Clinician adoption of Parent–Child Interaction Therapy: A systematic review of implementation interventions

Background

A degree of defiance, non-compliance and aggression is common in young children, particularly at certain stages of development. However, for some children, these problems persist and begin to detrimentally impact child and family wellbeing. Such conduct problems represent one of the most common reasons for children to be referred to mental health services (Coghill, 2013). The social and economic consequences of untreated childhood conduct problems are significant, and it has been suggested that ‘… there is no other commonly occurring childhood condition that has such far-reaching and pervasive consequences for later health, development and social adjustment’ (Fergusson et al., 2011, p. 60). Longitudinal studies indicate that left untreated, children with conduct problems may develop into adults who require intensive support from services across criminal justice, social welfare and health domains (Rivenbark et al., 2017). As such, ensuring the effectiveness of community-based treatment is essential (Coghill, 2013; Furlong et al., 2013).

Parent training interventions based on social learning theory and operant learning principles (also known as parent management training or behavioural parent training; hereafter referred to as ‘parent training’) are well established as an effective treatment for childhood conduct problems, being amongst the most comprehensively reviewed treatments for childhood psychiatric conditions worldwide (Leijten et al., 2019). These interventions typically include both a parent–child relationship enhancement component and behaviour management skills training for parents (Leijten et al., 2018). They have been shown to be effective in reducing child conduct problems (Furlong et al., 2013; Ward et al., 2020). Through improving parental mental health and psychosocial wellbeing, parent training interventions influence the nature and quality of the parent–child relationship, in turn contributing to enhanced child wellbeing (Barlow et al., 2014; Furlong et al., 2013). Parent training programmes may also reduce the incidence of child physical abuse by providing alternative, non-contact methods of discipline, reducing harsh or negative parenting practices and increasing positive parenting skills (Furlong et al., 2013; Knerr et al., 2013).

Parent–Child Interaction Therapy (PCIT; Eyberg and Funderburk (2011)) is a well-established example of a parent training intervention (Lieneman et al., 2019; Thomas et al., 2017; Ward et al., 2016). PCIT is somewhat distinctive from many parent training approaches in its use of live coaching of parents with their child by trained therapists, typically by way of a one-way mirror and coaching headset. Originally developed by Sheila Eyberg in the USA in the 1980s, PCIT has been disseminated internationally to countries including Germany, Hong Kong, Australia, New Zealand, the Netherlands and Japan. Accreditation in PCIT requires Master's level licensed clinicians to attend a five-day intensive training, followed by fortnightly supervision until the successful completion of two PCIT cases (PCIT International; www.pcit.org).

However, PCIT faces a similar challenge to other evidence-based parent training interventions, in that, despite demonstrated effectiveness, and considerable international dissemination efforts, uptake in usual care settings remains poor (Forgatch et al., 2013; Lyon et al., 2020; Michelson et al., 2013; Williams & Beidas, 2019; Woodfield et al., 2020). Instead, children and families may be offered clinician-preferred interventions, and/or lower-intensity delivery of evidence-based protocols, both of which are less likely to represent an efficacious treatment (Williams & Beidas, 2019). To address this, research into how to integrate parent training effectively and sustainably into clinical care settings is critical, particularly given suggestions that implementation research within mental health services has lagged behind other disciplines (Powell et al., 2014).

Implementation difficulties may occur at a patient, clinician, clinic, system or policy level or indeed, often across multiple levels (Bauer et al., 2015). Yet despite this complexity, human behaviour remains at the core of successful implementation (Presseau et al., 2019). A key requirement of successful implementation is ‘for someone (usually more than one person or group), somewhere (from organisational leadership through to those providing direct patient care) doing something (usually more than one thing) differently … taking up new evidence requires healthcare providers and other health system stakeholders to change their behaviour’ (Presseau et al., 2019, p. 2, emphasis original).

Implementation strategies are single-component or, more typically, multifaceted co-ordinated attempts to address identified determinants of practice, with the ultimate aim of bringing about behaviour change (Michie et al., 2011). Many implementation strategies working in harmony are often required to implement a treatment successfully (Rudd et al., 2020). Of note, the terms implementation strategy and implementation intervention have a variety of meanings in different contexts but are used interchangeably in this paper.

