Canine co‐therapy: The potential of dogs to improve the acceptability of trauma‐focused therapies for children

Abstract

Objective

Post‐traumatic stress disorder is a debilitating syndrome that effects approximately 30% of those who experience childhood trauma. Whilst effective treatments exist, they have high rates of attrition and non‐engagement. Augmenting traditional interventions such as trauma‐focused cognitive behaviour therapy is an approach often used to address this problem. The current study aimed to assess the acceptability of a novel adjunct, canine‐assisted therapy, which may be particularly useful with younger children.

Method

A community sample of Australian parents and caregivers (N = 267) read one of two vignettes describing problematic behaviour following sexual trauma in either a 6‐ or 13‐year‐old girl. Participants then rated their acceptability of three described treatment alternatives; traditional therapy alone, with medication and with canine‐assisted therapy, using the Treatment Evaluation Inventory–Short Form.

Results

Regardless of the age of the child traditional therapy with canine‐assisted adjunct had the highest acceptability. Adjunct medication was the least acceptable approach, particularly in those with no prior psychoactive medication experience.

Conclusions

Canine‐assisted therapy as an adjunct to traditional cognitive‐behavioural approaches holds promise as a way to improve acceptability. Implications and future research directions are discussed.

Key words:

• adjunct therapy
• animal‐assisted therapy
• canine‐assisted therapy
• child
• trauma
• treatment acceptability
• Treatment Evaluation Inventory

Please note Janene would like her married name to appear on this article (previously Arnold‐Hancock).

What is already known about this topic:

• Trauma‐focused cognitive behaviour therapy (TF‐CBT) is well evidenced and is currently the most common treatment for post‐traumatic psychological symptoms. However, TF‐CBT has also been reported to have high attrition rates and low acceptability especially for children following trauma.

• Adjunct therapies which can improve acceptability, promote treatment adherence, and reduce attrition rates in this age group would be beneficial.

• While inclusion of a trained animal, often a dog or horse, within a therapeutic intervention is increasingly being viewed as a useful adjunct with a relatively high degree of acceptability by professionals, little is known about consumer/lay acceptability of including an animal within treatment of childhood trauma.

What this topic adds:

• When asked to read a vignette which presented an overview of a child who had been sexually abused and, rate the acceptability of three possible therapeutic approaches, participants overwhelmingly indicated TF‐CBT with adjunct canine‐assisted therapy as more acceptable than either TF‐CBT alone or with adjunct medication.

• Neither age of the child in the vignette nor previous experience with dogs influenced acceptability ratings, while prior experience with psychoactive medication did.

• With the implementation of National Disability Insurance Scheme (NDIS) in Australia, parents/consumers will have greater control over treatment options. Improving the acceptability of TF‐CBT by including adjunct therapies such as canine‐assisted therapy (CAT) may mean that more seek, and adhere to, efficacious treatments such as TF‐CBT.

Experiencing trauma at any age can be significantly detrimental; however, it is particularly problematic during the formative years of childhood (4–18-years, De Bellis, Hooper, Woolley, & Shenk, 2010). Whilst most children are remarkably resilient in response to adverse life events, approximately 30% will develop persistent stress, fear, and avoidance, impeding their ability to function, and leading to clinically significant post‐traumatic stress disorder (PTSD) (Cohen, 2010).

