NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Collaborating Centre for Mental Health (UK). Challenging Behaviour and Learning Disabilities: Prevention and Interventions for People with Learning Disabilities Whose Behaviour Challenges. London: National Institute for Health and Care Excellence (NICE); 2015 May. (NICE Guideline, No. 11.)
Challenging Behaviour and Learning Disabilities: Prevention and Interventions for People with Learning Disabilities Whose Behaviour Challenges.
Show details8.1. Introduction
The assessment of behaviour that challenges is often complex and protracted because assessing the nature of the behaviour alone is rarely, if ever, sufficient to allow for the development of a support and intervention plan. Assessment needs to be able to adequately characterise the behaviour, its antecedents and its consequences, which may require a consideration of a person's developmental history, their mental and physical health, the social and physical quality of their environment, the nature of any care provided and the skills and capacities of those caring for them. It follows from this that the methods of assessment will need to be able to properly and reliably capture important dimensions of all these factors and that a range of assessment methods and skills will need to be available and may be best undertaken in a team context where teams members can draw on the skills and knowledge of each other and those of expert staff when needed. Central to assessment in this area is a consideration of the function of the behaviour, attempts to understand which are central to gaining an understanding of why the behaviour has emerged. However, occasionally assessment can be relatively straightforward; for example, understanding that an increase in aggressive behaviour results from a painful and treatable tooth abscess, which a person with a learning disability was otherwise unable to communicate other than by changing their behaviour.
To be effective, assessment has to more than simply set out an understanding of the function of the behaviour. It has to ensure the most appropriate means to involve service users, families and carers in the process so that not only is the assessment comprehensive and accurate but also that all involved can play an active part in the development of any support and intervention plan. In addition, if an assessment is to be comprehensive it means that skills of particular professionals (for example, GPs, psychiatrists, neurologists, paediatricians, speech and language therapists or psychologists) may be needed. The presence of neurodevelopmental disorders such as autism or attention deficit hyperactivity disorder may complicate assessment (for example, because of communication problems arising from the disorder or associated behavioural problems if the neurodevelopmental disorder is not recognised). As noted above, unrecognised or untreated physical health problems may underlie the problem—sometimes it may be a simple problem such as toothache but it may be a more complex and life-threatening disorder. Both neurodevelopmental and physical disorders can also complicate the identification of emerging mental disorders. Although the link between behaviour that challenges and mental illness is not well understood, new presentations of behaviour that challenges may be a manifestation of a new mental disorder or the relapse of a previously diagnosed one. However, the diagnosis of mental disorder in people with a learning disability poses difficulties resulting from communication problems, the developmental trajectory of a person with a learning disability and the presentation of the symptoms of mental disorders per se given the existing cognitive limitations.
Furthermore, behaviour that challenges may have an adverse impact not only on the person but also on those in caring roles. Therefore, it is acknowledged that the wellbeing of families and carers needs to be assured and an assessment of their ability to cope with the behaviour that challenges of the person they support is paramount. As part of the management of complex needs and behaviour that challenges in the community by secondary care mental health services, the care programme approach (Department of Health, 2008) may be implemented. A formal carer's assessment carried out by social care is part of such a coordinated approach to management.
Before provision of any interventions for behaviour that challenges, it is recognised that an assessment of carers' capacity and resources ought to be made and clear objectives set in order to not only manage expectations but also to monitor the implementation of the support and intervention plan (Ali et al., 2014)
8.2. Review question: In people with a learning disability, what are the key components of, and the most effective structure for, an assessment of the behaviour that challenges across a range of settings?
The review protocol summary, including the review question and the eligibility criteria used for this section of the guideline, can be found in Table 54. A complete list of review questions and review protocols can be found in Appendix F; further information about the search strategy can be found in Appendix H.
Table 54
Clinical review protocol summary for the review of the key components of, and the most effective structure for, an assessment of the behaviour that challenges across a range of settings.
8.2.1. Clinical evidence
No studies assessing the methods and structure of instruments for the assessment of behaviour that challenges displayed by people with a learning disability were identified by the systematic search of the literature undertaken for this guideline.
8.2.2. Clinical evidence statement
No evidence on the methods and structure of instruments for the assessment of behaviour that challenges displayed by people with a learning disability is available.
