RésuméL'approche mixte, pour la modélisation des processus de survie et des processus longitudinaux, connait de nos jours un succès grandissant, en particulier lorsqu';il existe une interdépendance entre ceux‐ci. Réputés pour leur efficacité, les modèles mixtes sont moins biaisés par rapport aux méthodes classiques naïves, d'où une littérature abondante sur le sujet. L'objet de cette étude est de présenter une revue de la littérature récente sur les modèles mixtes, en l'occurrence ceux utilisés pour l'estimation du temps d'échec dans les processus de survie. Nous présenterons une analyse détaillée de la gamme de modèles de survie utilisés pour la construction des modèles mixtes. Une attention particulière sera accordée aux progrès récents dans le développement des logiciels dédiés à la construction de ces modéles. Une illustration sur l'utilisation des progiciels JM et JoineR, développés dans l'environnement R, sera présentée pour le cas de l'analyse de survie des patients souffrant d'insuffisance rénale en phase terminale. Finalement, nous concluons avec une discussion sur plusieurs pistes de recherche potentielles dans le domaine.

Opacity is a property of many plasmas, and it is normally expected that if an emission line in a plasma becomes optically thick, its intensity ratio to that of another transition that remains optically thin should decrease. However, radiative transfer calculations undertaken both by ourselves and others predict that under certain conditions the intensity ratio of an optically thick to thin line can show an increase over the optically thin value, indicating an enhancement in the former. These conditions include the geometry of the emitting plasma and its orientation to the observer. A similar effect can take place between lines of differing optical depth. Previous observational studies have focused on stellar point sources, and here we investigate the spatially-resolved solar atmosphere using measurements of the I(1032Å)/I(1038Å) intensity ratio of O VI in several regions obtained with the Solar Ultraviolet Measurements of Emitted Radiation (SUMER) instrument on board the Solar and Heliospheric Observatory (SoHO)satellite. We find several I(1032Å)/I(1038Å) ratios observed on the disk to be significantly larger than the optically thin value of 2.0, providing the first detection (to our knowledge) of intensity enhancement in the ratio arising from opacity effects in the solar atmosphere. Agreement between observation and theory is excellent, and confirms that the O VI emission originates from a slablike geometry in the solar atmosphere, rather than from cylindrical structures.

No abstract

A349 transformed to common logarithms. Costs are stated in € (1€ = 25.5CZK). Results: A total 100 patients (about 10-15% of all HIV treated in the Czech Republic) were included. The mean age was 41.9 years (range 23-71), time since diagnosis 5.8 years (range 1-26), and period of assessment 30.5 months. Average total annual costs were highest for CD4+ <200 (22,905€ ; p<0.05) vs. CD4+ 200-499 (17,801€) and CD4+ >500 (17,043€). HIV medication accounted for the majority of costs in all three subgroups (range 62% in CD4+ <200 to 88% in CD4+ >500). Patients with a disease duration of >10 years had highest annual costs of 22,841€ (p<0.05) compared to 15,783€ (duration 5-10 years) and 16,777€ (<5 years). HIV-medication accounted for the majority of costs here as well (from 70% to 88%; NS). In this analysis 17 % of costs in patients with disease duration of >10 years were due to hospitalization vs. 5 % in both other groups (NS). ConClusions: Similarly to published literature, we found highest costs in patients with low CD4+ lymphocytes and longer disease duration.

No abstract