Trauma‐focused cognitive behaviour therapy (TF‐CBT) is the most evidence‐supported treatment for childhood PTSD, with a large number of randomised control trials supporting its efficacy compared to other trauma‐focused therapy protocols (Scheeringa, Weems, Cohen, Amaya‐Jackson, & Guthrie, 2011). However, the efficacy of TF‐CBT is overshadowed by significant rates of attrition, as high as 70%, in both clinical care and outcome trials (Wamser‐Nanney & Steinzor, 2017). This issue is thought to stem from the difficulty clinicians face forming an effective therapeutic alliance with trauma‐exposed children, which is shown to impact the child's motivation and commitment to therapy (Miller‐Graff & Campion, 2016). The quality of the therapeutic alliance is particularly vital in TF‐CBT where it has been shown to mediate efficacy (Ormhaug, Jensen, Wentzel‐Larsen, & Shirk, 2014). Trauma‐exposed children tend to experience great difficulty forming relationships with strangers and trusting adults, which can be further impaired by certain types of trauma, particularly childhood sexual abuse (CSA) (Green, 2006). For CSA victims, social and relational symptomatology can be significantly magnified by the abuse context, almost half have a trusting and affectionate relationship with the perpetrator prior to the onset of abuse (Kemp, Signal, Botros, Taylor, & Prentice, 2014). As a result, these children tend to develop rigid safety behaviours and avoid unfamiliarity as a means of coping with beliefs that the world is unsafe and unpredictable (Campbell & Simmonds, 2011).

Augmenting traditional prolonged exposure (PE) therapies is an approach often used to inhibit the avoidance behaviour that prevents victims from engaging with treatment (Lefkowitz, Paharia, Prout, Debiak, & Bleiberg, 2005). Currently, the most commonly used adjunct therapy is medication, in particular selective serotonin reuptake inhibitors (SSRI) (e.g., Hendriks, De Kleine, & Van Minnen, 2015). However, not only is there a lack of evidence supporting the efficacy of SSRIs as an intervention for trauma (Hendriks et al., 2015), but official practice guidelines for treating childhood trauma do not recommend the prescription of medication (Cohen, 2010). In addition, several studies have found that lay acceptability of medication as a psychological intervention is extremely low due to the prevalence of negative side effects and stigma surrounding their use (Farach et al., 2012). Acceptability is even lower for those with no prior psychoactive medication experience (Kehle‐Forbes, Polusny, Erbes, & Gerould, 2014). These issues highlight an important consideration in psychotherapy, that of treatment acceptability.

Treatment acceptability encompasses judgements regarding the fairness, intrusiveness, and appropriateness of interventions by treatment consumers (Kazdin, 1980). This judgement affects both sides of the therapeutic relationship with clinicians reluctant to recommend interventions with low acceptability to clients (Meyer, Farrell, Kemp, Blakey, & Deacon, 2014; Swift, Callahan, Ivanovic, & Kominiak, 2013), and for consumers, low acceptability reduces motivation to participate, non‐compliance with the treatment protocol, and compromised overall efficacy (Swift et al., 2013). For traumatised children, treatment acceptability also encompasses their parent's perceptions of that treatment, with researchers highlighting the role that parents play in initially accessing treatment and in adhering to treatment protocols (Wamser‐Nanney & Steinzor, 2017). Lay person/parental perceptions of treatment acceptability is particularly topical in Australia, where current changes in federal funding arrangements have given consumers an unprecedented amount of influence over treatment planning (Carey & Mathews, 2015). The introduction of the National Disability Insurance Scheme (NDIS, http://www.ndis.gov.au) represents a fundamental shift towards a consumer driven model of funding, where clients rather than clinicians will decide on the direction of government intervention spending (Thill, 2014). Despite the emphasis on choice, empirical evidence is still important due to the ethical imperative for professionals to recommend treatments with a solid evidence base (McIlwraith, 2014). As a result, TF‐CBT is unlikely to be superseded as a recommended treatment for childhood trauma, however, augmenting TF‐CBT with a complementary approach that improves acceptability may enhance the utility of this therapy.

It is suggested that animal‐assisted therapy (AAT) may represent an effective and acceptable complementary adjunct for TF‐CBT (O'Haire, Guérin, & Kirkham, 2015). AAT involves the goal‐directed use of a trained therapy animal in specific clinical interventions (Nimer & Lundahl, 2007). Whilst robust research on AAT is still limited (O'Haire et al., 2015), a recent meta‐analytic review by Hoagwood, Acri, Morrissey, and Peth‐Pierce (2017) concluded that canine‐assisted therapy (CAT) was an effective intervention for childhood PTSD.