8.3. Review question: In people with a learning disability and behaviour that challenges, what is the utility of methods and tools for assessment?
The review protocol summary, including the review question and the eligibility criteria used for this section of the guideline, can be found in Table 55. A complete list of review questions and review protocols can be found in Appendix F; further information about the search strategy can be found in Appendix H.
Table 55
Clinical review protocol summary for the review of the utility of methods and tools used to assess behaviour that challenges.
8.3.1. Clinical evidence
The search for evidence (supplemented by GDG advice) identified 57 studies that met the eligibility criteria for this review: Akande 1998 (Akande, 1998), Aman 1985a (Aman et al., 1985a), Aman 1985b (Aman et al., 1985b), Aman 1987a (Aman et al., 1987a), Aman 1987b (Aman et al., 1987b), Aman 1995 (Aman et al., 1995), Aman 1996 (Aman et al., 1996), Barnard-Brak 2013 (Barnard-Brak et al., 2013), Bihm 1991 (Bihm & Poindexter, 1991), Brinkley 2007 (Brinkley et al., 2007), Brown 2002 (Brown et al., 2002), Clarke 2003 (Clarke et al., 2003), Crawford 1992 (Crawford et al., 1992), Dekker 2002 (Dekker et al., 2002), Duker 1998 (Duker & Sigafoos, 1998), Durand 1988 (Durand & Crimmins, 1988), Einfeld 1995 (Einfeld & Tonge, 1995), Emerson 2005 (Emerson, 2005),Gonzalez 2009 (Gonzalez et al., 2009), Haynes 2013 (Haynes et al., 2013), Hill 2008 (Hill et al., 2008), Joosten 2008 (Joosten & Bundy, 2008), Kearney 1994 (Kearney, 1994), Kearney 2006 (Kearney et al., 2006), Koritsas 2013 (Koritsas & Iacono, 2013), Lecavalier 2004 (Lecavalier et al., 2004), Marshburn 1992 (Marshburn & Aman, 1992), Matson 1999b (Matson et al., 1999a), Matson 2007c (Matson & Boisjoli, 2007), Matson 2009 (Matson & Wilkins, 2009), Mohr 2005 (Mohr et al., 2005), Mohr 2011 (Mohr et al., 2011), Newton 1988 (Newton & Sturmey, 1988), Newton 1991 (Newton & Sturmey, 1991), Nicholson 2006 (Nicholson et al., 2006), Norris 2011 (Norris & Lecavalier, 2011), Oliver 2003 (Oliver et al., 2003), Oliver 2007 (Oliver et al., 2007), Paclawskyj 2000 (Paclawskyj et al., 2000), Paclawskyj 2001 (Paclawskyj et al., 2001), Rojahn 2001 (Rojahn et al., 2001), Rojahn 2003 (Rojahn et al., 2003), Rojahn 2010a (Rojahn et al., 2010), Rojahn 2010b (Rojahn et al., 2010b), Rojahn 2012 (Rojahn et al., 2012), Rojahn 2013 (Rojahn et al., 2013), Roy 2002 (Roy et al., 2002), Sansone 2012 (Sansone et al., 2012), Shogren 2003 (Shogren & Rojahn, 2003), Sigafoos 1994 (Sigafoos et al., 1994), Singh 1993 (Singh et al., 1993), Spreat 1996 (Spreat & Connelly, 1996), Thompson 1995 (Thompson & Emerson, 1995), Walsh 1999 (Walsh & Shenouda, 1999), Watkins 2013 (Watkins & Rapp, 2013), Zaja 2011 (Zaja et al., 2011) and Zarcone 1991 (Zarcone et al., 1991).
No studies provided data for the critical outcomes of sensitivity and specificity. Data for reliability and validity were reported for the following assessment instruments:
- Aberrant Behavior Checklist (ABC)
- Behavior Problems Inventory (BPI-01)
- Behavior Problems Inventory – Short Form (BPI-S)
- Challenging Behaviour Interview (CBI)
- Developmental Behaviour Checklist for Parents/Carers (DBC-P)
- Developmental Behaviour Checklist for Adults (DBC-A)
- Functional Analysis Screening Tool (FAST)
- Modified Overt Aggression Scale (MOAS)
- Motivation Assessment Scale (MAS)
- Nisonger Child Behavior Rating Form (NCBRF)
- Questions About Behavioral Function (QABF)
- Strengths and Difficulties Questionnaire (SDQ)
An additional instrument (the Brief Behavioural Assessment Tool) was identified during consultation. Because only preliminary evidence for reliability and validity have been published (Smith & Nethell, 2014), the GDG decided not to include it in this review.