Within the field of implementation science, there are calls for better specification and operationalisation of implementation strategies, and the field is also being challenged to better understand the intermediary in this relationship, that is, the mechanism by which a particular strategy may act on an implementation outcome (Lewis et al., 2020; Lewis et al., 2018; Powell et al., 2019). Hypothesised mechanisms of action are often not specified, let alone tested (Powell et al., 2019). Those working in the field are being encouraged to attempt to understand ‘the processes and mechanisms by which (implementation) strategies exert their effects rather than simply establishing whether or not they were effective’ (Powell et al., 2019), as connecting implementation strategies to hypothesised mechanisms of action can enhance understanding of why, and for whom, such strategies are effective (Lewis et al., 2020; Powell et al., 2019). By way of example, in Table 1, we have listed the Cochrane EPOC group's taxonomy of implementation strategies targeted at clinicians and have extended this with examples from the literature of the mechanisms thought to bring about behaviour change.

OPEN IN VIEWER

Table 1.

Examples and definitions of implementation strategies, and hypothesised mechanisms for action.

Example of behaviour change strategy targeting clinician behaviour (Cochrane Collaboration Effective Practice & Organisation of Care (EPOC) Group, 2015)	Definition (adapted from (Cochrane Collaboration Effective Practice & Organisation of Care (EPOC) Group, 2015)	Hypothesised or possible mechanism(s)
Audit and feedback	A summary of clinical performance over a period of time, may include recommendations for clinical action, provided via written, verbal or electronic means.	Alters clinician awareness of their current practice and clinical consequences of this; alters clinician's perceived social norms; impacts self-efficacy; draws clinician attention to a particular group of tasks and activities (Ivers et al., 2012).
Clinical incident reporting	A system for reporting critical incidents.	Organisational and individual learning from the accumulated data, used to reduce actual harm and potential harm (Stavropoulou et al., 2015).
Monitoring of the performance of the delivery of healthcare	Monitoring of service provision by healthcare organisations or individuals, e.g. comparing with an external standard.	Improves knowledge about performance, processes and results (Berwick et al., 2003).
Communities of practice	Groups or collectives of professionals with a common interest who interact regularly to deepen their knowledge or understanding in a particular area.	Knowledge generation and information sharing, collective problem-solving, reducing professional isolation, development of practice aids (Ranmuthugala et al., 2011).
Continuous quality improvement	An iterative process to evaluate, refine and improve healthcare that may include analysis of a system or process, use of data analysis to evaluate changes, and/or a structured problem-solving approach.	Shared valuing of clinical data for practice improvement, collective change efficacy and/or the development of a population health orientation (Schierhout et al., 2013).
Educational Games Materials Meetings Outreach visits or academic detailing Inter-professional education	Continuing education for clinicians, that may involve seminars, workshops, online learning, personal outreach visits, feedback on performance, educational materials to support clinical care, or interactive learning with other healthcare clinicians.	Improved awareness of best practice and/or new developments (O'Brien et al., 2007). Educational games may allow for ‘safe’ experimentation with problem-solving and decision-making in a controlled learning environment, stimulating use of analytical skills and reflective practice (Akl et al., 2013).
Clinical Practice Guidelines	Systematically and rigorously developed (often printed) resources to aid healthcare clinician decision-making around optimal healthcare for specific clinical circumstances.	Passive dissemination of scientific information to improve practice via increased awareness, knowledge, skills and changes in attitudes (Giguère et al., 2012).
Local consensus processes and/or local opinion leaders	Promotion of good clinical practice through the use of identified opinion leaders in the local context, or a consensus agreement on a local process.	Opinion leaders (cf. ‘champions’, who are selected by management) are respected by peers, and exert social influence to promote improved practice (Flodgren et al., 2019).
Managerial supervision	Routine supervision by healthcare staff in a management role.	Mechanism likely to depend on context – for example in low- and middle-income countries, may serve to reduce professional isolation and provide support to clinicians, or line manage logistics and finances around care provision (Bosch-Capblanch et al., 2011).
Patient-mediated interventions, including routine patient-reported outcome measures	Interventions which use information provided by/to patients and/or interactions with patients to change clinician performance.	Patient feedback may broaden information available to the clinician and prompt improvement. Informed patients may be more actively involved in their care, alongside increased self-efficacy and self-care (refer Table 1 in Fønhus et al., 2018 for a full description of possible mechanisms and positive and adverse effects).
Public release of performance data	Written or electronic performance data, released to inform the public about healthcare clinicians.	Informed patients may select better performing clinicians, perhaps providing an indirect stimulus or incentive for quality improvement (Metcalfe et al., 2018).
Reminders	Interventions (manual or computerised) which involve prompting clinicians to perform a particular behaviour, for example, during a consultation.	Serve as prompts for clinicians to recall known information, and/or present information in a more comprehensible format (Pantoja et al., 2019).
Tailored interventions	Interventions to change practice which are based on an assessment of barriers to change.	The implementation strategy is deliberately chosen to address specific pre-identified barriers or obstacles in a particular context or setting (Baker et al., 2015).
Financial incentives (Not an EPOC example, but often included – for example, Dombrowski et al. (2016))	Interventions which involve providing a monetary reward to clinicians for a particular time period; service, episode or visit; for providing care for a particular population group; a change in activity or quality of care – with the aim of increasing the quality and efficiency of care (Flodgren et al., 2011).	Extrinsic source of motivation, conditional on performing a particular behaviour or set of behaviours, though may inadvertently reduce intrinsic motivation (Flodgren et al., 2011).