Extant research on the positive effects of canine‐human interactions (including the production of oxytocin, often referred to as the relational hormone, Beetz, Uvnas‐Moberg, Julius, & Kotrschal, 2012) strongly support the potential for CAT to facilitate the goals of TF‐CBT. Younger children in particular seem to have an innate attraction to canines, which may result from the predictability and simplicity that is characteristic of the behaviour of both (Zilcha‐Mano, Mikulincer, & Shaver, 2011). Just like dogs that derive affection from touch through petting, very young children experience similar affection from their caregivers by way of gentle touch and physical comfort (Levinson, 1984). Therefore, it is suggested that the inherent value of CAT is its potential to help younger children form a commitment to therapy through forming a comforting attachment to the therapy animal (Levinson, 1984).

Literature on the general acceptability of AAT is, however, scarce. The majority of acceptability studies have sampled professionals, including hospital staff, physicians (Moody, King, & O'Rourke, 2002; Yap, Scheinberg, & Williams, 2016), and psychologists (e.g., Black, Chur‐Hansen, & Winefield, 2011). Some studies have revealed factors that might impact the acceptability of the practice, including concerns about the physical and psychological wellbeing of the therapy animal (Black et al., 2011; Lefkowitz et al., 2005), the client's orientation towards and prior experience with dogs (Yap et al., 2016), and the potential impact of animal misbehaviour (Moody et al., 2002). Despite these concerns, professionals appear to be generally positive about AAT as a psychotherapy approach.

Studies of the acceptability of AAT amongst lay or consumer populations are even rarer representing a significant gap in the literature. What studies do exist have mostly utilised qualitative methods, examining the opinions of individuals following participation in an intervention program (e.g., Stoffel & Braun, 2006; White et al., 2015). Whilst the feedback has been overwhelmingly positive, generalisability remains a concern.

In a notable exception to this, Rabbitt, Kazdin, and Hong (2014) surveyed parental acceptability of different intervention approaches, including AAT, for childhood disruptive behavioural problems. The study found that, whilst psychotherapy was rated the most acceptable intervention, AAT was more acceptable than medicinal approaches with little impact of prior pet ownership, animal‐related allergies or previous negative experiences with animals (Rabbitt et al., 2014). Despite providing vital preliminary insights into the acceptability of AAT for parents, there is a need to replicate and extend Rabbitt et al.'s study to provide a more nuanced understanding of acceptability within a PTSD/trauma context, and the use of dogs within therapy rather than generic ‘AAT’. Also, due to recommendations that AAT should complement rather than stand alone as a clinical intervention (Nimer & Lundahl, 2007), the current study will investigate the acceptability of AAT as an adjunct to traditional TF‐CBT.

Thus, the current study investigates the differential acceptability of TF‐CBT for Australian parents with the addition of two distinct adjunct therapies, psychoactive medication and CAT. Based on past literature that highlights potential variables of interest, the following hypothesises were proposed:

Hypothesis 1: Acceptability ratings of TF‐CBT with CAT will be greater than those of both TF‐CBT alone and TF‐CBT with medication.

Hypothesis 2: Prior experience with psychoactive medication will significantly affect overall acceptability ratings while prior experience with animals will not.

Hypothesis 3: Acceptability ratings across the three treatment approaches will differ as a function of the age of the child client.

Hypothesis 4: Attitudes towards animal welfare will significantly relate to ratings of the acceptability of TF‐CBT with CAT.

METHOD

Participants

Participants (N = 267) were Australian residents who identified as a current or past parent/caregiver of a child under the age of 18-years. This final sample size exceeded the minimum required sample size of 210 needed to establish a two‐tailed effect at p < .05, as determined by power calculations using GPower (Erdfelder, Faul & Buchner, 1996). Participants ranged in age from 18 to 99-years. The majority were female (90.6%), identified as White/Caucasian (85.5%), were married (55.1%) and had been a caregiver for three or less children (74.5%).