The evidence for each instrument is grouped within the following domains: behaviour that challenges (any), behaviour that challenges (aggression) and functional analysis. Further details about the characteristics and psychometric properties of each instrument can be found in Appendix L.
8.3.1.1. Behaviour that challenges (any)
8.3.1.1.1. Aberrant Behavior Checklist (ABC)
The ABC is a 58-item questionnaire completed by unpaid carers, paid carers or teachers. It was designed as a problem behaviour rating scale to assess treatment effects in people with a learning disability. There are 5 subscales including: Irritability, Lethargy/Social Withdrawal; Stereotypic Behaviour; Hyperactivity/Noncompliance; and Inappropriate Speech.
In a sample of participants with any learning disability the internal consistency of the ABC ranged from good to excellent (Irritability subscale, α = 0.92-0.93; Lethargy/Social Withdrawal subscale, α = 0.90-0.91; Stereotypic Behaviour, α = 0.84-0.90; Hyperactivity, α = 0.93-0.96; and Inappropriate Speech, α = 0.76-0.86 [Aman 1995; Aman 1985b; Marshburn 1992]). Test-retest reliability ranged from moderate to good. In Aman 1987a, inter-rater and test-retest reliability correlations varied markedly across subscales and raters, but were comparable to levels derived with other symptom checklists and were deemed to be adequate.
In a sample of participants with fragile X syndrome, internal consistency ranged from good to excellent (based on modified 6-factor solution: Irritability subscale, α = 0.94; Hyperactivity, α = 0.92; Lethargy/Social Withdrawal, α = 0.86; Stereotypic Behaviour, α = 0.87; Inappropriate Speech, α = 0.80; and a newly derived factor, social avoidance, α = 0.92 [Sansone 2012]).
The 5-factor solution of the ABC has been replicated with learning disability and autism samples (Aman 1987b; Aman 1995; Bihm 1991; Brinkley 2007; Newton 1988). Brown 2002 and Marshburn 1992 found a 4-factor solution to be most appropriate with a learning disability sample, as the inappropriate speech factor was not replicated. Moderate to excellent congruence has been found between the original ABC factor structure and that found with learning disability samples (0.62-0.97) (Aman 1987b; Aman 1995; Brown 2002; Marshburn 1992). Good convergent and divergent validity has been demonstrated by significant relationships between the ABC and the HoNOS-LD, VABS II, Reiss Screen, CBI, DASH-II and ABS (Aman 1985b; Hill 2008; Oliver 2003; Paclawskyj 1997; Rojahn 2003; Roy 2002; Walsh 1999).
A 6-factor solution, which adds a ‘social avoidance’ factor to the original ABC factors has been found in a sample of participants with fragile X syndrome (Sansone 2012).
8.3.1.1.2. Behavior Problems Inventory (BPI-01)
The BPI-01 is a 52-item respondent-based behaviour rating instrument. It is suitable for both children and adults with a learning disability and completed by unpaid carers, paid carers or teachers. It reports the frequency and severity of behaviour on 3 subscales: Self-Injurious Behavior; Stereotyped Behavior; and Aggressive/Destructive Behavior.
