Animal Assisted Interventions in the Green Care Framework: A Literature Review

Green Care (GC) and Animal Assisted Interventions (AAI) are recognised practices useful to enhance the wellbeing of people through interaction with nature and animals. This study aims at understanding the interconnections between GC and AAI by analysing deeply which interaction with animals is conducted. Therefore, we carried out a literature search through Web of Science and Google Scholar that allowed retrieval of 993 records; after the PRISMA selection process, 42 were included. Relevant information was extracted: year of publication, geographical location, objectives, settings in agricultural environment, animal species, characteristics of users involved, type of human–animal interaction, coexistence of other activities without animals, animal health and welfare issues. From the review emerged that research on GC with animals is common in high-income countries and that the line between AAI and occupational therapy is often vague. Moreover, the most common setting for these interventions appears to be the farm, and frequently animals involved are not selected according to their ethological characteristics. Users in this context are extremely various and not only involved in activities with animals. Within the included studies, we noted a lack in the consideration of animal welfare that indicates the need for increased awareness among practitioners and a more ethical approach when animals are involved.


Introduction
Green Care is a complex concept identified by a simple phrase. Since the early 2000s, it has started to appear in the scientific literature and became subject of research conducted by the European Cooperation in Science and Technology network that, by means of expert consultation, defined the Green Care conceptual framework [1]. Nowadays, we use Green Care as an umbrella term to include all the approaches in which the resources of nature provide health and wellbeing to people [2].
In this paper, our main interest is to investigate a specific branch of the Green Care framework, i.e., Animal Assisted Interventions (AAI). The field of AAI is usually considered independent because its development began in the 1960s with Boris Levinson experiences [3], and the scientific literature on AAI is mostly focused on the benefits of human-animal interaction [4]. An AAI can be defined as "a goal oriented and structured intervention that intentionally includes or incorporates animals in health, education and human services for the purpose of therapeutic gains in humans" [5]. This definition was drawn up by the International Association of Human-Animal Interaction Organizations (IAHAIO), an entity that has been engaged in research and the practice of human-animal bonds and AAI from the early 1990s, including members from all over the world. Nevertheless, when the Green Care concept started to be defined, it embedded AAI within the range of approaches in which natural resources (animals, in this case) are beneficial to human health and wellbeing [6]. From that moment, scientific literature concerning Green Care has been extremely various in the terminology spectrum used and frequently intertwined with AAI [7]. Obviously, this correspondence is not one-toone: not all Green Care activities involve animals and their relationship with humans (e.g., horticulture), and at the same time a large extent of scientific literature on AAI does not imply a natural setting, or it does not include the effects of the natural setting within the purposes of the study.
Therefore, we conducted a literature review in order to investigate the role of the interaction with animals in the Green Care context. Our aim is to clarify the interconnection between Green Care and AAI and to understand deeply what kind of interaction with animals is conducted in the Green Care context, focusing on the users, the settings and the species involved.

Eligibility Criteria
We considered only records describing activities or interventions that involved both the agricultural environment and interaction with animals (to which from now on, we will refer with the term "Green Care with animals"). Given the relatively recent conceptualization of Green Care, we considered only the timespan 2000-2020. Only literature published in English was included. Review articles were excluded to ensure to rely only on primary data.

Databases and Search Strategy
The literature search was carried out in April 2020 through Web of Science and Google Scholar. Search strings are reported in Table 1. The total number of records resulting from the Web of Science platform (613) and the first 400 results of Google Scholar (sorted by relevance) were retrieved for the screening phase. Web of Science 613 TI = ((care farm* OR social farm* OR social agriculture OR farm assisted OR green care OR nature based rehabilitation OR nature assisted therap* OR working in nature OR wilderness therap* OR farm based OR farm therap* OR rural welfare) AND (farm animal* OR animal* OR assisted intervention* OR animal assisted OR horse* OR donkey* OR cow* OR goat* OR sheep OR rabbit* OR chicken* OR pig* OR camelid* OR human-animal)) OR TS = (("care farm*" OR "social farm*" OR "social agriculture" OR "farm assisted" OR "green care" OR "nature based rehabilitation" OR "nature assisted therap*" OR "working in nature" OR "wilderness therap*" OR "farm based" OR "farm therap*" OR "rural welfare") AND ("farm animal*" OR animal* OR "assisted intervention*" OR "animal assisted" OR horse* OR donkey* OR cow* OR goat* OR sheep OR rabbit* OR chicken* OR pig* OR camelid* OR "human-animal")) Databases = WOS, KJD, MEDLINE, RSCI, SCIELO Timespan = 2000-2020; Search language = Auto TI = title TS = topic Google Scholar 400 care farm OR social farm OR social agriculture OR farm assisted OR green care OR nature based rehabilitation OR nature assisted therapy OR working in nature OR wilderness therapy OR farm based OR farm therapy AND farm animal OR animal OR assisted intervention OR animal assisted OR human-animal Timespan = 2000-2020; Search language = Auto

