-capping or lining cement. RCPC consists mainly of tetracalcium and dicalcium phosphates, ethoxylated bisphenol A dimethacrylate and pyromellitic glycerol dimethacrylate monomers and photo- and chemical initiators. OBJECTIVES: Here, the cytotoxic effects of RCPC were evaluated. The hypothesis was that RCPC induced only minor cytotoxic response in immortalized murine odontoblast and pulp cells, comparable to that produced by similar dimethacrylates due to unpolymerized dimethacrylate monomer present after curing. METHODS: Cytotoxicity was determined following the changes in cell succinate dehydrogenase activity after 24 h exposure to the cement components and after a 24 h recovery period. A fourfold range of concentrations was tested of the monomers, the eluate of cured RCPC leached in Dulbecco's modified Eagle's medium, and crushed cured cement in dimethyl sulfoxide. RESULTS: The monomers themselves had cytotoxicities similar to those reported for other dimethacrylates, although they are significantly less toxic than Bis-GMA. Differential cell sensitivity was demonstrated, with the pulp cells having greater sensitivity to the unpolymerized monomer than the odontoblast-like cells. The leached components have cytotoxicity similar to that of the free monomers. The crushed material demonstrated no apparent cytotoxicity at the dilutions tested. SIGNIFICANCE: These data demonstrate that RCPC has an in vitro cytotoxicity that is comparable to other materials containing dimethacrylate monomers and suggest that the material may be suitable for use in dental restorations. The data also indicate that the pulp cells appear more sensitive to dimethacrylates than the odontoblasts.     

The surgical risk of colectomy in patients with cirrhosis

The records of 54 patients with documented cirrhosis who underwent colectomy between January 1970 and January 1984 were studied to assess the operative risk and to determine the preoperative predictive risk factors. In-hospital mortality was 24 percent (13 patients), and postoperative complications occurred in 48 percent (26 patients). The risk of surgical intervention was significantly increased if encephalopathy, ascites, anemia, or hypoalbuminemia was present before operation. A simple operative risk index involving the presence of encephalopathy and ascites and the levels of hemoglobin and albumin is proposed to help distinguish a low-risk subgroup in whom postoperative mortality was 12.8 percent from a high-risk subgroup in whom postoperative mortality was 53.3 percent.