Table 1 Study Characteristics

Qualitative

 Casey (2018), Canada [34]

To explore the effects of an AAI program for residents with dementia on the work environment and perceived mental health and wellness of staff working with them

n = 20 staff

(purposive) in a long-term care facility

AAI program comprised three 1 h sessions per week for 8 weeks. Residents engaged with a range of farm animals including a sheep, rabbits, chickens and a goat

Y

In-depth interviews with eight open ended questions. Recorded and transcribed verbatim

Thematic analysis

 Cook (2013),

UK [35]

To evaluate the impact of the HENPOWER project on resident self-report of mental wellbeing, loneliness and depression

n = 21 care home residents, 1 relative, 22 staff, including managers

(purposive)

HENPOWER programme aimed to establish henkeeping in six care homes

N

Focus group interviews and individual interviews. Recorded and transcribed verbatim

Thematic analysis

Realist Evaluation

 Dookie (2013), Canada [36]

To explore the potential benefits of AAA on elders’ empowerment, self-esteem and quality of communication with caregivers in retirement homes

n = 10 residents who had a pet

(purposive)

Residents with a personal pet living in five semiautonomous retirement homes

N

Semi-structured interviews. Interpretative Content Analysis

 Fossey & Lawrence (2013), [37]

UK

To identify the perceived advantages of involving animals in the life of the care home; staff concerns regarding the inclusion of animals; factors that facilitate their presence; and barriers to their inclusion

n = 108 staff from 15 care homes

(purposive)

Staff with past and present experiences of a wide range of animals discussed the perceived benefits and difficulties associated with animals living at or visiting the care home

N

Fifteen focus groups. Data collection became progressively more focussed, and emerging themes were tested out in subsequent focus groups

Thematic analysis

 Freedman (2021),

UK [38]

To explore what the experience of keeping a personal pet in a care home means for residents’ sense of wellbeing

n = 7 residents from four different care homes, three had dogs and four had cats (purposive)

Residents with a personal pet living in care homes

N

Semi-structured interviews with residents

Interpretative Phenomenological Analysis

Phenomenology

 Gundersen & Johannessen (2018)

,

Norway [39]

To gain insight into the experiences and motivations of volunteer dog handlers and nurses who are involved in dog visits to nursing homes

n = 8 dog handlers, 10 nurses who had attended at least two dog visits

(purposive)

AAA comprised dog visits at four nursing homes with groups of three to five residents in a suitable room and at least one employee (nurse or nursing assistant). The handlers planned and led the visits

Y

Individual in-depth interviews with the dog handlers and group interviews with the nurses

Inductive approach to data analysis. Salutogenic theory

 Kawamura (2009),

Japan [40]

To examine how institutionalised older adults who have been participating in AAA for two years perceive the activity, and what relevance these perceptions have on clinical nursing practices in the AAA context

n = 8 residents (women) in a private nursing home living with dementia

(no details given)

AAA sessions twice per month for two years. At each session three or four dogs were taken to the nursing home and the residents were able to freely feed, hold and play with the dogs for approximately 30 min

Y

Semi-structured interviews (approximately 30 to 50 min). Interviews were recorded and transcribed. Colaizzi’s phenomenological methodology for data analysis

Phenomenology

 Kendziorski (1999),

USA [41]

To determine the effects of resident pets on residents living at a local nursing home, including perceived quality of life, and examine the relationship between the residents and the resident animals

n = 10 residents from one nursing home, all previous pet owners)

(purposive)

Informal interaction with the resident animals. Primary resident pet – Golden Retriever, also resident birds and fish

N

Semi-structured interviews following the hermeneutic process

Hermeneutic analysis for data analysis

 Kongable (1990),

USA [42]

To evaluate nursing staff feelings and attitudes about the use of a pet dog as a therapeutic agent for residents residing in a special care Alzheimer’s unit

n = 6 nursing staff associated with the pet therapy programme

Programme involved the use of pet therapy. The dog was brought to the unit for three hours, one day per week

Y

Interview with five open-ended questions (approximately 15 min)

Content analysis

 Pitheckoff (2018),

USA [43]

To obtain insider knowledge of the experience of participating in an existing animal-assisted activity program that provides older adults with an opportunity to interact with rabbits

n =  ~ 30 residents (observed) and 8 residents (interviewed)

AAA (Rabbit assisted activities). Activity typically involved three rabbits that were informally shared with residents in their rooms and in common areas. Activity session lasted for one hour