Data Collection Process
Two reviewers conducted the screening phases. The records (n = 993) were exported into EPPI-Reviewer Web software in order to remove duplicates and then titles and abstracts were screened to remove non-relevant ones. Subsequently, the remaining 67 records underwent full text screening. Finally, from the included records (n = 42) we collected in a Microsoft Excel Table the following information: year of publication and geographical location of the study, objectives of the study, settings in agricultural environment, animal location of the study, objectives of the study, settings in agricultural environmen species and characteristics of the users involved, kind of human-animal interactio activities conducted without animals, animal welfare issues and health perspecti

Overview
The selection process is summarized in Figure 1 with a PRISMA inspired flow di [8]. A total of 42 records were included in this review: 33 peer reviewed journal a six book chapters published by Springer, one peer reviewed conference proceedi book and one report. All publications are listed in the Appendix A. A synthesis o information extracted is reported below. Most records included in this review were published in 2006 and in 2014 (F Of the six records considered for 2006, five were collected in a book titled "Farm Health" [9]. The year with the maximum number of papers about human-anima tion in the Green Care framework published by different research groups was 20 As for geographical location, most studies were conducted in Europe, especially in Norway (30.9%), the United Kingdom (19%) and the Netherlands (14.2%). Outside of European countries, the studies fulfilling the predefined eligibility criteria were conducted in the United States of America (9.5%) and South Korea (4.8%) ( Figure 3). Comparing years of publication and countries, we can see that some nations published on this theme only in one year (Austria, Germany and Sweden in 2006 and Italy in 2019), while others, such as Norway, the Netherlands and the United Kingdom, published more or less every year. We did not find a trend of publication related to the time in any country.

Objectives of the Studies
Within the records analysed, around 36% (n = 15) aimed at mapping Green Care providers and/or at describing Green Care practices in a geographical area (e.g., a country or a region) [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24]. The remaining approximately 64% (n = 27) described the perceived benefits of Green Care with animals programs by both users and providers. In detail, in six of these studies, Green Care with animals programs were compared to other approaches (with the same users' category) to highlight their benefits [25][26][27][28][29][30]. In three records, the main focus was the evaluation of the efficacy of the Green Care with animals programs, but experimental designs lack of control groups [31][32][33]. Other 13 records analysed the benefits of Green Care with animals for users by means of qualitative methods [34][35][36][37][38][39][40][41][42][43][44][45][46].  As for geographical location, most studies were conducted in Europe, especially in Norway (30.9%), the United Kingdom (19%) and the Netherlands (14.2%). Outside of European countries, the studies fulfilling the predefined eligibility criteria were conducted in the United States of America (9.5%) and South Korea (4.8%) ( Figure 3). As for geographical location, most studies were conducted in Europe, especially in Norway (30.9%), the United Kingdom (19%) and the Netherlands (14.2%). Outside of European countries, the studies fulfilling the predefined eligibility criteria were conducted in the United States of America (9.5%) and South Korea (4.8%) ( Figure 3). Comparing years of publication and countries, we can see that some nations published on this theme only in one year (Austria, Germany and Sweden in 2006 and Italy in 2019), while others, such as Norway, the Netherlands and the United Kingdom, published more or less every year. We did not find a trend of publication related to the time in any country.