In Rojahn 2010b, the BPI-01 showed good reliability between teacher informants, but it was poor between parent and teacher informants. Gonzalez 2009 found that the inter-rater and re-test reliability coefficients of the Self-Injurious Behavior items and subscale were generally good, whereas the overall inter-rater and test–retest reliability coefficients of the Aggressive/Destructive Behavior items and subscale were good to excellent. The Stereotyped Behavior items and subscale had fair to low inter-rater and test-retest reliability coefficients (Gonzalez 2009). Internal consistency values range from poor to acceptable for the Self-Injurious Behavior subscale, poor to excellent for the Stereotyped Behavior items and acceptable to good for Aggressive/Destructive Behavior (Gonzalez 2009; Rojahn 2001; Rojahn 2010b; Rojahn 2012b). Good convergent and divergent validity has been demonstrated by significant correlations in predicted directions between the BPI-01 and measures including the ABC, NCBRF, Inventory for Client and Agency Planning, Autism Spectrum Disorders-Behaviour Problems for Intellectually Disabled Adults and DASH-II (Hill 2008; Rojahn 2003; Rojahn 2010a; Rojahn 2010b; Rojahn 2012b).There have been mixed findings regarding structural validity. Rojahn 2001 and Gonzalez 2009 replicated a 3-factor solution and Hill 2008 found a 6-factor solution that mapped onto the 3-subscale structure. However, Rojahn 2010b failed to replicate a 3-factor solution. Barnard-Brak 2013 used confirmatory factor analysis to indicate acceptable model fit for each latent construct suggesting support for the one-dimensional nature of each trait. Individuals with a diagnosis of PDD had higher scores on the Self-Injurious Behavior and Stereotyped Behavior subscales than those without; in addition, they also had elevated aggression/destruction scores. Higher stereotyped behaviour scores among people with a diagnosis of stereotyped behaviour disorder, compared with those without, can be considered as another sign of validity of the BPI-01.
Rojahn 2013 included a sample of participants with Cornelia de Lange's syndrome only. In this study internal consistency values ranged from questionable to excellent (α = 0.66-0.90) and there was evidence of a sufficient factor structure for each of the subscales identified by the BPI-01.
8.3.1.1.3. Behavior Problems Inventory – Short Form (BPI-S)
The BPI-S is a shortened 30-item version of the BPI-01 completed by unpaid carers, paid carers or teachers. It is used for children and adults with a learning disability and contains the same 3 subscales as the BPI-01: Self-Injurious Behavior; Stereotyped Behavior; and Aggressive/Destructive Behavior.
Internal consistency was found to be acceptable for the Aggressive/Destructive and Stereotyped Behavior subscales of the BPI-S. For the Self-Injurious Behaviour subscale, values ranged from unacceptable to acceptable (Rojahn 2012b). Confirmatory factor analysis results indicated an acceptable model fit for each latent construct suggesting support for the one-dimensional nature of each trait (Barnard-Brak 2013). Good convergent and divergent validity has been demonstrated by significant correlations in predicted directions between the BPI and measures including the ABC, NCBRF, Inventory for Client and Agency Planning and DASH-II (Rojahn 2012b).
8.3.1.1.4. Challenging Behaviour Interview (CBI)
The CBI is a 19-item instrument completed by paid carers or teachers, which measures the severity of behaviour that challenges in children and adults with a learning disability. It is divided into 2 parts. Part I identifies the occurrence of 5 clearly operationalised forms of behaviour that challenges that have occurred in the previous month. Part II assesses the severity of the behaviours identified on 14 scales measuring the frequency and duration of episodes, effects on the person with a learning disability and others and the management strategies used by carers.
The CBI has been found to demonstrate good inter-rater reliability (kappa = 0.50-0.80) and test-retest reliability (kappa = 0.70-0.91). The CBI has also been found to be significantly correlated with the ABC showing good convergent validity (Oliver 2003).
8.3.1.1.5. Developmental Behaviour Checklist for Adults (DBC-A)
The DBC-A is a 107-item instrument completed by unpaid or paid carers. It assesses a comprehensive range of emotional, behavioural and mental health problems in adults with mild, moderate and more severe levels of learning disability. The manual and supplement cost £64.92 and a pack of 10 checklists cost £5.90.
The DBC-A has shown substantial agreement between family members (ICC = 0.72; Mohr 2005) and acceptable agreement between paid carers (ICC 0.69; Mohr 2011). Test-retest reliability has been found to be good, ranging from 0.75-0.85 (ICC; Mohr 2005). A strong positive correlation has been demonstrated between the DBC-A and both the PAS-ADD and ABC, providing evidence of good convergent validity (Mohr 2005).