N

Direct observation of two activity sessions conducted several weeks apart

Eight face-to-face, in-depth, semi-structured interviews (generally lasted 45 min)

Thematic Analysis (Triangulation of observation data and interview data)

 Pooley (2007), UK [44]

To produce an explanatory model to improve understanding of the psychological benefits of pet ownership by investigating the significance of companion animals for older people who move with them into homes for older people

n = 9 residents (living in nursing, residential and sheltered housing, 7 managers

(purposive)

Residents with a personal pet living in care homes

N

Semi-structured interviews with residents; managers of the homes; observations in the homes were conducted in the communal areas (one hour)

Grounded theory methodology

Grounded theory

 Roenke & Mulligan (1998), USA [45]

Explore the qualitative aspects of pet therapy in order to better understand what it is about such experiences that is perceived as beneficial from the participant’s perspective

n = 4 residents, 1 pet therapy volunteer

(purposive)

Two sessions of pet therapy (one with a five month old kitten, and one with two eight week old kittens)

N

Participant observation of two sessions of pet therapy; group interviews of three residents, individual interview with one resident, individual interview with pet therapy volunteer. Interviews consisted mainly of open-ended questions

Thematic analysis

 Savishinsky (1985),

USA [46]

Describe the results of a long-term, intensive research project conducted on the Cornell Companion Animals Program (CCAP), a university-based, volunteer organisation

Residents in three nursing homes (n = 72, 75 and 250). Approx 10% -30% resident involvement

Volunteers conducted weekly visits using either their own pets or borrowed animals offered to CCAP. In the smaller homes, pet visitation was in a group format, in the larger home, individual residents met with pets in their own rooms or lounge areas

N

Direct participant-observation, unstructured interviewing and life history work. Researcher also completed participant observation on non-pet related recreational activities at the three homes

No details on data analysis

 Swall et al. (2015),

Sweden [47]

To illuminate the meaning of the lived experience of encounters with a therapy dog for persons with Alzheimer’s disease

n = 5 residents living in a nursing home with medium to severe Alzheimer’s Disease

AAT – Ten visits per person with therapy dog. Prescription is individualised with a specific justification for each planned visit, for example, to increase alertness or to decrease anxiety

Y

Video observations

The first step of the analysis was a naïve reading, followed by structural analysis and a comprehensive understanding

Phenomenological hermeneutics

Phenomenology

 Swall et al. (2016),

Sweden [48]

To illuminate meanings of the lived experiences of dog handlers’ when visiting older persons with dementia with their therapy dog

n = 9 dog handlers

As above

Y

Interviews with open-ended questions (as above)

Phenomenological hermeneutics

Phenomenology

 Swall et al. (2017),

Sweden [49]

To illuminate meanings of care for people with Alzheimer’s Disease in their encounters with a therapy dog

n = 5 residents with medium to severe dementia

As above

Y

Video observations focusing on the interaction between the person with AD and the therapy dog (as above)

Phenomenological hermeneutics

Phenomenology

 Swall et al. (2019),

Sweden [50]

To describe the impact of therapy dogs on people with dementia in the final stages of life from the perspective of the dog handler

n = 7 dog handlers

(convenience)

As above

Y

Semi-structured interviews

Participants were asked to talk about situations when they had visited people with dementia with their therapy dog

Qualitative content analysis

 Wong (2015)

New Zealand [51]

To examine the experience of animal therapy in two residential aged care homes that receive animal visits from an animal welfare organisation

n = 7 residential aged care home residents

(convenience)

AAT visiting programme involving puppies, kittens, adult dogs, rabbits and guinea pigs, rats and adult cats. Visits were either once a week or once a fortnight and each visit lasted approximately an hour

N

In-depth, semi-structured interviews

Narrative analysis

Narrative gerontology

 Wong & Breheny (2021), New Zealand [52]

To examine the experience of animal therapy in two residential aged care homes that receive animal visits from an animal welfare organisation

n = 7 residential aged care home residents

(convenience)

AAT visiting programme involving puppies, kittens, adult dogs, rabbits and guinea pigs, rats and adult cats. Visits were either once a week or once a fortnight and each visit lasted approximately an hour

N

In-depth, semi-structured interviews

Narrative analysis

Narrative gerontology

 Zando (2017), USA [53]

To explore how residents diagnosed with Alzheimer’s disease respond to contact with a certified therapy dog

n = 5 nursing home residents

(purposive)