Objectives of the Studies
Within the records analysed, around 36% (n = 15) aimed at mapping Green Care providers and/or at describing Green Care practices in a geographical area (e.g., a country or a region) [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24]. The remaining approximately 64% (n = 27) described the perceived benefits of Green Care with animals programs by both users and providers. In detail, in six of these studies, Green Care with animals programs were compared to other approaches (with the same users' category) to highlight their benefits [25][26][27][28][29][30]. In three records, the main focus was the evaluation of the efficacy of the Green Care with animals programs, but experimental designs lack of control groups [31][32][33]. Other 13 records analysed the benefits of Green Care with animals for users by means of qualitative methods [34][35][36][37][38][39][40][41][42][43][44][45][46].  Comparing years of publication and countries, we can see that some nations published on this theme only in one year (Austria, Germany and Sweden in 2006 and Italy in 2019), while others, such as Norway, the Netherlands and the United Kingdom, published more or less every year. We did not find a trend of publication related to the time in any country.

Settings in Agricultural Environment
Studies included in our review were mainly carried out in farms (Figure 4). Just a few records mentioned a "natural environment" in general [11,12,15,34,39] and only one was conducted in the prison context [19]. Finally, five of these records asked for the opinion of providers on the efficacy and benefits of Green Care with animals programs [47][48][49][50][51].

Settings in Agricultural Environment
Studies included in our review were mainly carried out in farms (Figure 4). Just a few records mentioned a "natural environment" in general [11,12,15,34,39] and only one was conducted in the prison context [19].

Animal Species Involved
In 25 records, the animal species involved were more than four or generally identified as "farm animals". Other eight records did not even specify the animal species involved, even if the users interacted directly with them [17,18,20,22,24,26,31,42]. Finally, nine records reported specifically the animal species involved in their activities: in four cases two or three species were mentioned: goats, cattle and pigs [37], chickens and pigs [47], chickens and rabbits [21] or cattle, hens and alpacas [33]. In five records, authors referred only to one species: horses [27,35,39] or dairy cows [28,40].

Kind of Human-Animal Interaction
Seventy-six percent of records classified human-animal interactions as animal taking care, animal feeding, animal husbandry or a mix of these activities. In 12%, authors declared that users were involved in Animal Assisted Therapy [11,17,39,48,49]. Only one record referred to "riding horses" [27] and one to "training animals" [19] whereas in 7% human-animal interactions were not described in detail [13,22,50]. These results are summarized in Figure 5.

Animal Species Involved
In 25 records, the animal species involved were more than four or generally identified as "farm animals". Other eight records did not even specify the animal species involved, even if the users interacted directly with them [17,18,20,22,24,26,31,42]. Finally, nine records reported specifically the animal species involved in their activities: in four cases two or three species were mentioned: goats, cattle and pigs [37], chickens and pigs [47], chickens and rabbits [21] or cattle, hens and alpacas [33]. In five records, authors referred only to one species: horses [27,35,39] or dairy cows [28,40].

Kind of Human-Animal Interaction
Seventy-six percent of records classified human-animal interactions as animal taking care, animal feeding, animal husbandry or a mix of these activities. In 12%, authors declared that users were involved in Animal Assisted Therapy [11,17,39,48,49]. Only one record referred to "riding horses" [27] and one to "training animals" [19] whereas in 7% humananimal interactions were not described in detail [13,22,50]. These results are summarized in Figure 5

Animal Welfare and Health Perspectives
In 36 out of 42 records we analyzed, there is no reference to the health and welfare of the animals involved and the main focus was the benefit of the Green Care program for the human side. Only few authors described animal housing and husbandry conditions: herd and free access to pasture for horses [27,35,39], free-range stall for dairy cows [28]. Finally, only two records highlighted the need of a change of perspective, of improving the study of animal welfare to understand the impact of this kind of programs on animals [10,44]. Moreover, we noticed that no papers referred to the health risks for patients who interact with animals in a farm environment and there were no references to hygienic procedures and zoonosis control programs.

