In our study, we attempted to review the available methods for quantification of the effectiveness of AAI, based on published literature. Since AAI is a diverse field, both qualitative and quantitative research methodologies were used in the reviewed publications. Therefore, we considered it necessary to review both evidence-based studies and published methods aiming to measure the effectiveness of AAT.

To summarize the original findings of our review, although there are multiple research about AAT and many best practices were published, the quantitative comparison of best practices is not possible because of the lack of standardized methodology and lack of standardized research protocols. As the quality assessment scores of the reviewed studies show there are only a few appropriate research methods in this field. Our original findings suggest that AAT is an emerging adjunctive treatment in the rehabilitation of various disorders, which has gained significant popularity. Future studies with homogenized patients’ groups and standardized methodology are needed to increase the comparability of the studies and therefore provide more robust evidence of the efficacy of AAT. Thus, based on the results, we may conclude that AAI significantly improved both the physical and mental well-being of participants, but the mental improvement was more outstanding. According to the results, the physical rehabilitation achieved by AAI is slower, while psychological coping can be facilitated in the short term by the therapy animal.

Of all the research studies we reviewed, three studies examined the effects of animal-assisted intervention on participants based on how the intervention complementing the usual treatment impacted them (Beinotti 2010, 2013, Machová). In addition, two articles examined the effect of animal-assisted interventions carried out in addition to standard therapy, and the effects of other additional interventions besides standard therapy (Bukentrop 2017, 2019).

This is supported by Palsdottir’s research. In their study, participants spent 30 min with effective equine-assisted therapy once a week, complementing this by patients spending 60 min connecting with each other, attuning to the therapy animal, and preparing for the joint activity (social connection, building bridges). This is an important intervention from both the musculoskeletal and the mental perspectives. From Palsdottir’s research, it is clear that AAI induced significant positive changes in the patients’ mental state. Machová et al. indicated that those patients who received AAT felt better despite showing no significant changes in heart rate or blood pressure. The beneficial effect of AAT on patients' well-being is also demonstrated by the subjective mood ratings on a Likert scale that patients completed before and after the session. By improving the mood of the patients, AAT has an indirect impact on the successful rehabilitation of patients. It encourages proactivity in the treatments they receive, improves their interaction with other therapists, and thus improves their relationship with the whole team providing the treatment (Machová, 2019).

In the studies we reviewed, the use of horse-assisted interventions was over-represented compared to dog-assisted ones; however, the results of the research studies suggest that horses are more likely to assist with physical development, while dogs are more likely to assist with mental and social development. Beinotti, Bunketorp, Lee, and Sunwoo et al. supported the beneficial effects of equine-assisted therapy on balance, gait function, and hand and arm use. However, Peppe’s small sample size study found that a 5-week course of equine-assisted therapy in Parkinson’s patients improved motor skills and also reduced anxiety and apathy among participants. The developmental impact of a dog-assisted intervention is supported by the results of Machova and Macaulay. They found that dog therapy increased motivation, mood, and self-reliance in stroke survivors.

Clarity is made difficult by the fact that concepts of AAT are still not applied uniformly, as experts do not refer to exactly the same things by them. Despite this incongruity, we tried to use the terminology related to animal-assisted intervention consistently. Thus, we used the term “dog-assisted therapy” for interventions with dogs and “equine-assisted therapy” for interventions with horses, but the activities carried out in the studies with dogs and horses differ from one study to another.

Another important limitation is the diversity of measurement tools. Physical (gross motor function, fine motor function, and balance) and mental changes (mood, motivation, and well-being) are measured very differently, making it difficult to compare results. From the publications we analyzed, we have identified approximately thirty different scales, such as FAC, Fugl-Meyer Scale, BBS, NIHSS SIS, TUG, Timed 10 m-WT, M-MAS (UAS) for the measurement of physical changes, and STAI, OVal-pd, EQ-VAS/EuroQol-VAS SMBQ for the measurement of mental changes. There are also tools that measure both physical and mental states, such as the MOSF36 SF. There are also measurement tools that we identified in several publications, e.g., TUG, Barthel index, and Timed 10 m-WT.

Also, an important limitation is the disproportionality of studies with different species. In the literature reviewed, equine-assisted therapy is over-represented (3 therapies with dogs, 10 therapies with horses), but we have no information on the number of AAI in the world and the proportion of dog-assisted therapies and equine-assisted therapies, and on how many of these attempts to measure the impact of animal therapy. Our research team could analyze only published papers. It is a well-known fact that in past centuries, in the case of dogs and horses (as opposed to other animals), cooperation with humans was a main selection criterion during the course of breeding, and both species have a long tradition of being involved in tasks and being specifically trained.