8.3.1.1.6. Developmental Behaviour Checklist for Parents/Carers (DBC-P)
The DBC-P is a 96-item instrument for the assessment of behavioural and emotional problems of children and young people with developmental and learning disabilities completed by parents or unpaid and paid carers. It takes 10 to 15 minutes to administer. The starter kit, which consists of a manual and a packet of checklists and score sheets, costs £77.46.
Internal consistency has been found to be questionable for the Antisocial subscale (α = 0.67) and acceptable to excellent for the remaining subscales (α = 0.73-0.91) based on the original 6-factor solution (Einfeld 1995). Internal consistencies for a revised 5-factor solution have been found to range from questionable for the Anxiety subscale (α = 0.66) to excellent for the Disruptive/Antisocial and Self-absorbed subscales (α = 0.91) (Dekker 2002). Inter-rater reliability for parent ratings was moderate to substantial (ICC = 0.75-0.80) and poor to substantial for teacher ratings (ICC = 0.30 – antisocial subscale; ICC = 0.74 – self-absorbed subscale) (Einfeld 1995). Test-retest reliability was found to be moderate to substantial (ICC = 0.75-0.80) (Einfeld, 1995).
Post-treatment change as measured by the DBC has been found to be strongly correlated with change as rated by an experienced clinician (Clarke 2003). Einfeld 1995 produced 6 clinically meaningful and factorially valid subscales using principle components analysis: Disruptive, Self-absorbed, Communication disturbance, Anxiety, Social relating and Antisocial. However, Dekker 2002 suggested that a 5-factor solution was more appropriate, which included the following subscales: Disruptive/Antisocial, Self-absorbed, Communication Disturbance, Anxiety and Social relating. Dekker 2002 suggested that this revised scale structure constitutes an improvement over the original structure given that it is based on a larger sample and one that better represents all levels of learning disability. Strong positive correlations have been found between the DBC and the Adaptive Behavior Scale (0.72) and the Scales of Independent Behaviour (0.72 p < .001 in each case). Pearson product-moment correlations between the DBC total score and psychiatrist ratings has been found to be significant (0.81, p < .001) (Einfeld 1995).
8.3.1.1.7. Nisonger Child Behavior Rating Form (NCBRF)
The NCBRF is a standardised instrument for assessing child and adolescent behaviour completed by families, carers or teachers. It has 76 items and a scoring time of 8 minutes. The instrument is available for free.
Poor inter-rater reliability for the NCBRF Prosocial subscales has been found between teacher and parent-teacher ratings. For the Problem Behavior subscales teacher-teacher agreement was fair, but parent-teacher agreement ranged from poor to moderate (Aman 1996; Rojahn 2010b). Rojahn 2010b found fair reliability for the Prosocial and Problem behavior subscales. Internal consistency has been found to be fair to good for the Prosocial subscales and good for the Problem Behavior subscales, based on a learning disabilities sample (Aman 1996; Norris 2011; Rojahn 2010b). Based on a sample of participants with autism, Lecavalier 2004 found questionable to good consistency for the Adaptive Social subscale (α = 0.63-0.79) and acceptable to good consistency for the Compliant/Calm subscale (α = 0.79) based on parent and teacher ratings, respectively. Studies indicated strong convergent and divergent validity between the NCBRF and BPI-01, ABC and DBC (Aman 1996; Norris 2011; Rojahn 2010b). There have been mixed findings regarding the factor structure of the NCBRF. Lecavalier 2004 and Norris 2011 replicated a 2-factor structure for social competence items based on autism and learning disabilities samples. But Rojahn 2010b found the fit for a 2-factor solution to be poor. Lecavalier 2004 found a 5-factor solution to be more appropriate than the original 6-factor solution for problem behaviour items. Other studies have demonstrated poor fit for both 5- and 6-factor solutions for this scale (Norris 2011; Rojahn 2010b).
8.3.1.1.8. Strengths and Difficulties Questionnaire (SDQ)
The SDQ is one of the most widely used brief questionnaires for assessing mental health problems in children and young people. It has 25 items and is divided into 5 subscales: Emotional Symptoms; Conduct Problems; Hyperactivity; Peer Problems; and Pro-Social Behaviour. It can be self-completed or administered by families, carers and teachers, and is available for free.