Discussion
Data analyzed in this review highlighted that the line between AAI and occupational therapy is often vague. The importance of the human-animal interaction is stressed in all the papers we included, but only 22 of them examined in detail how animals are beneficial in the Green Care context . Six out of eight records focused only on the humananimal interaction in the Green Care context were published in Norway, a country where the research topic of Green Care with animals was developed between 2007 and 2014 providing social services to the community [25,27,28,30,40,51]. Certainly, the co-presence of the natural environment and interaction with animals is the main source of benefits for the patients/users, but none of the studies investigated the possible inner mechanisms involved. The selected timespan for the literature search results to be correct, as we did not find scientific literature dated before 2006. A book was published in that year whose five chapters are included in this review [9], highlighting that human-animal interactions in farm context were already carried out in practice, but the scientific interest for the topic developed later and only in some western countries with well-developed agro-economic systems.
Taking into account that papers were selected for reporting programs of Green Care with animals, as a matter of fact not every record discussed the benefits of these practices.

Animal Welfare and Health Perspectives
In 36 out of 42 records we analyzed, there is no reference to the health and welfare of the animals involved and the main focus was the benefit of the Green Care program for the human side. Only few authors described animal housing and husbandry conditions: herd and free access to pasture for horses [27,35,39], free-range stall for dairy cows [28]. Finally, only two records highlighted the need of a change of perspective, of improving the study of animal welfare to understand the impact of this kind of programs on animals [10,44]. Moreover, we noticed that no papers referred to the health risks for patients who interact with animals in a farm environment and there were no references to hygienic procedures and zoonosis control programs.