Researcher met with each resident in privacy of own room, introduced the therapy dog and engaged the resident in a conversation about the dog. There were three visits and each visit lasted for no more than 15 min

Y

Recorded observations of physical and/or verbal responses the resident had to the dog. Observations were coded to identify significant themes, categories and patterns. A consensus process with peer colleagues was used to support the identification of themes

Quantitative

 Andrysco (1982),

USA [54]

To determine the effectiveness of animal assisted therapy in a nursing home and explore appropriate methods for assessing resident-animal interaction

n = 46 residents in 1 long-term care home. Age range 70–99 yrs

Individualised AAT. Residents received 1 × 20 min session with a dog over 10 weeks. Control group had a visit without a dog

N

RCT. Primary outcome was interaction and behaviour (assessed by questionnaire). Secondary outcome social interaction (assessed by video). Analysis by (3-way ANOVA)

 Banks et al. (2008),

USA [55]

To compare the ability of a living dog and robotic dog (AIBO) to treat loneliness in resident living in long-term care

n = 42 residents in 1 care home. No age metrics – described as elderly

Individualised AAT. AAT groups received weekly visits lasting 30 min from either a dog robot (AIBO) or a living dog for 8 weeks

N

RCT. Primary outcomes of loneliness (UCLA questionnaire) and attachment (Lexington attachment to Pets Scale). Means were compared by analysis of variance

 Banks (1998; 2002),

USA [56]

To determine whether AAT is effective in combating loneliness among elderly adults

N = 15 residents in 3 long-term care facilities (70% > 75 yrs)

Individualised AAT. AAT groups received either weekly or thrice weekly visits lasting 30 min with a dog. The control group received no visits

N

RCT. Primary outcomes of loneliness (UCLA-LS questionnaire) and pet history questionnaire also completed. Analysis using ANCOVA

 Bumsted (1988),

USA [57]

To investigate whether pet therapy has a favourable effect on resident self care

n = 20 from 1 large nursing home. Mean age 83yrs (I), 86 yrs I

Individualised AAT. Residents received 6 × 25 min sessions over 3 weeks with a dog. Control group had no visits

N

RCT. Primary outcomes were self-care and physical self-maintenance (Bespoke questionnaires). Analysis by MANOVA

 Briones (2021), Spain [58]

To assess the effectiveness of AAT on quality of life in residents with dementia in a care home

n = 39 residents with dementia in 1 care home. Mean age 88yrs(I) 89yrsI

Group AAT. Residents had 1 × 50 min session per week for 9 months. Control had usual care, no visits

Y

RCT. Primary outcomes was quality of life (QOL-AD). Functional status also assessed (Barthel Index). Analysis using ANCOVA

 Colombo (2006),

Italy [59]

To determine whether pet therapy had a favourable effect on psychopathological status and on perceived quality of life

n = 144 from 7 nursing homes. Mean age 78 yrs (I and C)

Companion pet. Residents either had a canary or plant to look after in their room, or usual care for 3 months

N

RCT. Primary outcome was quality of life (LEIPAD-short version). Brief symptom inventory (BSI) also completed. Analysis by variance analysis

 Friedman (2015),

USA [60]

To evaluate the use of structured activities with a dog (PAL intervention), to prevent deterioration of physical function, emotional and behavioural function

n = 40 residents with dementia from 7 long-term assisted living facilities. Mean age 80 yrs (I) 82yrs I

Group AAT. Residents had 2 × 60–90 min session per week with a dog over 12 weeks. Control group had reminiscence sessions

Y

CLUSTER RCT. Primary outcomes of activities of daily living (Barthel Index) and depression (CSDD) and agitation (CMAI). Analysis using linear mixed models

 Greer (2002), USA [61]

To examine the effect of live

cats compared to toy cats on the verbal communication of individuals with dementia

n = 6 residents in 1 nursing home with Alzheimer’s disease. Age range 84 -90yrs

Group AAA. Residents (groups of 3) had 3 × 10 min session with 2 live cats. Control phase involved 2 toy cats

Y

CROSSOVER RCT. Primary outcome was the number of words spoked (observation). Analysis approach not specified

 Johnson (1997), USA [62]

To determine whether pet encounter therapy (PET) has a favourable effect on mood states and social facilitation

n = 100 residents in 6 skilled nursing facilities. Mean age 85yrs (I) and 78yrs I

Individual AAT. Residents received either an animal (rabbit/kitten), stuffed toy, or person only visit of 10 min. Control had no visit