Materials and procedure

Following receipt of ethical approval via Central Queensland University Human Research Ethics Committee (approval number: H17/05‐086) participants were recruited using snowball sampling through posting and sharing the study information via the author's personal networks and university Facebook and research pages. Interested individuals were provided with information about the project, including a trigger warning (sexual abuse of a child) and a consent statement. Informed consent was indicated by selecting Yes to a participation question that then took participants to the survey which began with one of two vignettes (see Appendix A).

Trauma vignette

Participants were instructed to take the perspective of the child's parents whilst reading one of the vignettes which differed only in the age of the child victim (6 vs 13-years old). Both vignettes described the girl displaying behavioural changes following sexual abuse by a family friend. The behaviours described included social withdrawal, school avoidance, nightmares, trouble sleeping in her own bed, and disruptive, aggressive behaviour. Survey Monkey randomly allocated participants to a vignette with 147 participants (55.1%) reading the 6‐year‐old, and 120 (44.9%) reading the 13‐year‐old, vignette.

Treatment descriptions

After the vignette, brief descriptions of three potential treatment protocols for the child were presented in randomised order. The base description outlined a typical TF‐CBT approach while the other two added either psychoactive medication (i.e., taking a once‐a‐day antidepressant to aid in the reduction of anxiety and sadness to allow persistence in therapy) or a trained therapy dog to the TF‐CBT description. The dog was described as helping the child feel more at ease in the clinical environment, more motivated to attend therapy and providing physical proximity to help the child persist during the difficult parts of therapy.

Treatment Evaluation Inventory—Short Form (TEI‐SF)

Each treatment description was followed by the presentation of the nine‐item Treatment Evaluation Inventory‐Short Form (TEI‐SF) which was designed to assess parental acceptability towards analogue treatment descriptions for childhood syndromes (Kelley, Heffer, Gresham, & Elliott, 1989). Each item is answered via a 5‐point Likert scale with one reverse‐scored item to detect potential response bias. The TEI‐SF utilises more simplified wording than the original TEI in order to enhance understanding of the question content (Kelley et al., 1989). The internal constancy of the TEI‐SF is reported to be high (Cronbach's α = .85, Kelley et al., 1989). For the current study, internal consistency was acceptable, ranging between α = .78 and α = .92 for the three different measurements of the scale.

Demographic questionnaire

Participants answered several demographic questions regarding their age, gender, educational attainment, ethnicity, marital status, employment status, and number and ages of their children. They were also asked to indicate whether they or their dependent child had ever taken medication for a mental health problem (yes/no) or if they have ever had a pet dog (yes/no). Demographic questions followed those of treatment acceptability to prevent any potential priming effects.

Brief Animal Attitudes Survey (AAS‐10)

The brief, 10‐item version of the Animal Attitudes Survey (AAS‐10) was used to assess participant attitudes towards the ethical use of animals. The AAS‐10 is a shortened version of the original 20‐item questionnaire, designed to evaluate attitudes towards animal welfare and protection (Herzog, Grayson, & McCord, 2015). Items are answered via a 5‐point Likert scale with half of the items reverse‐scored. The internal consistency of the AAS‐10 is excellent (Cronbach's α = .90, Herzog et al., 2015). For the current study, internal consistency was acceptable at α = .74.

RESULTS

A total of 385 responses were collected, however, 107 of these were deemed too incomplete to meaningfully contribute to the analysis and were subsequently excluded; whilst a further 11 responses were excluded during assumption testing. The data was analysed using IBMs Statistical Package for the Social Sciences (SPSS) v24 (IBM, Armonk, NY, USA).

Main analyses

To examine the first three hypotheses, two mixed‐design, factorial analyses of variance (ANOVAs) and one Pearson's correlation analysis were used. The first analysis was a three‐way, mixed‐design, factorial ANOVA, with one within‐subjects factor of treatment type (TF‐CBT, with medication, with CAT) and two between‐subjects factors of experience with medication (yes, no) and experience with dogs as pets (yes, no). The two different vignette groups were collapsed into a combined sample for this analysis. Mauchly's test of sphericity was significant, χ ²(2) = 67.87, p = <.001; as the Greenhouse–Geisser estimate was ε = >.75, Huynh–Feldt corrections were used to adjust F‐ratios for the within‐subjects factor (Field, 2013).