The quality of the published research varies widely. In the present review article, we examined thirteen publications with the following results: 9 publications were rated good, and 4 publications were rated poor. One possible reason for this may be the scarcity of available literature on the topic that meets the criteria and the novelty of the research field. Due to the small number of papers published on this topic and meeting our quality criteria, we did not limit the scope of the present review to one group of animals, one disease, or one type of animal-assisted intervention. The low number of quality articles published posed a challenge to our research. On one hand, this paints an incomplete/distorted picture of this form of complementary rehabilitation; on the other hand, it also draws attention to the fact that there have been very few studies of appropriate excellence on this topic. Since this field has been undergoing explosive development, our study calls attention to the fact that many more qualitative research studies are necessary to provide an overview of the subject. Besides the low number of quality publications, the other major limitation was methodological heterogeneity. Although the definitions are provided and there is literature that specifies criteria, e.g., for animal-assisted therapy, on reviewing the research reports we found that these definitions are not used consistently in practice. Furthermore, we could find no information on whether there are comparable forms of quality control measurement in place in different countries, and due to their absence, the use of measurement tools and interventions may differ from one country to another, or even from one research study to another. The length of musculoskeletal and social interventions varies. The interventions range from a few weeks or months to complex programs lasting a year. This makes it difficult to compare the effectiveness of AAIs of varying lengths.

In most studies, AAT was used as an adjuvant to conventional medical therapy, but the studies do not provide clear-cut information on whether patients with similar illness severity were selected in the experimental and control groups. Therefore, as mentioned earlier, quantitative comparison of efficacy has major limitations. Although Parkison’s disease seems to be associated with a higher efficacy of AAT than MS, we cannot draw definite conclusions due to significant confounders such as the great variability of the severity, symptoms, and course of the conditions. Affected brain areas can be very different even in the case of one particular disease, which may have direct effects on psychomotor functions. There are many psychological confounding factors (mood, apathy, level of motivation, locus of contol, etc.) which also affect the response to any intervention. Another important factor is the great variability of the studies and the lack of standardized methodology for measuring efficacy. Future studies with standardized assessment and methodology and more homogenous participants groups considering psychomotor functions, disease severity, and mental health state can be eligible for quantifying efficacy differences.

As interest in AAT has grown rapidly and the number of scientific publications on the subject has increased significantly over the last 15 years, there has been an increasing need for both providers and users to review the state of the art in AAT. An important strength of the paper is that it attempts to investigate the effectiveness of AAT by summarizing highly heterogeneous studies that differ in methodology, quality, and target group. The publication following the 2006 peer-reviewed paper, which met the selection criteria in terms of quality, is from 2010, followed by papers published in 2013 and 2014. The fact that 7 of the publications we analyzed were written in the past 7 years suggests that interest in the topic has increased in recent years, with more and more people using AAI. The number of research studies on AAI is growing, and its literature is expanding at a faster pace. A further strength of our paper is that the approach presented in the review provides a quality framework for the development of methodological features for future research. If the presented framework is followed in future studies, comparability, and evidence-gathering would be much easier.

There are few evidence-based studies available, and their summary is insufficient to draw any far-reaching conclusions; however, we may venture to conclude that AAT is an effective adjuvant treatment for the studied diseases, adding significant value to conventional therapies, particularly for psychiatric symptoms and quality of life, and, in the case of equine-assisted therapy, for movement rehabilitation. Therefore, we need to emphasize the importance of developing a uniform measuring methodology, a larger sample size, and detecting significant changes. In the case of control group studies, it would be important to use simultaneous complementary therapies, such as dog-assisted therapy and art therapy in the two analyzed groups. By developing evidence-based research methodology and standardizing the research setting, the effects of animal-assisted therapy could be effectively measured as part of healthcare practice, which would bring about significant benefits in the rehabilitation of patients in need. Interventions of different lengths could be subject to further research in order to identify the length of intervention that is the most effective, and the systemic change a 12-month intervention can trigger in patients’ lives.

It would be a step forward in quantifiable interventions with measurable impact if professionals used the definitions uniformly. It would also be important for research and interventions to have a regulated professional framework for animal-assisted therapy, within which the professionals have the flexibility to tailor the therapy to the clients’ needs.

In connection with the measurements, the idea may arise that control group studies conducted in health care institutions would be more fit for the purpose. However, given the heterogeneity of the patient group and the pharmacotherapy used, the implementation of this cannot be standardized as of yet. In contrast, biometric measurements could be helpful in measuring the impact of AAI.

We would like to confirm that although this is a difficult area to assess, at the same time, an evidence-based approach is indispensable. Psychometric tools (depression, anxiety) are key in measuring changes in mental health.