The SDQ has been found to show acceptable internal consistency overall (α = 0.71), ranging from unacceptable (α = 0.30 for the Peer Problems subscale) to good (α = 0.87 for total impact) (Emerson 2005). Inter-rater reliability has been found to be modest for child ratings when compared with parent and teacher ratings (0.11 for the Peer Problems subscale – 0.49 for Hyperactivity) (Emerson 2005). Self-reported difficulties have been found to be significantly correlated with ICD-10 diagnoses (Emerson 2005). In a population of children with a learning disability, Haynes 2013 found that a 3-factor model was a better measure than the original 5-factor model.
8.3.1.2. Behaviour that challenges (aggression)
8.3.1.2.1. Modified Overt Aggression Scale (MOAS)
The MOAS is designed to measure aggressive behaviours in adults and children. It is a 20-item instrument divided into 5 subscales: Verbal Aggression Towards Others; Verbal Aggression Towards Self; Physical Aggression Against Objects; Physical Aggression Against Self; and Physical Aggression Against Others. The MOAS differs from the original Overt Aggression Scale by modifications to wording and the addition of items measuring verbal aggression toward self. It is completed by unpaid or paid carers and is available for free.
The MOAS has been found to have a high level of agreement between raters for Verbal Aggression Towards Others and Verbal Aggression Towards Self (ICC = 0.90), Physical Aggression Against Others (ICC = 0.90) and for total MOAS score (ICC = 0.93). Levels of agreement on the other 2 subscales have been found to be lower but still in the moderate range (ICC = 0.49-0.56) (Oliver 2007). There were no data available for the validity of the measure.
8.3.1.3. Functional assessment
8.3.1.3.1. Functional Analysis Screening Tool (FAST)
The FAST is a functional assessment tool designed to assess 4 functional properties of problem behaviour in adults with a learning disability. The 4 subscales are called: Social (Attention/Preferred Items), Social (Escape from Tasks/Activities), Automatic (Sensory Stimulation) and Automatic (Pain Attenuation). It has 16 items and is completed by a paid carer, family carer or teacher. It takes approximately 10 minutes to score and is available for free.
The FAST has been found to have unacceptably low internal consistency (α = 0.05-0.77 for each subscale with a mean of 0.39) especially for the Social Attention and Social Escape subscales (Zaja 2011). Correlations for inter-rater agreement have been found to range from poor to good (ICC = 0.48−0.71) (Zaja 2011). Test-retest correlation coefficients have been found to range from fair to excellent for total FAST scores (0.55-0.82) (Zaja 2011). Convergent and discriminant validity (Spearman p) has been found to be better between the Functional Assessment for Multiple Causality and the QABF (0.80) than between the FAST and the Functional Assessment for Multiple Causality (0.50) or the FAST and the QABF (0.51) (Zaja 2011).
8.3.1.3.2. Motivation Assessment Scale (MAS)
The MAS is a 16-item instrument completed by unpaid and paid carers or teachers. It is designed to provide information about the function of the target behaviour of children and adults with a learning disability. Each item refers to one of 4 potential functions, with each item rated on a 7-point Likert scale. The MAS is supposed to reveal whether the target behaviour is related to sensory, escape, attention or tangible variables. The instrument takes approximately 10 minutes to score and is free.
Internal consistency has been found to range from questionable to good for the sensory items (α = 0.67-0.83), questionable to good for escape (α = 0.68-0.88), questionable to excellent for attention (α = 0.69-0.96) and good to excellent for tangible items (α = 0.80-0.91) (Bihm 1991; Duker 1998; Koritsas 2013; Newton 1991; Shogren 2003; Spreat 1996). There have been mixed findings concerning inter-rater reliability with levels of agreement ranging from poor to almost perfect. However, the majority of studies report poor agreement (Akande 1998; Crawford 1992; Duker 1998; Durand 1988; Kearney 1994; Koritsas 2013; Newton 1991; Shogren 2003; Sigafoos 1994; Spreat 1996; Thompson 1995; Zarcone 1991). The MAS correlates with functionally analogous scales of the QABF, offering evidence of convergent validity (Koritsas 2013; Paclawskyj 2001; Shogren 2003). There have been mixed findings about the factor structure of the MAS. Several studies have failed to replicate the original factor structure of the MAS (Duker 1998; Kearney 2006; Joosten 2008; Koritsas 2013) and others have offered support for the structure in institutional but not school samples (Bihm 1991; Singh 1993). Durand 1988 found that teachers' ratings on the MAS predicted their students' behaviour in experimental conditions.