The second analysis was a two‐way, mixed‐design, factorial ANOVA, with one within‐subjects factor of treatment type (TF‐CBT, with medication, with CAT), and one between‐subjects factor of hypothetical child's age (6‐year‐old, 13‐year‐old). Mauchly's test of sphericity was again significant, χ ²(2) = 70.61, p = <.001; as the Greenhouse–Geisser estimate was ε = >.75, Huynh–Feldt corrections were used to adjust F‐ratios for the within‐subjects factor (Field, 2013). The third analysis was a Pearson's correlation analysis, exploring whether attitudes towards animal welfare related significantly to the acceptability of TF‐CBT with CAT.

Bonferroni's correction was used to compensate for the problem of multiple comparisons, consequently, p = <.016 was accepted as the critical value for a statistically significant result. Due to a violation of sphericity and the effect this can have on the accuracy of contrasts using within‐subjects variables estimated marginal means were used rather than statistical contrasts to examine any significant ANOVA interactions (Field, 2013).

Treatment type

There was a significant main effect of treatment type on ratings of acceptability, F(1.66, 435.48) = 72.63, p = <.001, ${\eta }_p^2$ = .22. It was found that TF‐CBT with CAT (M = 34.30, SE = 0.46) was more acceptable than both TF‐CBT alone (M = 32.16, SE = 0.48) and TF‐CBT with medication (M = 26.22, SE = 0.73).

Experience

There was a significant treatment type‐by‐experience with medication interaction, F(1.66, 435.48) = 4.68, p = .014, ${\eta }_p^2$ = .02, indicating that acceptability ratings of the different treatment types varied significantly between those with and without experience with medication. It was revealed that those with medication experience rated TF‐CBT alone as less acceptable (M = 31.48, SE = 0.80) than those without experience (M = 32.83, SE = 0.53). Although TF‐CBT with medication was the lowest rated treatment in both groups, ratings were higher amongst those with medication experience (M = 27.62, SE = 1.22) than those without (M = 24.83, SE = 0.80). TF‐CBT with CAT was rated roughly the same between both those with (M = 34.25, SE = 0.77) and without (M = 34.34, SE = 0.51) medication experience.

The treatment type‐by‐experience with dogs as pets interaction was not significant, F(1.66, 435.48) = 0.20, p = .773, ${\eta }_p^2$ = .00, indicating that acceptability ratings of the different treatment types were roughly equal for those both with and without experience with dogs as pets. There also proved to be no significant treatment type‐by‐hypothetical child's age interaction, F(1.64, 433.15) = 3.59, p = .037, ${\eta }_p^2$ = .01, indicating that acceptability ratings of the different treatment types varied roughly equally between the different vignette groups. In contrast, a Pearson's correlation analysis revealed a significant small‐sized positive relationship between attitudes towards animals and the acceptability of TF‐CBT with CAT, r = .177, p = .004.

Demographic variables

In addition to the main analyses, several mixed factorial ANOVAs were conducted to explore if mean acceptability ratings across treatment types differed significantly as a result of demographic group membership. Only one demographic variable, marital status, significantly interacted with acceptability ratings for the different treatment types. Specifically, while acceptability ratings for TF‐CBT with CAT were similar in all groups, TF‐CBT alone was much less acceptable for the single/never married group. Additionally, whilst TF‐CBT with medication was the least acceptable for all groups, the change in acceptability ratings was far less prominent for the single/never married group.

