Evaluation of Zoonoses Trends by Confidence Calculation in GermanyGermany

allemand. Since its designation in 1996, the National Reference Laboratory for the Epidemiology of Zoonoses isresponsible to prepare a yearly German Report on Trends and Sources of Zoonotic Agents accordingto the Council Directive on Zoonoses (former 92/117/EEC, now 2003/99/EG). Germany has toprepare a yearly trend report as a contribution to be submitted to the EU Commission. To evaluatethe trends, questionnaires for the collection of data on zoonoses were distributed among the supremegovernment authorities of the Länder at the end of the year in co-operation with the Federal Ministryof Consumer Protection, Food and Agriculture. This contribution will demonstrate the evaluations ofthe reported investigations of zoonoses agents (e.g. Salmonella, Campylobacter, E.coli -VTEC/STECor Listeria monocytogenes) in foods sampled under a sampling plan with a new application using aformula to compare the results with the previous year by the confidence intervals.The standard calculations includes sums, percentages of positives of the investigated samples. As anew feature in the present report, the confidence intervals and accepted error have been included inthe tables of the zoonoses trend report. Calculations were performed by means of modification of thecalculations according to Spoorenberg et al (1996). Confidence intervals have been stated only if thenumber of samples was 385 or more. This corresponds to the minimal calculation for 5 % acceptederror and an unknown prevalence fixed at 50 %, according to Spoorenberg et al. (1996) for foodsamples collected under a sampling plan and a comparable standard investigation method (e.g. incase of Salmonella in Germany). This evaluation strategy will be demonstrated using some food categories reported from 2004 andcompared on a basis of significance to previous years (Hartung, 2004). Using data with comparablesampling schemes and comparable investigation methods a calculation of trends from the formeryears to the recent year is possible on a statistical basis. Y2 - 2005/07/19/21.07.2005