8.3.1.3.3. Questions About Behavioral Function (QABF)
The QABF is a 25-item report completed by unpaid and paid carers. It is designed to identify behavioural functions that are important in maintaining aberrant behaviour in children and adults. The 5 subscales of the assessment relate to 5 possible variables influencing problem behaviour: Attention; Escape from Task Demands or Social Contact; Non-social Reinforcement; Physical Discomfort; and Tangible Reinforcement. The instrument is available for free.
Internal consistency has been found to be generally acceptable to excellent for all subscales (Koritsas 2013; Nicholson 2006; Paclawskyj 2000; Shogren 2003; Zaja 2011), although Paclawskyj 2000 found that it was questionable for the test as a whole (α = 0.60). Inter-rater reliability for subscales has been found to range from poor to almost perfect (kappa = 0.21-0.95) (Koritsas 2013; Matson 2007c; Matson 2009; Nicholson 2006; Paclawskyj 2000; Shogren 2003; Zaja 2011). Scores have been found to be stable over time indicating good test-retest reliability (Paclawskyj 2000; Zaja 2011). The Motivation Assessment Scale (MAS) and Functional Assessment for Multiple Causality have been found to correlate with functionally analogous scales of the QABF, offering evidence of convergent validity (Koritsas 2013; Paclawskyj 2001; Shogren 2003; Zaja 2011). Watkins 2013 also demonstrated that the QABF identified the same behavioural functions in participants when compared with a brief functional analysis. Participants with treatments developed from functional assessment (QABF results) have been found to improve significantly when compared with controls receiving standard treatments not based on functional analysis (Matson 1999b). Paclawskyj 2000 replicated the original 5-factor solution. Nicholson 2006 also found 5 factors that corresponded to the 5 subscales of the QABF, however the analysis suggested the existence of a 6th factor with a high loading from only a single item, concerning the repetitive nature of the behaviour. The proposed explanation for this was that respondents differentiated repetitiveness of behaviour from aspects suggesting sensory or other automatic reinforcement.
8.3.2. Health economic evidence
No studies assessing the cost effectiveness of methods and tools for the assessment of behaviour that challenges displayed by people with a learning disability were identified by the systematic search of the literature undertaken for this guideline. Details on the methods used for the systematic search of the economic literature are described in Chapter 3.
8.3.3. Clinical evidence statements
- For the ABC instrument, there was evidence from 16 studies demonstrating adequate reliability and validity, although evidence for inter-rater and criterion validity were not available.
- For the BPI-01 instrument, there was evidence from 8 studies demonstrating adequate reliability and validity, although evidence for criterion validity was not available.
- For the BPI-S, there was evidence from 2 studies demonstrating adequate internal consistency and validity, although evidence for inter-rater reliability, test-retest reliability and criterion validity was not available.
- For the CBI, there was evidence from 1 study demonstrating adequate reliability and validity, although evidence for internal consistency and criterion validity was not available.
- For the DBC-A there was evidence from 2 studies demonstrating adequate reliability and validity, although evidence for internal consistency and criterion validity was not available.
- For the DBC-P there was evidence from 3 studies demonstrating adequate reliability and validity.
- For the NCBRF there was evidence from 4 studies demonstrating adequate test-rest reliability, internal consistency and convergent validity, however inter-rater reliability was poor, structural validity was unclear and criterion validity was not available.
- For the SDQ there was evidence from 1 study demonstrating adequate internal consistency and criterion validity, however inter-rater reliability was poor and test-retest reliability and structural validity were not available.
- For the MOAS there was evidence from 1 study indicating adequate reliability, although evidence for test-retest reliability, internal consistency and validity was not available.
- For the FAST there was evidence from 1 study demonstrating adequate reliability, however internal consistency was poor and construct validity was mixed. Criterion validity was not available.
- For the MAS there was evidence from 17 studies demonstrating adequate internal consistency and convergent validity, however test-rest reliability was mixed and there was no evidence for inter-rater reliability and criterion validity.