DISCUSSION

The overall aim of the current study was to evaluate differences in perceived parental acceptability of TF‐CBT, the current gold standard intervention for childhood trauma, both alone and with the addition of either psychoactive medication or CAT. Only three of the four hypotheses were supported. Firstly, as predicted, participants rated TF‐CBT with CAT as more acceptable than either TF‐CBT alone or with TF‐CBT with medication; in fact, medication as an adjunct was rated considerably lower than both other approaches. Predictions regarding experience were also supported, finding that experience with medication interacted significantly with acceptability ratings of TF‐CBT alone and TF‐CBT with medication, whilst experience with dogs as pets did not. Lastly, the significant positive relationship between attitudes towards animal welfare and the acceptability of CAT indicates that concern for animal welfare relates positively to attitudes towards utilising canines in psychotherapy. Contrary to predictions, age of child was not found to impact treatment acceptability.

Whilst these findings support the predictions of the current study, they only somewhat replicate the findings of Rabbitt et al. (2014), who concluded that psychotherapy was more preferred than AAT. However, this may have resulted from the approach taken here whereby CAT was presented as an adjunct, rather than standalone, intervention. This approach was taken in recognition of the fact that there is not currently enough empirical backing to support AAT or CAT as standalone interventions, and consequently, clinicians may be unlikely to use or recommend them as such (Geist, 2011; Hoagwood et al., 2017; Nimer & Lundahl, 2007). What the current study demonstrates is that consumers appear to share this view, and may hold similar hesitations towards abandoning trusted and familiar treatment options, such as TF‐CBT. However, augmentation may provide a suitable compromise, particularly given low engagement and high attrition rates amongst children in therapy (Wamser‐Nanney & Steinzor, 2017). If the efficacy of empirically robust therapies like TF‐CBT can be enhanced by improving acceptability, as is suggested by research (Swift et al., 2013), then augmenting TF‐CBT with CAT may hold the key to improving outcomes for children who have experienced trauma. This also represents a much more achievable goal for researchers keen to solve these complex therapy engagement problems.

Overall, medication was the least acceptable adjunct therapy, and whilst those with prior experience with psychoactive medications were more accepting of this option, the acceptability of TF‐CBT alone was much higher. This aligns with prior studies where negative attitudes towards medication as a psychological intervention have also been noted (e.g., Kehle‐Forbes et al., 2014). A consideration of treatment acceptability is vital in this circumstance, as it is found to significantly impact eventual efficacy through lack of motivation and compliance (Swift et al., 2013). The current findings must be viewed in light of this; if augmenting TF‐CBT with medication reduces its acceptability then this may not be a beneficial approach. This is particularly salient in Australia given the shift in intervention decision‐making that is expected due to NDIS. Therefore, a useful agenda may be to examine other, more acceptable, complementary therapies to inform future clinical practice as has been aim of the current study. Along this line, it may be useful to examine whether combining CAT with medication as co‐adjunct therapies might improve acceptability of pharmacological therapy in future studies.

In contrast to medication, prior experiences with dogs as pets had little effect on the acceptability of TF‐CBT, with and without the addition of CAT. While this supports suggestions of prior research that there is a generalised tendency towards acceptance of AAT (e.g., White et al., 2015), it also contradicts others studies which suggested that a client's attitude towards CAT might hinge on their perceptions of, and prior experiences with, dogs (Black et al., 2011; Yap et al., 2016). These worries are not necessarily groundless; many individuals dislike interacting with dogs (Lefkowitz et al., 2005; Muris, Van Zwol, Huijding, & Mayer, 2010). In particular, individuals from culturally and linguistically diverse backgrounds can have complex relationships with some animal species, including dogs, resulting from religious beliefs and fears of uncleanliness (Jegatheesan, 2015). Whilst the current study did not find evidence of such concerns, further investigation is needed. In particular, a replication of the current study with participants from cultures known to have differing perceptions of animals (Jegatheesan, 2015).

