Canine scent detection—Fact or fiction?

Abstract

Dogs have been used in a variety of scent detection tasks for hundreds of years. However, methodological differences in the design of studies concerning canine scent detection make it difficult to directly compare and to evaluate their results. We set out to (1) evaluate the quality and comparability of published literature concerning canine scent detection according to criteria of evidence-based medicine and (2) to determine the influence of the testing system on the outcome of a scent detection task considering two different testing systems. For the systematic literature evaluation we retrieved 31 studies. After applying specific exclusion criteria 14 studies were left for final evaluation. A check list detailing relevant information about the study design and the training and testing process was used. Our results demonstrate many differences in methodology and a high variability of the results of those studies leading to diversity in respect to relevant quality criteria. For the second part of our study seven dogs were trained by means of positive reinforcement to detect black tea (LIPTON Earl Gray, Unilever Deutschland GmbH, Hamburg, Germany) as target scent in two different testing systems, a testing platform and a scent detection board. Our data show that using an optimized training strategy high sensitivity (92.1%) and specificity (97.4%) can be achieved in a short time. Sensitivity and specificity for the detection of a target substance (i.e. black tea) was similar for the two testing systems.