The Research of Standardized Protocols for Dog Involvement in Animal-Assisted Therapy: A Systematic Review

Simple Summary Animal-assisted therapies (AATs), as discussed in this review, are structured interventions, involving pets, for patients suffering from different diseases. Although many studies have highlighted the beneficial effects of these interventions on the well-being and health of the humans given the dogs, there are few studies that highlight the involved dogs themselves. Therefore, in this study, we carried out a systematic review to investigate the characteristics of the dogs involved in AATs. Based on the results, in most papers, there is a lack of general information on the dog(s) involved, including the methods used to choose and train the dog and the animal’s health status. These results highlight the need for standardized, specific methods to choose and train the dog and also suggest the need for univocal health protocols to ensure the animal’s welfare, as well as the final results of the therapeutic intervention. Abstract Dogs are considered the most important species involved in animal-assisted therapy (AAT), and the scientific literature focuses on the benefits linked to the involvement of dogs in various therapeutic areas. In this study, we carried out a systematic review according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, exploring the scientific literature from the last 5 years (2016–2021) on three databases (PubMed, Scopus, and Web of Science) to highlight the characteristics of the dogs involved in AATs. Based on the scientific literature relevant to such dogs, we considered different parameters (i.e., number, age, sex, breed, temperament, methods of choice and training, health status, research goals, and activities with dogs) to include studies in our paper. After screening 4331 papers identified on the searched databases, we selected 38 articles that met the inclusion criteria. Analysis of the included articles showed that the characteristics of the dogs were neglected. Our findings indicated a lack of information about the dogs, as well as the absence of standardized and univocal criteria for dog selection, training programs, and health protocols.


Introduction
According to some archaeological evidence, the domestication of dogs took place more than 30,000 years ago and was fundamental for the evolution of humans [1]. The dog was the first species to be domesticated. This domestication process was born from an extraordinary event in which suitable ecological conditions allowed the coexistence and, therefore, the formation of a relationship between human and dog. The coevolution between human and dog has created a unique interspecific relationship that has spread across cultures around the world and has survived by changing over the years [2]. Dogs were In our study, published scientific literature was selected by searching three electronic scientific databases: PubMed [35], Scopus [36], and Web of Science [37].
Other related papers, including references from the selected papers, were revised and used as sources of supplementary information.

Study Selection and Data Extraction
The relevant information was extracted from each paper included in this systematic review to achieve our stated goal. All data were entered into an Excel data set. We included data related to the characteristics of the dogs (i.e., number, age, sex, and breed), whether the dog was handled or owned, methods for choosing the dog, the temperament and training of the dog, and the health and behavioral status of the dog. Additional data were extracted to facilitate identification of each study (i.e., first name and year of publication).
The search query identified 114 articles (69 in PubMed,32 in Scopus, and 13 on the Web of Science). After evaluating the titles and abstracts of all articles, we selected the papers; after removing irrelevant and duplicate papers, only papers published in journals indexed on the Pubmed [35], Scopus [36], and Web of Science [37] were included. Ultimately, a total of 38 papers met the inclusion criteria. Figure 1 illustrates the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow-chart process [34] for study selection.
The inter-judge agreement was calculated (and independently identified by two judges) as a measure of reliability and assessed by Cohen's kappa. The minimum reliability value was ≥0.81, indicating strong agreement between the judges. Every disagreement was solved through intervention of the senior author (A.S.).

Results
The preliminary database search returned a total of 4331 documents. After removin the duplicates and irrelevant results, 114 articles were obtained for the complete revisio of the full text. Following the final evaluation, 38 articles were included in the systemat review. The results are shown in three tables. Table 1 includes the general characteristi of the dogs. Table 2 contains information regarding the research goals and activities wi the dogs. Table 3 provides the selection, training methods, and information about th health status of the dogs, including whether or not the dog belonged to the handler.
An overview of the results is presented in the following histogram chart ( Figure 2)