In prior AAT acceptability research, medical and psychiatric staff raised concerns about the welfare of therapy animals suggesting that this could be a major barrier to implementation (Black et al., 2011; Lefkowitz et al., 2005). However, the current study has found that even those with a high concern for animal welfare were accepting of incorporating dogs into psychological therapy. These findings have important implications for the direction of future research and practice. Whilst concern for therapy animal welfare is an ethical imperative, and should be considered thoughtfully by AAT clinicians (Black et al., 2011), animal welfare might not factor into the evaluation of treatment options by consumers.

The fact that predictions regarding the age group most suited for CAT‐augmented interventions were not supported only lends further support to the versatility of the practice. Research indicates that younger children might benefit more from TF‐CBT with CAT due to difficulties that exist engaging them in traditional talk‐based psychotherapies (Campbell & Simmonds, 2011; Miller‐Graff & Campion, 2016). However, the fact that parents have similar positive attitudes towards this approach for older children suggests an even larger scope of use which is backed by prior research. Studies have found that adolescents feel much more at ease interacting with a dog than an adult therapist (Beetz et al., 2012). Dogs are attentive to humans, regardless of their age, and children can share stories with dogs without worry of repercussions, which may be particularly useful in the case of CSA (Zilcha‐Mano et al., 2011). Animals introduce a relationship dynamic that may be absent from the traditional child‐psychotherapist dyad which has been found to enhance the child's feelings of purpose and mastery (Hamama et al., 2011). This is a key outcome sought in other AATs as well, particularly equine‐facilitated psychotherapy, where the taming of such a large animal is a favoured approach for use with older children (Mueller & McCullough, 2017). Going forward, it may be useful to gauge the acceptability of equine therapies for different age groups as well, which may contribute to building a more comprehensive framework of the uses of different AAT approaches for various clinical populations.

Overall, this has been the key focus of the current study to contribute to shaping a more comprehensive framework of the applicability of AAT therapies for children. The finding that marital status may impact the acceptability of different therapies is intriguing, suggesting that family dynamics may affect attitudes towards different therapy approaches. However, this also strengthens the argument for the usefulness of AAT as, unlike the other suggested interventions, the acceptability of TF‐CBT with CAT remained unaffected by marital status.

Limitations identified in the current study may provide additional directions for subsequent research. Firstly, a significant proportion of participants (approximately 28%) failed to complete the survey. It seems possible that presenting CSA as the source of trauma in the vignettes was particularly confronting for parents. If this is so, then it is possible that the current findings may be specific to post‐CSA therapy rather than childhood trauma more generally. It would be useful in subsequent studies to compare adjunct treatment acceptability between different types of traumas, presenting conditions and whether the parent currently has a child engaging in therapy. The difficulties engaging children in traditional talk therapies may not necessarily be exclusive to traumatised populations, as these problems seem to stem more from developmental immaturity more so than the syndrome (Green, 2006).

Further to this, the inequitable sizes of several demographic sub‐groups made it difficult for the current study to draw robust conclusions about the potential effect of these variables. Two variables in particular would benefit from a more balanced examination in subsequent AAT/CAT research, ethnicity and gender. Prior studies have revealed that ethnicity is a particularly influential demographic characteristic in regard to the use of and orientation towards animals (Signal, Taylor, Botros, & Lazarus, 2013). The current study was unable to find any meaningful impact of ethnicity, which most likely resulted from the small number of participants from minority groups. Gender suffered from a similar disproportion in sample size. Whilst it is generally more common for women to respond to survey questionnaires than men (Porter & Whitcomb, 2005), it makes it somewhat problematic to suggest that the current findings generalise optimally to the attitudes of both male and female caregivers. As different genders can have vastly different approaches to psychotherapy (Wang et al., 2007) as well as parenting practices (McKinney & Renk, 2008), future studies should aim for more equitable gender samples.

The fact that current, potentially effective, interventions fail to engage many CSA victims suggests a need for a creative approach (Wamser‐Nanney & Steinzor, 2017). An intervention can only be useful if it is, in fact, used. Which raises an important issue that should form the apex of continuing intervention‐planning research, that of ensuring consumers feel able to connect with the interventions that can assist them to recover. This study suggests that augmenting traditional therapies is one way in which this agenda can be accomplished.

