Please activate JavaScript!
Please install Adobe Flash Player, click here for download

ISCB2014_abstract_book

104 ISCB 2014 Vienna, Austria • Abstracts - Poster PresentationsTuesday, 26th August 2014 • 10:30-11:00 Monday25thAugustTuesday26thAugustThursday28thAugustAuthorIndexPostersWednesday27thAugustSunday24thAugust Main Outcome Measure: Under five child mortality was the primary end point. Results: Increase in Mother’s age corresponded with an increase in child survival. Breastfeeding increases survival significantly by 71% (HR 0.29, 0.24-0.34, p=0.001). Standard of Living improved 22% child survival with medium (HR 0.78, 0.65-0.93, 0.005) and 43% with high (HR 0.57, 0.44- 0.74, p=0.001) standard of living in comparison to low standard of living. Received prenatal care and breastfeeding health nutrition education were found significant protective factors for child mortality (HR 0.42, 0.31-0.67, p=0.001) and (HR 0.45, 0.31-0.67, p=0.001) respectively. Conclusion: In a nationally representative sample of households in India, mother’s age, breastfeeding, standard of living, prenatal care and breast- feeding Health Nutrition education were associated in reducing child mortality.   P2.1.161 Disease evolution of spinocerebellar ataxia type 2 patients: interruption of follow-up considerations S Tezenas du Montcel1,2 , S Molho2 , H Jacobi3 , T Klockgether3 1 UMPC - Paris 6, Paris, France, 2 Groupe Hospitalier Pitié-Salpétrière, AH-HP, Paris, France, 3 University Hospital of Bonn, Bonn, Germany   Aim: Spinocerebellar ataxia type 2 is a genetically heterogeneous ataxia disorder with a well-established clinical phenotype. A European cohort was formed to establish the disease evolution of some spinocerebellar ataxias. Follow-up was continued in order to establish long-term follow- up. However some patients were lost of follow-up due to center or pa- tients reasons. The aim of the current study is to establish long term dis- ease evolution taking in account the follow-up interruptions. Method: Patients were included between May 2005 and September 2006 from 18 European centers and followed on a yearly basis for the first 3 years and then based on center practice. All data closed at 2013/08/15 were analyzed. Linear mixed models were used to analyze the longitudi- nal data. Results: From the 146 patients followed at least one year (2 to 8 visits, mean: 2 years ± 2), 26% patients were followed less than 3 years: 11% pa- tient withdrawn consent, 27% patients interrupted for disease reasons. The patients followed less than 3 years appeared more severe at baseline. The missing completely at random hypothesis was rejected. As a linear evolution could be evoked, we considered a missing at random process. Using pattern mixture models, we estimated that the disease evolution was of 1.46 ± 0.08 per year and 1.41± 0.08 after adjustment by the initial score. Age at onset and the genotype were associated with the disease evolution. Conclusion: Taking in account the missing process we were able to esti- mate the unbiased disease evolution and the determinants of this evolu- tion. For the Eurosca Group.   P2.2 Methodological issues and case studies in epidemiology P2.2.16 Use of propensity score in plastic and reconstructive research; rare complication events in elective surgeries S Bagher1 , T Zhong1,2 , SWMC Maass1 , AC O’Neill1,2 , SOP Hofer1,2 1 Plastic & Reconstructive Surgery, University Health Network, Toronto, Canada, 2 Plastic & Reconstructive Surgery, University of Toronto, Toronto, Canada   Aim: This study introduces propensity score (PS) methods and their ap- plication in studies with small sample size and rare events. A real-case dataset was evaluated, exploring two breast reconstruction (BR) methods (MS-Tram and DIEP) and their postoperative complication rates. Method: Data was pulled from a prospectively maintained institutional database. Regression adjustment with and without PS method was stud- ied. Complications were categorized as abdominal or breast. One PS was estimated for both breast and abdominal complications. The choice of variables to include in the PS model was examined (treatment, outcome or both); final variables used in the estimation of PS were associated with both treatment and outcome. To capture the effect of other key covariates their association was exam- ined in the final multivariate model and they were not included in the treatment effect model. Inclusion of interaction terms was examined. For missing data other than outcome, a multiple imputation procedure was performed. Complete cases analyses were performed as sensitivity test. Results: There were 83 (28%) complications, (20% breast; 8% abdomen). Using PS, the adjusted odds of abdominal complications were still signifi- cantly higher in free MS-TRAM vs. DIEP flaps (OR=2.73). With prior chemo- therapy, increase in BMI significantly increased abdominal complications (OR=1.16).There was no significant association between reconstruction method and breast complications; diabetics had significantly increased breast complications (OR=4.19). Conclusion: To handle the risk of calculation bias, PS methods can be used in evaluating elective surgeries with rare events. PS analysis indicated sig- nificantly higher abdominal complications in free MS-TRAM compared to DIEP flaps. P2.2.29 Pharmacoepidemiological characteristics of influenza-like illness in hospitalized children in Russia: a case-control study VA Bulgakova1 , VF Uchaikin2 , VV Maleev3 , OV Shamsheva2 , EA Osipova4 , IV Prostyakov4 1 Scientific Centre of Children’s Health, Moscow, Russian Federation, 2 Russian National Research Medical University, Moscow, Russian Federation, 3 Central Research Institute of Epidemiology, Moscow, Russian Federation, 4 JSC Pharmstandard, Moscow, Russian Federation   Objectives: To evaluate pharmacoepidemiological picture of the course of influenza and other acute respiratory infections (ARI) in children during the first post-pandemic season. Methods: A retrospective analysis of medical histories of patients hospi- talized with influenza and other ARI was performed. Using the random sample method 2044 case histories of patients aged 1-18 years were se- lected. All the data were entered into a database for statistical analysis. Patients with laboratory confirmation of another inflectional disease were not included in the study. Continuously distributed variables were com- pared using criterion Mann-Whitney, categorical data or proportions were compared using χ2.

Pages Overview