Results
The preliminary database search returned a total of 4331 documents. After removing the duplicates and irrelevant results, 114 articles were obtained for the complete revision of the full text. Following the final evaluation, 38 articles were included in the systematic review. The results are shown in three tables. Table 1 includes the general characteristics of the dogs. Table 2 contains information regarding the research goals and activities with the dogs. Table 3 provides the selection, training methods, and information about the health status of the dogs, including whether or not the dog belonged to the handler.    An overview of the results is presented in the following histogram chart (Figure 2). Table 3. Results of information on "Choice of dog", "Dog training", "Health Status", and "Dog ownership" in the included papers.    the duplicates and irrelevant results, 114 articles were obtained for the complete revision of the full text. Following the final evaluation, 38 articles were included in the systematic review. The results are shown in three tables. Table 1 includes the general characteristics of the dogs. Table 2 contains information regarding the research goals and activities with the dogs. Table 3 provides the selection, training methods, and information about the health status of the dogs, including whether or not the dog belonged to the handler. An overview of the results is presented in the following histogram chart (Figure 2).

Figure 2.
Overview of the results. Figure 2. Overview of the results.

Research Goals and Activities with Dog
In the AAT activity reports, almost all the works described the activities carried out with the dog; only one study did not release any information [40]. The sessions described had a duration between 10 min and 1 h. Activities included social [56], physical [39,55,59,60,62,66], and free [42,43,48,49,51,57,64,68,73] interactions; activities of rehabilitation [46,76]; psychomotor activities; and socialization [52]. Other activities included grooming, feeding, and stroking [54,58,61,67,70,75]; playing with the dog; and engaging in obedience exercises [45,50,65,72,74,75]. In one study, children read in the company of the dog [47]. Finally, in most studies, the role of the dog was focused on reducing anxiety and stress and improv-ing mood. In four studies, the dogs worked with autistic patients [41,48,60], and in one study, the dogs also worked with Down-syndrome patients [52]. Four papers focused on the elderly to improve the physiological parameters of heart disease [66], Alzheimer's disease [62], the perception of pain in geriatrics [59], and cognitive functions among patients with dementia [60]. In two papers, the dogs supported patients with oncological diseases [67,70]. Three studies carried out psychological rehabilitation [46] with inmates [72] and schizophrenia [74]. Other work focused on improving communication and social skills [44,45,47,49,51,63,76]. Finally, one study explored the use of support dogs during dental sessions [61]. As a method for choosing the dog, five papers used test administration [41,44,51,65,72], while three papers evaluated AAT certifications [50,53,56]. Seven studies evaluated accredited programs as a method of inclusion [39,40,43,61,69,73,76]. All other studies did not include this information. To validate the dog's temperament, most studies tested docility, socialization, or type of training. Thirty-two papers did not report any information on the temperament of the dogs. For training, almost all papers reported membership in accredited associations or programs as the source of dog training, but the type of training was rarely mentioned. Nine papers did not offer any information on the type of dog training [45,47,52,57,58,61,69,70,74]. Furthermore, dogs were examined by a behavioral veterinarian in only seven studies [41,44,63,67,72,74,76]. In terms of health protocols, almost all the reviewed works applied vaccinations and parasitic treatments. In 19 studies, no information on health protocols was provided.
Finally, in only three studies [55,70,76] was the dog handler also the owner. In most of the studies, whoever worked with the dog was only a handler, or it was not specified whether the handler was also the owner of the dog. The remaining 12 studies did not include this information.