N

RCT. Primary outcome was affective mood (MAACL). Analysis using ANOVA

 Le Roux (2009), South Africa [63]

To explore the effect of AAA on the depression and anxiety levels of residents in a long-term care facility

n = 16 residents from 1 home. All > 65 yrs

Group AAA. Residents spent time with a dog for 1 × 30 min, over 6 weeks. Control group had no visits

N

RCT. Primary outcomes were depression (BDI) and anxiety (BAI). Analysis using Wilcoxon signed rank test

 Olsen (2016), Norway [64]

To examine the effects of AAA on depression, agitation and QoL in nursing home residents with dementia or cognitive impairment

n = 58 residents from 10 nursing homes. Mean age 83yrs (I) and 84yrs (C)

Group AAA. Residents received 2 × 30 min dog visits per week, for 12 weeks. Control group had no visits

Y

CLUSTER RCT. Primary outcomes were depression (CSDD), agitation (BARS) and quality of life (QUALID). Analysis using ANOVA

 Panzer-Kaplow (2000),

USA [65]

To investigate the effects of AAT on levels of depression and morale among residents of a geriatric nursing facility

n = 35 residents from 1 nursing home. Mean age 72yrs (I) and 73yrs (C)

Individual AAT. Residents received 1 × 15 min dog visit per week, for 10 weeks. Control group received no visits

N

RCT (Matched pairs). Primary outcomes were depression (BDI &GDS) and morale (PGCM). Analysis using t-tests

 Pope (2016), USA [66]

To compare the effectiveness of AAT versus human interaction only on social behaviours and engagement among elderly patients with dementia in a long-term care facility

n = 44 residents from 1 skilled nursing facility. Mean age 80yrs

Individual AAT. Residents had 2 × 10 min/week sessions for 2 weeks with a dog, Control phase involved visits with human only

Y

CROSSOVER RCT. Primary outcome was social behaviour (observation) and engagement (MPES). Analysis using ANOVA

 Thodberg (2016a & b), Denmark [67, 68]

To explore the effects of a real animal (dog) compared to robot/ toy animals on the psychiatric wellbeing and sleep of residents with dementia

n = 124 residents from 4 nursing homes. Overall mean age 85yrs

Individual AAT. Residents received 2 × 10 min visits with a dog, robotic seal or stuffed toy cat,

Y

RCT. Primary outcomes were behaviour (GBS), sleep quality, depression (GDS) and cognitive functioning level(MMSE). Analysis by linear regression

 Travers (2013), Australia [69]

To assess the effect of dog-assisted therapy for people with dementia living in aged care facilities on mood and quality of life

N = 67 residents from 3 aged care facilities. Mean age 85yrs (I and C)

Group AAT. Residents received 2 × 40/50 min sessions a week, for 11 weeks with a dog. Control group had therapist visit only

Y

RCT. Primary outcomes were depression (GDS), quality of life (QOL-AD) and psychosocial functioning (MOSES). Analysis by linear regression

 Valenti Soler (2015),

Spain [70]

To compare the effectiveness of AAT (dog) versus therapeutic sessions with a robotic animal (PARO) or usual care for residents with moderate to severe dementia

n = 110 residents from 1 nursing home. Mean age for all 85yrs

Individual AAT. Residents received 2 × 30-40 min sessions per week for 12 weeks with a dog or robotic seal. Control group had usual care

Y

RCT. Primary outcomes were behaviour changes (NPI), apathy (APADEM-NH) and quality of life (QUALID). Analysis with a mix of non-parametric tests

 Wall (1994), USA [71]

To explore whether a purposeful goal orientated interaction with a companion animal facilitates positive changes in residents wellbeing

n = 80 residents from 11 skilled nursing facilities. Mean age 88yrs(I) and 83yrs(C)

Individual AAT. Residents received 3 × 6-10 min sessions with a dog over 3 weeks. Comparator groups (stuffed toy or human)

N

RCT. Primary outcomes were speech activity (audio) and mood (MSE). Analysis by ANOVA

 Zulauf (1987), USA [72]

To explore the effects of a dog-facilitated therapy visitation program on psychosocial wellbeing

N = 40 residents from 1 nursing home. Age range all: 60-92yrs

Individual AAT. Residents received 1 × 30 min visit/ week with a dog for 6 weeks. Control group human visit

N

RCT. Primary outcomes were depression (GDS), Morale (PGCM) and self esteem (SES). Analysis by ANoVA