Existing reviews

Despite a recent upswing in interest in implementation research in child psychology and psychiatry, there are very few existing reviews of implementation strategies to increase therapist use of parent training approaches in usual care settings. Forgatch et al. (2013)'s review related solely to the implementation of the Parent Management Training-Oregon Model (PMTO) programme and illustrated the process of implementing PMTO in three locations: Norway, Michigan and Kansas. The review was narrative rather than systematic and aimed to highlight important influences on successful implementation, along with challenges. The authors wrote of the importance of ‘combin[ing] the puzzle pieces, captur[ing] the key components, and build[ing] a complete picture’ (p. 692) – a suggestion to move the parent training implementation field forward through collaboration and integration. As applied to the review context, this suggests a need to rigorously and systematically synthesise individual implementation studies. However, this may need to occur through synthesising the implementation literature relating to a specific programme, as implementation strategies are ideally tailored to specific treatments and contexts (Baker et al., 2015). A rigorous and systematic review of the literature relating to an individual parent training intervention sets the scene for the field-wide collaboration that Forgatch et al. (2013) promote.

Also, a new look at the evidence is timely. Baumann et al.'s (2015) review of four evidence-based manualised programmes: Parent–Child Interaction Therapy, The Incredible Years, Parent-Management Training-Oregon Model (PMTO) and the Positive Parenting Programme (Triple P) included articles published up to January 2013. Their review assessed the extent to which implementation strategies and implementation outcomes have been evaluated when integrating these parent training interventions into usual care. At that time, the authors suggested that parent training research had been disproportionately weighted towards effectiveness studies but was ‘poised’ to undertake more research into implementation strategies (Baumann et al., 2015). Given that implementation, science is a new and rapidly evolving field, and the importance of tailoring interventions to specific treatments, an up-to-date review of implementation strategies specific to PCIT is particularly important.

The aim of this systematic review was to synthesise the existing literature on implementation strategies designed to increase clinician use of PCIT in the treatment of childhood conduct problems. Specifically, we aimed to:

Describe the nature of implementation strategies used and their constituent components according to items from the Template for Intervention Description and Replication (TIDieR) checklist (Hoffmann et al., 2014).

Methods

Results

Article selection

The search was run on 10 October 2020. The article selection and screening process is outlined in the PRISMA flow diagram (Figure 1). The search yielded 2150 results and after duplicates were removed, 769 abstracts were manually and independently reviewed by MW and HK. Full texts of 17 papers were retrieved, and reviewed by MW, SH and SM. A total of 13 papers relating to three studies were included (Table 2).

OPEN IN VIEWER

Figure 1.

PRISMA flow diagram.

OPEN IN VIEWER

Table 2.

Summary of included studies.