- For the QABF there was evidence from 10 studies demonstrating adequate reliability and construct validity, however inter-rater reliability was mixed and criterion validity was not available.
8.3.4. Economic evidence statements
No evidence on the cost effectiveness of methods and tools for the assessment of behaviour that challenges displayed by people with a learning disability is available.
8.3.5. Recommendations and link to evidence
The recommendations that were developed from this section and the link to the evidence are at the end of the chapter (see Section 8.5). The GDG considered the review of the utility of methods and tools used to assess behaviour that challenges alongside the reviews of other instruments because they saw the benefit of developing an integrated approach to assessment.
8.4. Review question: In carers of people with a learning disability and behaviour that challenges, what is the utility of methods used to assess and monitor their capacity to support the person?
The review protocol summary, including the review question and the eligibility criteria used for this section of the guideline, can be found in Table 56. A complete list of review questions and review protocols can be found in Appendix F; further information about the search strategy can be found in Appendix H.
Table 56
Clinical review protocol summary for the review of the utility of methods used to assess and monitor carers' capacity to support the person.
8.4.1. Clinical evidence
The search for evidence (supplemented by GDG advice) identified 8 studies that met the eligibility criteria for this review: Chao 2011 (Chao et al., 2011), Friedrich 1983 (Friedrich et al., 1983), Hastings 2004 (Hastings et al., 2004), Hatton 1995a (Hatton et al., 1995), Hatton 1995b (Hatton & Emerson, 1995), Honey 2005 (Honey et al., 2005), Knussen 1992 (Knussen et al., 1992) and Scott 1989 (Scott et al., 1989).
No studies provided data for the critical outcomes of sensitivity and specificity. Data for reliability and validity were reported for the following assessment instruments:
- Maslach Burnout Inventory (MBI)
- Shortened Ways of Coping (Revised) Questionnaire (SWC-R)
- Ways of Coping Questionnaire – Revised (WC-R)
- Questionnaire on Resources and Stress – Friedrich edition (QRS-F).
The evidence is organised by instrument and grouped within the following domains: carer burnout, carer needs and carer stress. Further details about the characteristics and psychometric properties of each instrument can be found in Appendix L.
8.4.1.1. Carer burnout
8.4.1.1.1. Maslach Burnout Inventory (MBI)
The MBI is a self-report instrument with 22 items developed to assess burnout in professional paid carers. The licence to conduct 50 and 500 paper and pencil administrations costs £59.59 and £214.51, respectively. The licence to use the online version for 50 and 500 administrations costs £71.50 and 257.42. respectively. The manual for the MBI costs £23.83.
The MBI has been found to have acceptable to good internal consistency for the Emotional Exhaustion subscale (α = 0.87-0.90) and the Personal Accomplishment subscale (α = 0.76). Internal consistency for the Depersonalisation subscale has varied from unacceptable to acceptable (α = 0.68-0.71) (Chao 2011, Hastings 2004).
Chao 2011 found that while a 3-factor solution suggested an acceptable fit for the data, a 4-factor solution provided a better fit than the original 3-factor solution. Items on the 3 subscales all had positive loadings greater than 0.40 on the anticipated factors. Of the 22 items, 19 loaded above 0.40 on the appropriate factor and less than 0.40 on the other factors.
8.4.1.2. Carer needs
8.4.1.2.1. Shortened Ways of Coping Questionnaire – Revised (SWC-R)
The SWC-R a 14-item self-report questionnaire for adults to represent thoughts and actions used to deal with the demands of a stressful encounter. The measure is scored on 2 subscales which represent distinct ways of coping: Practical Coping and Wishful Thinking.
Internal consistency for the SWC-R has been found to range from poor to good for the Wishful Thinking subscale (α = 0.52-0.82), and acceptable to good for the Practical Coping subscale (α = 0.70 − 0.80) (Hatton 1995b). Subscale scores were stable over time demonstrating good test-retest reliability: paired t-tests showing no significant differences between measurements over a 16-month period (Hatton 1995b).
A significant association has been found between 1991 Wishful Thinking scores and 1993 distress scores (Hatton 1995b).
8.4.1.2.2. Ways of Coping Questionnaire – