Positive Behaviour Support: RCT finds no effect, but fidelity is poor


Imagine this: You read the results of a randomised controlled trial (RCT) that evaluates the effectiveness of a new drug in reducing symptoms of depression. The RCT reports that this new pill performed no better than the current treatment-as-usual. However, when describing the results, the authors mention that the medication was much weaker than intended and only contained one third of the expected ingredients. Would you think that the medication was ineffective, or be unable to judge because the drug wasn’t administered as planned?

This same issue is evident in the RCT recently published by Hassiotis et al. (2018). The UCL PBS study was designed to evaluate the effect of a training intervention on the challenging behaviour of adults with learning disabilities (LD). Professionals from community learning disability teams (CLDTs) were trained to deliver Positive Behaviour Support (PBS) in a pragmatic trial designed to assess the practicality of using existing resources to deliver PBS.

Prevalence studies indicate that 10-15% of people with LD are likely to engage in “challenging behaviour” that physically harms themselves or those around them or limits normal community opportunities. PBS is a multi-component framework that incorporates techniques based on the science of applied behaviour analysis, such as functional assessment, to address challenging behaviour and improve quality of life (Gore N. et al, 2013). The core components of PBS include a data-driven approach, a focus on understanding why challenging behaviour is occurring and a commitment to teaching skills that enable the individual to meet their needs effectively, without having to rely on challenging behaviour.

Use of PBS-based approaches is recommended as part of the Transforming Care agenda (Transforming Care and Commissioning Steering Group, 2014), by the Local Government Association and NHS England (2014) and in a number of relevant NICE guidelines (e.g. National Institute for Health & Care Excellence, 2018). Delivery of PBS usually involves specialist or consultant-level behavioural assessment and intervention planning, leading to PBS plans that are implemented by direct support staff or family members/carers.

The question posed by the UCL PBS study was this:

Can the varied professionals usually found in a community learning disability team:

  • be trained to complete functional assessments
  • write associated behaviour support plans
  • effect a reduction in challenging behaviour occurrence for people with LD?
10-15% of people with learning disabilities are likely to engage in "challenging behaviour".

10-15% of people with learning disabilities are likely to engage in “challenging behaviour”.


A published trial protocol (Hassiotis A. et al, 2014) describes the methodology of the single-blind, parallel, two-arm cluster RCT in detail. Twenty-three CLDTs participated, with a cluster of 11 teams allocated to receive the intervention and deliver PBS plus treatment-as-usual (TAU), while the remaining 12 delivered TAU only. A total of 21 professionals received PBS training before completing assessments and support plans for 108 people with learning disabilities. Plans were implemented by family members (18%) or paid carers (82%).

The intervention applied in the trial involved training professionals of various backgrounds (such as speech and language therapists, occupational therapists and clinical psychologists) to complete specialist/consultant-level PBS activities, namely:

  • completing functional behavioural assessments
  • developing multi-component function-based PBS plans
  • supporting family members and paid carers to implement plans effectively.

A training manual supported this process and specified the intervention and behavioural support plan components to be completed when putting the training into practice.

Training was delivered for a total of six days in three separate two-day workshops. A 30% reduction in usual caseload was requested by the research team to enable professionals to complete assessments and interventions as trained.

Challenging behaviour reduction, as measured by informant interview using the Abberant Behaviour Checklist-Community, was the primary outcome measure. Secondary measures included indicators of mental illness, as measured by the Mini PAS-ADD. Measures were completed at baseline, 6-month and 12-month time points.


No differences were detected between the intervention and control groups on any of the primary or secondary outcome measures. Simply put, the results indicate that the intervention was ineffective; training CLDT staff in PBS had no impact on challenging behaviour for the participants.


The authors conclude that:

training [CLDT] services staff in PBS, as delivered in this study, was no more effective than TAU in reducing challenging behaviour.

However, a number of limitations described in the paper cast significant doubt over the fidelity of the intervention.

This RCT found no effect from the intervention, but was PBS actually implemented as trained?

This RCT found no effect from the intervention, but PBS was not actually implemented as trained.

Strengths and limitations

This pragmatic trial addresses an important question: Can CLDT professionals who are not behavioural specialists be trained to deliver assessments and intervention plans, using a PBS framework, resulting in significant challenging behaviour reduction? The RCT was well-designed, with a clear CONSORT diagram reporting the recruitment, allocation and follow-up of participants. Evaluation of the quality of completed intervention reports and PBS plans represents a further strength.

Intervention fidelity is the most obvious limitation to this study. The authors report that professionals were trained as specified in the trial protocol. The aim of the training intervention was to teach skills in conducting valid functional assessments, developing function-based behaviour support plans and overseeing their implementation by others. Did this happen in practice?

Functional assessments

Less than a third (31%) of functional assessment reports were completed as trained, with all others missing at least one component of the expected content (e.g. observational data). Functional assessment is a core component of PBS. Incomplete assessment is likely to hamper both the understanding of the context in which challenging behaviour occurs and the development of function-based intervention strategies.

Behaviour support plans

The quality of behaviour support plans completed by professionals was evaluated independently, using the BIP-QEII evaluation tool (Browning Wright D. et al, 2003). All plans received the lowest possible rating (‘weak’). This means that, of the plans made available for review, none were of acceptable quality. A turnover rate of 42% of the paid carers in the intervention arm of the study was noted. Without a clear plan to follow, it is unlikely that a detailed intervention will be sustained over time, especially in the context of significant staff turnover.

In summary, the training that was delivered did not result in the production of intervention reports or good-quality behaviour support plans. Insufficient training may explain this, but the authors also state that the planned 30% caseload reduction did not happen consistently. Professionals who were trained may simply not have had the time to implement this approach.

Implications for practice

In contrast to the UCL PBS study, Hassiotis et al. (2009) previously reported an RCT that investigated the effectiveness of a specialist behaviour support team, predominantly staffed by professionals with diplomas or master’s degrees in applied behaviour analysis. Functional assessment and supported intervention plans implemented by this specialist team proved more effective than TAU in reducing challenging behaviour. Two-year follow-up data indicated that these outcomes were maintained over time. This team had both the training and resources needed to effectively implement PBS.

The conclusion I draw from the UCL PBS trial is that practitioners responsible for completing functional behavioural assessments and writing related PBS plans need to 1) be sufficiently trained and 2) have enough time to complete the process effectively. The study is useful in indicating that existing CLDT professionals may require more detailed training and/or dedicated time to do this work well. In contrast, specialist PBS services (e.g. Toogood S et al, 2015) are already well-equipped to implement this approach.

Future research should address the fidelity issues with this study. Ensuring high-quality delivery of PBS will allow for a fair evaluation of the effectiveness of this approach.

High quality trials can't tell us about effectiveness of PBS until these treatment fidelity issues are solved.

High quality trials can’t tell us about effectiveness of PBS until these treatment fidelity issues are solved.

Conflicts of interest

Dave O’Regan works in the field of Positive Behaviour Support.


Primary paper

Hassiotis APoppe MStrydom A. et al (2018) Clinical outcomes of staff training in positive behaviour support to reduce challenging behaviour in adults with intellectual disability: cluster randomised controlled trial (PDF). Br J Psychiatry 2018 212(3) 161-168.

Other references

Browning Wright, D, Saren, D, Mayer, G (2003) Behavior Intervention Plan Quality Evaluation Scoring Guide II (PDF). Available at http://www.pent.ca.gov.

Gore N, McGill P, Toogood S. et al (2013) Definition and scope for positive behavioural support (PDF). Int J Posit Behav Support 2013 3(2) 14–23.

Hassiotis A, Robotham D, Canagasabey A. et al (2009) Randomized, single-blind, controlled trial of a specialist behaviour therapy team for challenging behaviour in adults with intellectual disability (PDF). Am J Psychiatr 2009 166(11) 1278–1285.

Hassiotis A, Canagasabey A, Robotham D. et al (2011) Applied behaviour analysis and standard treatment in intellectual disability: 2-year outcomes (PDF). Br J Psychiatry 2011 198(6) 490–491.

Hassiotis A, Strydom A, Crawford M. et al (2014) Clinical and cost effectiveness of staff training in Positive Behaviour Support (PBS) for treating challenging behaviour in adults with intellectual disability: a cluster randomised controlled trial (PDF). BMC Psychiatry 2014 14:219.

Local Government Association & NHS England (2014) Ensuring Quality Services. Core Principles for the Commissioning of Services for Children, Young People, Adults and Older People with Learning Disabilities and/or Autism who Display or are at Risk of Displaying Behaviour that Challenges (PDF). London: Local Government Association.

National Institute for Health & Care Excellence (2018) Learning disabilities and behaviour that challenges: service design and delivery (PDF). NICE guideline (NG93).

Toogood S, O’Regan D, Saville M. et al (2015) Providing positive behavioural support services: referral characteristics, resource allocation, case management and overview of outcomes. Int J Posit Behav Support 5(2) 25-32.

Transforming Care & Commissioning Steering Group (2014) Winterbourne View – Time for Change: Transforming the Commissioning of Services for People with Learning Disabilities and/or Autism, Chairman: S Bubb (PDF).

Photo credits


Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+