	Protocol	Primary publication(s)	Secondary publications
1		Herschell (2003) Evaluation of techniques for disseminating PCIT (PhD dissertation)	Herschell, A. D., McNeil, C. B., Urquiza, A. J., McGrath, J. M., Zebell, N. M., Timmer, S. G., & Porter, A. (2009). Evaluation of a treatment manual and workshops for disseminating, parent–child interaction therapy. Administration and Policy in Mental Health and Mental Health Services Research, 36(1), 63-81.
2	Funderburk et al. (2011) Trial protocol (NCT01294488) Using technology to augment the implementation and effectiveness of PCIT	Funderburk, B., Chaffin, M., Bard, E., Shanley, J., Bard, D., & Berliner, L. (2015). Comparing Client Outcomes for Two Evidence-Based Treatment Consultation Strategies. Journal of Clinical Child & Adolescent Psychology, 44(5), 730-741.	Nelson, M. M., Shanley, J. R., Funderburk, B. W., & Bard, E. (2012). Therapists’ attitudes toward evidence-based practices and implementation of parent–child interaction therapy. Child Maltreatment, 17(1), 47–55.
3	State wide trial to compare three training models for implementing an evidence-based treatment (NCT02543359; Herschell; ClinicalTrials.gov) Herschell, A. D., Kolko, D. J., Scudder, A. T., Taber-Thomas, S., Schaffner, K. F., Hiegel, S. A., …  Mrozowski, S. (2015). Protocol for a statewide randomized controlled trial to compare three training models for implementing an evidence-based treatment. Implementation Science, 10(1), 1–16.		Scudder, A. T., Taber-Thomas, S. M., Schaffner, K., Pemberton, J. R., Hunter, L., & Herschell, A. D. (2017). A mixed-methods study of system-level sustainability of evidence-based practices in 12 large-scale implementation initiatives. Health research policy and systems, 15(1), 1–13. Jackson, C. B., Herschell, A. D., Schaffner, K. F., Turiano, N. A., & McNeil, C. B. (2017). Training community-based clinicians in parent–child interaction therapy: The interaction between expert consultation and caseload. Professional Psychology: Research and Practice, 48(6), 481. Jackson, C. B., Macphee, F. L., Hunter L. J., Herschell, A. D., & Carter, M. J. (2017). Lessons learned: Enrolling family participants in a statewide implementation trial of an evidence-based treatment. Progress in Community Health Partnerships: Research, Education, and Action 11(3), 233–241; Brabson, L. A., Herschell, A. D., Kolko, D. J., & Mrozowski, S. J. (2019). Associations among job role, training type, and staff turnover in a large-scale implementation initiative. The Journal of Behavioral Health Services & Research, 46(3), 399–414. Jackson, C. B. (2020). Making Implementation Last: Understanding the Sustainability of an Evidence-based Treatment. Dissertation. Brabson, L. A., Herschell, A. D., Snider, M. D., Jackson, C. B., Schaffner, K. F., Scudder, A. T., …  Mrozowski, S. J. (2020). Understanding the Effectiveness of the Cascading Model to Implement Parent-Child Interaction Therapy. The Journal of Behavioral Health Services & Research, 1–19.

Discussion

Overall, the studies identified in this systematic review of PCIT implementation interventions were modest in number (three). Strengths included the ‘real world’ context within which studies were based, and the detailed examination of characteristics associated with successful training and post-training consultation methods. Implementation outcomes have not yet been published from one large study (Herschell et al., 2015), were not directly reported in another (Funderburk et al., 2015) or were limited to those assessed at the end of the final day of training and not upon the clinicians’ return to their treatment delivery context (Herschell et al., 2009).

In recent years, there has been increasing recognition of the importance of describing mechanisms of action of implementation strategies. However, mechanisms tended not to feature prominently in the design of these studies, which in some cases were carried out some time ago. Robust training of clinicians in a particular treatment approach is often a necessary pre-condition for successful implementation in a clinical context and warrants research attention (Frank et al., 2020). Training serves an important role – it may increase clinician knowledge and satisfaction with an approach (Valenstein-Mah et al., 2020). However, it remains unclear whether and how training results in enhanced delivery of treatments by clinicians, and better outcomes for clients (Valenstein-Mah et al., 2020). Future research into training ought to seek to understand mechanisms by which training does (or does not) result in clinician behaviour change as it relates to adoption and sustainment, along with how other contextual factors (such as organisational context) inter-relate with training experiences to result in adoption (Frank et al., 2020). As an example of a description of a hypothesised mechanism, training and training-related consultation are thought to impact implementation behaviour by way of expert modelling/rehearsal of skills, problem-solving, improving engagement and motivation, therapist accountability, adaptation or tailoring to improve fit, planning for sustainability of the treatment (Motamedi et al., 2021). Herschell et al. (2009) justifiably note that several limitations in the design and execution of their study were a result of its ‘real world’ nature, ‘which is both a strength and weakness of the current study … (there is a) need to balance rigor with relevance in dissemination and implementation studies, which leads to necessary compromises at all levels’ (p. 77). Indeed, while researchers may be aware of the importance of testing hypothesised mechanisms of action, factors outside of researchers’ control, for example, contractual agreements and/or community relationships, can influence the design and execution of an implementation study (Herschell et al., 2009).

There was an emphasis in all included studies on training as an implementation strategy. Funderburk et al. (2015) importantly observed that post-training drift (i.e. poor fidelity to the treatment protocol) and poor transferability of traditional training models from university to community settings are challenges facing the field. But across the three studies, there does not appear to have been a systematic assessment of barriers to implementation of PCIT in the particular context (ideally in partnership with stakeholders, including clinicians), followed by a deliberate selection of implementation strategies to address the specific barriers in the setting. Considering clinician perspectives is particularly relevant to the delivery of evidence-based interventions (Motamedi et al., 2021). In a recent mixed-methods study, the majority of implementation strategies perceived by child therapists as essential for sustained delivery of interventions (connection to a community of trained therapists, ongoing consultation and supervision, availability of internal supervisors trained in the intervention) were indeed significantly associated with the use of evidence-based practices, while only one (access to intervention-related resources or materials) was not (Motamedi et al., 2021). It is apparent that comprehensive consultation is required with those charged with PCIT implementation (i.e. clinicians) to better understand barriers, and shape and prioritise implementation interventions to address these.

