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

ISCB2014_abstract_book

ISCB 2014 Vienna, Austria • Abstracts - Poster Presentations 107Tuesday, 26th August 2014 • 10:30-11:00 Monday25thAugustTuesday26thAugustThursday28thAugustAuthorIndexPostersWednesday27thAugustSunday24thAugust P2.2.115 Timing for definitive cure in clinical trials for visceral leishmaniasis R Omollo1,2 , N Alexander3 , N Omolo2,4 , P Oleche2 , M Wasunna1 , T Edwards3 1 DNDi, Kenya Medical Research Institute, Nairobi, Kenya, 2 Maseno University, Private Bag, Maseno, Kenya, 3 London School of Hygiene & Tropical Medicine, London, United Kingdom, 4 Jaramogi Oginga Odinga University of Science & Technology, Bondo, Kenya   Background: Visceral Leishmaniasis (VL) is a parasitic neglected tropi- cal disease, endemic in developing countries. Treatment outcomes in VL clinical trials are measured at two time points: a) initial cure at end of treat- ment (EOT) and b) definitive cure (DC) measured commonly at six months post EOT. Additional assessment is commonly conducted at either one or three months post EOT for patient monitoring. This paper investigates justification for shortening of the six month follow-up time point in the assessment of DC. Methods: A three state Markov model has been applied to estimate the transition probabilities for DC using data from a phase III clinical trial con- ducted in East Africa to compare the safety and efficacy of three different treatment regimens. At baseline all patients were untreated but their sta- tus changed to either treatment failure (TF) or success (TS) at EOT, month 3 or 6 follow-up with TF considered as an absorbing state. Results: Conditional on a patient ever having a TS, 98.5% of observations have a TS outcome at the next visit while that of ever having a TF is 61.4%. From the transition probabilities, approximately 1% of patients change cure status between month 3 and 6. Overall, there was high stability for the treatment outcomes (>97%). Conclusion: Advantages of shortening follow-up time include a possible reduction in loss to follow-up and time to availability of effective new treatments. Limitations to be addressed in simulation studies include the possibility of false negative results and differences in relapse rates be- tween treatment regimens.   P2.2.127 Etiological age-components of cervical cancer in Finland in 1953-2011 K Seppä1 , J Pitkäniemi1 , N Malila1 , M Hakama1 1 Finnish Cancer Registry, Helsinki, Finland   The objective of the study was to assess whether age incidence of cervical cancer in Finland is consistent with two etiological components and how the changes in the incidence is concordant with the organized screening programme. Thestudyutilizedapopulation-basedregisterofcervicalcancercasesfrom the Finnish Cancer Registry diagnosed at 20-84 years of age in 1953-2011 and female population at risk. Age-specific incidence of cervical cancer was estimated by using a Poisson regression model with the assumption of two normally distributed latent components. A hierarchical Bayesian model was applied and life time net risks and crude numbers of cancer cases of the two components were estimated from the joint posterior. Before the screening started the life time net risks were 0.5% for younger and 1.3% for older (RR=0.43, 95% CI 0.29-0.61) with actual cases of 154 and 206 per year, respectively. The component of incidence occurring in younger women disappeared in 1970s after the organized mass screening program was initiated (in 1973-1977, RR=0.02, 95% CI 0.00-0.08). Since that, the risk for younger increased to 0.2% whereas the risk for older decreased to 0.5% in 2008-2011 (RR=0.58, 95% CI 0.25-0.94) with actual cases of 80 and 61 per year, respectively. Existence of the two components is likely to be due to different etiological exposures. The trend in risk of the both components followed closely both the effects of organized screening and changes in sexual mores. The increase in the first is consistent with increase in HPV exposure. P2.2.132 Location-scale tests for non-negative data with skewed distribution, with focus on parasitology research J Reiczigel1 , S Yehia2 1 Szent Istvan University, Faculty of Veterinary Science, Budapest, Hungary, 2 Tanta University, Faculty of Science, Tanta, Egypt   Two-sample comparison of parasite infection data is usually made by loca- tion tests. As more infected samples have both higher mean and higher SD, location-scale tests might be more powerful. We investigated this by simulation. Our results may also be relevant in other fields where skewed distributions are common, e.g. in the analysis of health care cost data. Methods: We compared Cucconi’s location-scale test (CU) with 3 com- monly used location tests: Welch-t-test (WT), Mann-Whitney-test (MW), and bootstrap-t-test (BT), for 5 right-skewed theoretical distributions and 5 empirical parasite distributions. Sample sizes varied from 10 to 100. Alpha was compared assuming that the two samples came from identi- cal distributions. Power was compared assuming that differences would appear at higher infection levels, reflecting to the observation that even in heavily infected populations many hosts (those with good defence) re- main free or almost free. Results: Alpha error rate of MW, CU, and BT was acceptable (<6% at nomi- nal 5%) if ratio of sample sizes (SSR) was below 2, and slightly elevated (<7.3%) for SSR=3. WT showed alpha=8.9% for SSR=3, and 7.2% for SSR=2, therefore we excluded WT from the power comparison. Power depended strongly on the distribution. For theoretical distributions CU performed best. For empirical parasite distributions there was no clear “winner” but both CU and BT had considerably higher power than MW. Conclusion: Location-scale tests may be useful in parasitology. CU was more powerful than the location tests, among which BT was best. All tests had elevated alpha levels for unbalanced designs.   P2.2.155 The challenges of conducting a multidisciplinary trilogy of studies: diagnostic accuracy & use of the SeHCAT test in England JA Summers1,2 , A Pascoal2 , S Keevil2 , C Lewis2 , G Vivian3 , R Logan3 , V Cornelius1 , J Peacock1,2 1 King’s College London, London, United Kingdom, 2 King’s Health Partners AHSC, London, United Kingdom, 3 King’s College London Hospital NHS Foundation Trust, London, United Kingdom   Bile Acid Malabsorption (BAM) is common in many diarrhoea conditions, however, robust data on the prevalence of BAM and care pathways for di- agnosis and treatment of the condition do not exist. Confirmation of BAM can be made using SeHCAT (tauroselcholic [75 sele- nium] acid), a radiolabelled synthetic bile acid. The SeHCAT test measures the retention of radioactivity in the patient following administration of a capsule containing SeHCAT. Bile acid sequestrants (BAS) are a treatment option for BAM. The aim of this research is to gather and assess evidence regarding the use of SeHCAT in the diagnosis of BAM in NHS centres. Several challenges have been identified in conducting this ongoing re- search: • Patchy adoption of SeHCAT testing across NHS centres; • Limited information on outcome of patients diagnosed as BAM negative; • Poor patient adherence to BAS treatment; • Current lack of evidence on clinical value of the SeHCAT test; • Expertise is required in various areas: nuclear medicine, gastroenterol- ogy, study design, statistics and database development and manage- ment; • Statisticians working and coordinating successfully with clinicians in an area with little background knowledge. Based on the identified challenges, a trilogy of sequential studies is being

Pages Overview