Animal olfactory detection of human diseases: Guidelines and systematic review

Abstract

Animal olfactory detection of human diseases has attracted an increasing amount of interest from researchers in recent years. Because of the inconsistent findings reported in this body of research and the complexity of scent detection research, it is difficult to ascertain the potential value of animal detectors in operational diagnostic algorithms. We have outlined key factors associated with successful training and evaluation of animals for operational disease detection and, using these key factors as points for comparison, conducted a systematic review of the research in this area. Studies that were published in peer-reviewed outlets and that described original research evaluating animals for detection of human diseases were included in the review. The majority of relevant studies have assessed dogs as detectors of various forms of cancer. Other researchers have targeted bacteriuria, Clostridium difficile, hypoglycemia, and tuberculosis. Nematodes and pouched rats were the only exceptions to canine detectors. Of the 28 studies meeting inclusion criteria, only 9 employed operationally feasible procedures. The most common threat to operational viability was the use of a fixed number of positive samples in each sample run. Most reports included insufficient information for replication or adequate evaluation of the validity of the findings. Therefore, we have made recommendations regarding the type of information that should be included when describing research in this area. The results of this systematic review suggest that animal detectors hold promise for certain diagnostic applications but that additional research evaluating operationally viable systems for olfactory detection of human diseases is necessary.

Canines (Canis lupus familiaris) as biodetectors for conservation work: Can they discriminate the rock ptarmigan (Lagopus muta) from the willow grouse (L. lagopus) in a yes/no task?

“…Several factors can influence a dog-handler team´s accuracy rate [32,49,53]. Working dogs have been stated to have increased success when it comes to careful selection in relation to breed and personality traits as e.g.…”

“…Faecal pellets from the ptarmigans were collected in the field for the genetic study and directly placed in plastic jars (Nalgene, 15 x 38 mm or 30 x 43 mm, Thermo Scientific™, Norway). Later, they were taken out of the jars for seven days to dry and cross-contamination of the pellets was avoided by separating them by species and only interacting with them using sterilised equipment [32]. Time from sampling in the field, to when the drying process was initiated, could range from five hours to two weeks.…”

“…To emphasis that the dogs learnt to both recognise and discriminate the rock ptarmigan from other odours, a collection of control samples were used in the platform [29,32,38]. Faecal pellets from birds in the same family (Tetraonidae) and the same genus (Lagopus) were used, and other laboratory environmental control samples used during the preparation of samples (e.g.…”

“…From here on, the dog-handlers were blind to the presence and position of the target scent as an experimenter placed the samples. The experimenter was situated in the room next door watching the dog-handler team through a video monitor connected to cameras in the laboratory, hence no potential cuing to the dog-handler team and double-blind training [32,54]. The experimenter evaluated the team´s responses, and if a correct response was made, a click from a clicker held by the experimenter that could be heard across the rooms was provided, and the dog rewarded.…”

Automated Canine Scent-Detection Apparatus: Technical Description and Training Outcomes

“…However, they are also associated with some disadvantages, particularly those related to potential cuing (see Lit et al, 2011); subjectivity involved in the identification of the indication response and data collection in general; and the immediacy, accuracy, and reliability of reinforcer delivery. For these reasons, Edwards et al (2017) suggested that an optimal laboratory-based approach to scent-detection research and operations is fully automated. However, few, if any, automated solutions for applied scent-detection research with dogs have been described in the literature and no such apparatus is available commercially.…”