Discussion
Data analyzed in this review highlighted that the line between AAI and occupational therapy is often vague. The importance of the human-animal interaction is stressed in all the papers we included, but only 22 of them examined in detail how animals are beneficial in the Green Care context . Six out of eight records focused only on the human-animal interaction in the Green Care context were published in Norway, a country where the research topic of Green Care with animals was developed between 2007 and 2014 providing social services to the community [25,27,28,30,40,51]. Certainly, the co-presence of the natural environment and interaction with animals is the main source of benefits for the patients/users, but none of the studies investigated the possible inner mechanisms involved. The selected timespan for the literature search results to be correct, as we did not find scientific literature dated before 2006. A book was published in that year whose five chapters are included in this review [9], highlighting that human-animal interactions in farm context were already carried out in practice, but the scientific interest for the topic developed later and only in some western countries with well-developed agro-economic systems.
Taking into account that papers were selected for reporting programs of Green Care with animals, as a matter of fact not every record discussed the benefits of these practices. In the majority of them, benefits were presented more as an opinion of the authors or the providers than as a scientific evidence [34][35][36][37][38][39][40][41][42][43][44][45][46][47][48][49][50][51]. Only six studies had an experimental design with a control group and a well-described methodology, which confirmed the efficacy of the Green Care approach with animals [25][26][27][28][29][30]. Generally, the main benefit described was an improvement of users' well-being due to the non-institutional context. Anyway, the lack of internal and external validity of all papers analyzed addressed to the need of a more structured and well-designed research to investigate the real positive effects of Green Care.
We noticed that the descriptions of Green care programs' settings were very generic. The word "farm" was used referring to a rural and productive context where many activities were provided to the beneficiaries. The only exception are "prison farms" which are well described by Furst [19]. The author investigated programs involving animals offered in the U.S. prisons: one of them was provided in a farm environment where inmates worked with farm animals. In some other records the word "farm" was replaced by "natural environment" without any additional description of the context. We can suppose that such kind of settings are not related to agriculture or livestock production and at the same time far away from the traditional idea of a "healthcare setting".
Additionally, the animal species involved was not specified in the majority of records we analyzed. It seems that animal features are not so relevant in the Green Care context. Except for few studies focused on horses [27,35,39] and dairy cows [28,40], other animal species were not involved for their ethological characteristics but only for the rural context in which they were housed and their "non-human" nature. Without a priori choice of animal species, projects can be easily adapted to the users' needs and will maintain flexibility for many beneficiaries' categories.
Actually, users involved belonged to a wide range of categories including specific physical and mental disabilities [12,15,16,25,26,28,30,33,40,49,51], people with poor social inclusion [19,46], abused or affected by addiction [23,35]. More than a half of the records analyzed described human-animal interaction programs without specifying category of users, highlighting their great flexibility. Others, involving horses [27,35] and dairy cows [28,40], were focused on patients/users with specific features, fitting the animal species and the nature of the interaction with the standard therapeutic program.
In many cases, the kind of services provided with animals were not described in detail but generally referred to "animal taking care activities" (e.g., brushing, cleaning stables, feeding animals). These activities give new occupational skills to the users and increase their responsibility toward another living being, but they are far away from AAI, which are based on the development of a human-animal relationship [5], that is used as a tool to achieve therapeutic, educational or recreational goals, specifically established for each AAI project. In Green Care context, the non-judging interaction with animals is mostly limited to the possibility of physical contact, with an underspending of the AAI potentiality.
We highlighted that in 13 records human-animal interaction was not the only activity provided, but in the same context users could benefit from horticulture and gardening, which are relevant in Green Care [10,13,17,18,[21][22][23][24]33,36,38,42,44]. This framework created an interconnection between "taking care of plants" and "taking care of animals" offering a holistic approach to the beneficiary who is completely surrounded by the rural environment.
Anyway, in the records we analyzed, we noticed lack of attention towards key issues such as animal welfare and health risks related to human-animal interactions in rural context. Interactions with beneficiaries with disabilities and without specific knowledge and skills on animal husbandry could be a source of stress for animals, especially when not well managed and trained by animal handlers [52]. If precautions are not taken, the interaction with humans can be physically and mentally demanding for animals [53]. Poor description of animal housing and husbandry, as well as the absence of clear reference to animal welfare status and monitoring, suggest the need to raise awareness among practitioners towards a more ethical approach in the involvement of animals in such kind of activities. Moreover, health issues related to zoonosis or pathogens with antimicrobial resistance transmissible to humans in rural contexts is another crucial topic due to the possible vulnerability of users' health. Hygienic procedures and health monitoring of farm animals were not mentioned at all in the records included in this review, highlighting an underestimation of the problem.
Finally, given the heterogeneity of information sources and the type of studies included, the methodology and level of detail provided by each record are various; this factor can represent a limitation of this review. However, the question underlying this review was not focused on the effectiveness of GC with animals, but more on the description of this emerging phenomenon. Therefore, it was considered to include also different types of information sources: as previously reported, of the 42 included voices, 33 are peer reviewed journal articles, six are book chapters published by Springer, one is a peer reviewed conference proceeding, one is a book, and one is a report. For all these records, it was possible to identify the study purposes (as described in 3.2), which in some cases have been achieved through qualitative methods, in other cases through quantitative ones. In particular, as given ahead, only six of them had an experimental design with a control group and a well-described methodology [25][26][27][28][29][30]. Nevertheless, by adopting a critical perspective on the selected literature, some considerations may arise. Analyzing the characteristics of the records included, the studies often lack relevant information: on one hand, the detailed description of patients (type of disease, age, etc.), operators and animals involved (species, type of training and management, etc.); on the other hand, the specification of the activities that are carried out during the interaction with animals. In the interests of increased consistency of the scientific literature which investigates the phenomenon, in particular as regards efficacy studies, the purpose of standardizing the design and reporting of the studies conducted can undoubtedly contribute to shed light on the complex interaction between humans, animals and the environment in a one health/one welfare perspective.

Conclusions
From our literature review we can conclude that interaction with animals in the Green Care Context is present and extremely various. The countries of publication of the records included demonstrate that these practices are typical of developed areas. In some studies, the human-animal interaction is minimal, but in others is connected properly to a therapeutic goal and could be considered as an AAI (even if we do not know if in these cases there was an adequate support of professionals). Moreover, we have to say that when studies describe the benefits deriving from Green Care with animals it is difficult to distinguish which of these come from the human-animal interaction or from the natural environment. The incorporation of AAI into Green Care will benefit form a greater awareness of the animal species involved and its ethology; for future research, it is desirable that more attention will be paid to the planning of interventions, to animal welfare and the health aspects of the interaction, in order to guarantee safety for both users and animals.