The most important contribution of the current study is the support it provides for the fact that consumers favour the inclusion of animals in psychotherapy. Whilst this study has focused on the use of canines as an animal and children with trauma as a clinical population, other animal species and other therapeutic approaches have also shown promise (Kemp et al., 2014; Signal et al., 2013). Currently, the state of research on the efficacy of AAT in psychotherapy is poor (Hoagwood et al., 2017). Whilst a lack of empirical rigour has made drawing solid conclusions difficult, the lack of a uniformed research agenda has made it almost impossible for standardised and replicable procedures to be designed and tested (Geist, 2011). Consequently, there is a need for a more focused approach to research on the topic of AAT, which begins with assessing the viability of the therapeutic approach amongst its potential consumers and ends with the development of empirically rigorous and manualised interventions.

Notes

Please note Janene would like her married name to appear on this article (previously Arnold‐Hancock).

Taylor's parents are becoming increasingly worried about a dramatic change in her behaviour. Their usually confident and social 6‐year‐old/13‐year‐old now seems to have trouble being around other people.

Taylor has begun refusing to leave the house, especially to go to places where there are lots of people, like the shopping centre. She tries everyday to get out of going to school, sometimes by pretending to be sick but mostly by crying and begging her parents to let her stay home. When she does go to school, teachers say she has become very disruptive and at times can be aggressive towards other kids. At lunch breaks, she always sits on her own now.

Taylor becomes very afraid and shy around adults, especially ones she has never met before. She clings tightly to one of her parents and won't look them in the eye or speak, even when she is asked questions.

Taylor's behaviour is the most problematic at night time. She no longer wants to sleep in her own bed and cries and begs to sleep in her parents bed. She awakens several times during the night screaming because of terrible nightmares.

When Taylor's parents discover that she has been sexually abused by a family friend, they visit the family doctor who worries that her change in behaviour might be the result of PTSD. The doctor suggests that Taylor and her parents visit a psychologist.

APPENDIX A Treatment descriptions

TF‐CBT: Taylor, along with her parents, attend clinic‐based therapy sessions—some together and some individually—to help them accept and cope with her traumatic experience. Taylor learns strategies to use when difficult thoughts and feelings arise that cause strong emotions like fear, stress, sadness, and anger. Throughout the sessions, she re‐processes her trauma safely by working through thoughts and beliefs under the guidance of the psychologist. The goal is to learn to accept and live with what happened rather than continuing to avoid reminders of the experience. Over time, thinking about the trauma becomes less and less painful for both Taylor and her parents.

TF‐CBT adjunct medication: Taylor takes an anti‐depressant medication whilst attending clinic‐based therapy sessions with her parents—some together and some individually—to help them to cope with her traumatic experience. The medication alters chemical reactions in the brain which helps to reduce the anxiety and sadness when thinking about the traumatic experience. Throughout the sessions, Taylor re‐processes her trauma safely by working through thoughts and beliefs under the guidance of the psychologist. The goal is to learn to accept and live with what has happened rather than continuing to avoid reminders of her experience. Because of the medication, Taylor feels more prepared to participate in these difficult parts of therapy and, over time, thinking about the trauma becomes less and less painful for her and her parents.

TF‐CBT adjunct canine therapy: Taylor and her parents attend therapy sessions with a psychologist and trained therapy dog—some together and some individually—to help them accept and cope with her traumatic experience. The dog helps Taylor to feel more at ease in the clinic environment and more motivated to attend therapy. Throughout the sessions, she re‐processes her trauma safely by working through thoughts and beliefs under the guidance of the psychologist. The goal is to learn to accept and live with what has happened rather than continuing to avoid reminders of her experience. The dog can sense when Taylor is experiencing difficult emotions like fear and sadness and provides physical proximity and comfort which helps her to persist through the difficult parts of therapy. Over time, thinking about the trauma becomes less and less painful for both Taylor and her parents.