Discussion
Dogs are some of the most frequently involved animals in AATs due to their long history of coevolution with humans and their numerous beneficial effects, which have been widely described in the literature [14][15][16].
However, although studies on dogs are numerous, many important aspects remain overlooked [20]. In most of the selected articles, there was often a lack of information regarding the general characteristics of the dog, such as the number of dogs involved in the AAT, as well as the age, sex, and breed of each dog. This is probably because other aspects of research are given importance, while the dogs have been overlooked while representing an integral and fundamental part of the AAT's team [14][15][16]. In particular, data on the dog's temperament are presented in very few studies [45,49,50,52,59,63,65,68,73], thereby neglecting the importance of who the dog is; in this way, the dog's vision as the "other", a unique subject with its own personality, is lost. The methods used to select dogs are also not exhaustively presented. Accredited methods of selection are rarely mentioned. Some works cited various tests carried out [44,51,65,76], others cited certifications [41,43,50,53,56,61], and others cited self-contained training programs [39,69]. This vast heterogeneity in choice protocols is linked to the lack of univocal tests to select a dog most suitable for a certain type of AAT and the overall lack of standardized selection criteria for the dogs involved. Furthermore, most of the works limited themselves to only mentioning the names of accredited associations or programs without describing the training methods of the dogs in detail. This result indicates a lack of unique educational training programs among the dogs involved.
In terms of the relationship between dog and handler [11,16,77], based on our results, the handler was reported to be the owner of the dog in very few studies [55,70,76], while most of the remaining studies did not specify whether the handler was also the owner of the dog. This is a very important aspect, as the relational dimension is essential for an effective intervention [16,[77][78][79]. Furthermore, it was previously shown that the relational factors between dogs and owners influence the performance of the dog [78]. As other studies already reported, good attachment and a harmonious relationship between dog and owner can increase safety, reduce stress factors in the dog, and improve the dog's performance [79]. Therefore, an interspecific relationship in an AAT can positively affect the final results of the intervention [80]. An equally important aspect that was neglected in the analyzed works concerns the applied health protocols. Few studies preventively examined dogs using a behavioral veterinarian, the only professional figure able to certify the suitability of a dog with respect to the absence of behavioral diseases [41,44,49,63,67,72,74,76]. This aspect is of fundamental importance for the safety of users involved in AAT to prevent zoonotic risks [81][82][83]. In addition, in some studies, the health protocols used to ensure the safety of the setting and the users/patients involved in the interventions were not mentioned or specified [40,42,43,45,[47][48][49][50]52,[55][56][57]59,60,[64][65][66]71,74,75]. This result indicates the absence of standardized reference health protocols, which could yield dangerous consequences linked to zoonotic risks, especially since AATs are often given to immunosuppressed patients [84][85][86][87].
In terms of the types of activities carried out by the dogs during AATs, our results are in accordance with previous research findings [88], which highlighted the limitations of current studies due to a lack of information on the welfare indicators of dogs-particularly how different interventions (in terms of duration and frequency), the environmental factors of the setting, human manipulation, animal characteristics, etc. can influence these indicators. We also noted the physiological and behavioral variations that can occur in response to the actions of the handler or the patient, suggesting that we should measure the personality and stress levels not only of the dogs but also of the people involved, which represents a potential area of future research on interspecific relationships. Therefore, we strongly suggest a greater awareness of scholars on the need to investigate on welfare, considering not only stress parameters, but also the pleasure indicators of dog during an AAT intervention.
In today's society, dogs are recognized as sentient beings, as they are endowed with their own personalities [89,90] and emotions [91]. The data reported by Horowitz [89]similar to the data reported in the present review-underline that in human-animal interactions, dogs are strongly neglected; however, since dogs are now considered to be sentient beings, their roles must be more seriously considered.

Limitations
This study has several limitations. The heterogeneity of the data collected in this systematic review did not allow for adequate statistical evaluation of the different characteristics analyzed in each paper, as the examined data did not refer to universally shared protocols. Moreover, certain characteristics were not always presented or described. However, these data can be considered an important index of the heterogeneity that characterizes the scientific literature related to the characteristics and roles of dogs in AATs. Furthermore, these characteristics reinforce the need to develop and disseminate general protocols to facilitate the use of globally recognized tools and allow, at the same time, a greater focus on the dog.

Conclusions
Based on our results, the characteristics of the dogs involved in therapy are rarely considered, as such factors were neither described nor investigated in detail in most of the reviewed studies. With a contemporary and scientifically validated understanding of dogs as sentient beings, species characteristics and dog individuality must be seriously considered. Our results show that the parameters most requiring in-depth analyses are related to the choice, temperament, and training of the dog and health protocols. Furthermore, greater attention should be given to interspecific relationships and the factors that influence them, as well as the interspecific relational skills of each individually chosen animal and its relationship with the owner, such as attachment styles and relational reciprocity between the animals, patients, and members of the working group. All these factors will affect the animal's welfare, the safety of the therapeutic setting, and achievement of the study's goals. The strength of these therapies lies in the involvement of dogs, which offers its own languages and abilities that determine activation of the emotional, cognitive, and behavioral dynamics responsible for the various positive effects observed in previous studies. Therefore, it is necessary to implement standardized and univocal protocols that emphasize the role of the dog and facilitate a paradigm change capable of underscoring the importance of and respect for the individuality of each dog, not only the benefits the dog can bring.