Also, while the included studies ostensibly explored training and training-related consultation as implementation strategies, their intervention often had other – potentially very influential – components. For example, inclusion criteria for participating agencies typically required that the agency had PCIT equipment in place and a commitment to a suitable referral stream (e.g. Funderburk et al., 2015), which meant some of the more common barriers which may be faced by PCIT clinicians downstream, such as having an appropriate client flow and installing the equipment needed, were pre-emptively addressed (Barnett et al., 2021; Woodfield et al., 2020). It may have been useful to consider this as an implementation strategy in and of itself. Though it is likely that the authors intentionally addressed logistical or practical barriers in order to examine the impact of training conditions alone and minimise confounding factors.

Within the field of implementation science, there is a need for increased consistency in relation to the terms used and their definitions, as overlapping and variable use of terms is common and problematic (Lewis et al., 2020). Mechanisms may be referred to as mediators, and determinants or barriers as outcomes – for example, poor fidelity was highlighted as a barrier to be addressed by the implementation intervention in studies included in this review and is also an implementation outcome (Proctor et al., 2011). We share the call by Lewis et al. (2020) for increased clarity in distinctions between determinants, strategies and outcomes. Finally, different implementation outcomes are relevant at different stages of an implementation process, with factors influencing the sustainment of an intervention likely to be different from those influencing its initial adoption (Shoesmith et al., 2021). While adoption is, by definition, necessary for sustainment, few would define a successful implementation as one which simply results in the initial adoption of an intervention without any sustained use. A decline in the use of evidence-based practices such as PCIT following implementation is often described, but not well understood (Birken et al., 2020). Yet apart from Herschell et al. (2015)'s study, the outcomes of which are yet to be published, there are no data on implementation outcomes after the PCIT training period. As such, while these are valiant and important studies, more work is required to better understand downstream (i.e. post-training) influences and processes.

Limitations

We limited our search to publications in English only and did not search the grey literature. It is possible that relevant publications may have been missed with these search parameters. The studies included in this review were relatively homogeneous, in that they were all conducted in the USA, and did not include specific consideration of the needs of minority groups such as Indigenous clinicians. The review resulted in a modest, and heterogeneous pool of included studies. This is a key limitation to any conclusions that can be drawn relating to PCIT implementation interventions. Also of note, one researcher (Dr Amy Herschell) was involved in the design and execution of two of the three included studies, which introduces the possibility that effects may be associated with her considerable expertise, rather than the implementation interventions themselves. This may also limit the generalisability of the findings of this review.

The process of conducting a systematic review requires one primary outcome to be identified, for the purposes of data analysis. We elected to use adoption as a primary outcome, and this decision may warrant further exploration. As stated earlier, it has been suggested that the quality of adoption is as important as the quantity of adoption (Beidas & Kendall, 2014). Also, implementation outcomes are known to be inter-related, with some outcomes even thought to be predictive of others. For example, acceptability, feasibility and appropriateness of the intervention being implemented may predict adoption by clinicians (Proctor et al., 2011).

Conclusion

In summary, while the methodology of studies is increasingly robust, and there has been an increase in implementation-related PCIT research in the past decade, it is clear that the focus remains on training and training-related consultation as implementation strategies, with increases in clinician knowledge and/or skills, and acceptability or satisfaction of the training itself as implementation outcomes. The stage is set for consideration of how best to intervene to sustain clinician use of PCIT in community settings post-training.

PCIT has a compelling evidence-base, and the potential to effectively treat childhood conduct problems – a relatively commonly occurring and potentially debilitating condition. However sustainably embedding evidence-based interventions such as PCIT in usual care environments requires thoughtful and deliberate effort, and rarely occurs naturally (Shelton et al., 2020). Existing PCIT implementation initiatives have focussed on training and training-related consultation strategies and the stage is set for implementation interventions that attempt to improve sustainment and even attempt ‘re-implementation’ where implementation has failed or fallen short in real-world settings. This process ought to involve pre-identified barriers identified in partnership with stakeholders, and set out to test specific mechanisms of action, to not only understand what works in implementation interventions, but why and how.

Supplemental Material

Supplemental Material