Research Article: Rejections in an non-purpose bred assistance dog population: Reasons, consequences and methods for screening

Date Published: June 13, 2019

Publisher: Public Library of Science

Author(s): Evelien Bogaerts, Christel P. H. Moons, Filip Van Nieuwerburgh, Luc Peelman, Jimmy H. Saunders, Bart J. G. Broeckx, Simon Russell Clegg.


Assistance dogs aid people with various impairments on a daily basis. To become an assistance dog, a strict selection procedure and intensive training period must be successfully completed. Consequently, not every dog acquired for this purpose, becomes an assistance dog. The purpose of this study was to investigate reasons for failure and the financial consequences thereof for assistance dog associations that do not have a dedicated breeding program for their dogs. Data were collected for a total of 537 dogs enlisted between 2001 and 2015 and purchased out of the general dog population by five Belgian assistance dog associations. Only 60 percent of the dogs actually became an assistance dog and the main reasons for failure were related to undesirable behavioural characteristics and orthopaedic disorders. The estimated average financial loss per rejected dog was found to be 10524 euro. A detailed comparison of the two most popular breeds (Golden Retriever and Labrador Retriever) within the guide dogs and mobility assistance dogs revealed no significant difference in probability of successfully completing the training. However, a comparison of orthopaedic screening methods revealed a higher rejection with computed tomography for elbow dysplasia and laxity-based radiographical techniques for hip dysplasia compared to radiography and the standard ventrodorsal hip extend radiograph alone, respectively. Based on these results, we provide several suggestions to increase the probability of success.

Partial Text

Assistance dogs are defined as dogs trained to perform tasks for an individual with a disability.[1] All over the world these dogs aid people with various impairments. One example of assistance dogs are the guide dogs (GD) that aid blind or visually impaired handlers during daily activities (e.g. by avoiding obstacles). While GDs are generally well known, they represent only one of the five different subtypes of assistance dogs discussed here.[2] Mobility assistance dogs (MAD) provide practical support to people with mobility impairments by performing tasks like opening doors, retrieving objects that are out of their handlers reach and turning on lights. Hearing dogs aid their hearing impaired or deaf handlers by drawing their handlers’ attention with physical contact upon the occurrence of specific sounds and/or leading them towards the source of the sound. Seizure alerting–and response dogs are trained to alert the handler before the onset of, for example, epileptic seizures or hypoglycaemia or react by warning somebody if a seizure is going on. Autism assistance dogs support children with autism spectrum disorder. Given the wide spectrum of assistance these dogs offer, it is clear that the characteristics of these dogs and the training programs they follow, can differ substantially.[1–3]

A questionnaire was designed to obtain information on dogs that are purchased and trained to become assistance dogs in Belgium. A first part of the questionnaire focussed on general characteristics of the different ADA. General information included the type(s) of assistance dogs trained by the association, average purchase price of a puppy, and the duration and estimated cumulative cost after each of the three different life phases (host family, orthopaedic screening, training). ADA were asked whether they routinely performed screening for medical and/or behavioural traits and, if this was the case, which methodology was used, the age at which screening is performed and which criteria are used when deciding on the suitability of the dog.

This retrospective cross-sectional study contains two parts: a general overview of the characteristics of the participating ADA, obtained with a questionnaire, followed by a detailed evaluation of the dogs.

A total of five Belgian organisations participated in this study. These are relatively small ADA without a breeding program to provide puppies. As a consequence, puppies are purchased from individual dog breeders without a pre-existing purpose to become assistance dog. Two out of these five organisations mainly trained GDs, two others MADs and one trained small amounts of MADs, hearing dogs, seizure alert dogs and autism assistance dogs. Prior to acquisition, puppies were routinely subjected to a behaviour test. Three ADA provided details of their test: they were variations of the Campbell puppy selection test, but the execution and scoring criteria differed between ADA.[14] Irrespective of the purpose of the dog, all organisations followed the same sequence of phases: host family, orthopaedic screening and training with the duration of the different life phases being similar between the different associations. In more detail, the host family phase lasted from eight weeks until 18 months, with one exception where this phase ended at the age of ten months. Most of the dogs were screened for orthopaedic disorders at the age of 12 months. Behavioural screening was, in general, not performed at predetermined ages but rather continuously. The final phase, which is the training phase, took five to ten months and it was at the end of this period that a fully trained dog was paired with a disabled person.

Based on our results, it is clear that the rejection rate of assistance dogs originating from the general dog population is substantial and has a high financial impact. With a rejection rate of 40 percent, only three out of five dogs complete training successfully and become accepted. Whereas, to our knowledge, no literature is available on assistance dog organizations without a breeding program, rejection rates have been published for some organizations with a breeding program. For example, a study which included five guide and service dog organizations in the USA reported a success rate of 30 to 50 percent.[17] Bray et al (2019) reported a success rate of 43 percent over 13 years in an American organization with different types of assistance dogs, after exclusion for medical reasons and the breeders necessary to maintain a breeding program.[18] At first, this might seem a surprising result as more dogs are rejected than in our study. However, a possible explanation for this higher rejection rate could be a more strict screening protocol within organizations with a breeding program, in combination with higher standards for inclusion of a dog in training and earlier rejection due to the availability of larger numbers of puppies at the start. Where these organizations can actually be very stringent in which dog to select for training, obtaining puppies can be very difficult for organizations without a breeding program, especially if they are small. It has been reported to us that even releasing one or several puppies can have huge consequences for the functioning of these organizations. This might lead to maintaining less suitable dogs in training and even graduating them. Where this might result in a higher number of dogs rejected in their actual working phase, we cannot assess this as this data is not available to us. Overall, it is however clear that rejection rates are substantial.

Overall, we conclude that the rejection rate of purchased puppies before matching with a disabled person is high. Only 60 percent of the puppies selected by Belgian organizations actually become an assistance dog. With an average cost of 10524 euro per rejected dog, the subsequent financial loss is substantial. The main reasons for rejection are related to behaviour and orthopaedic disorders. Insecurity and/or anxiety were mainly reported as the behavioural reasons for rejection, while during musculoskeletal screening, HD was the most common reason, followed by ED. While the causes of these high rejection rates seem diverse, we suggest several approaches that can be used individually or combined to improve the success rates.




Leave a Reply

Your email address will not be published.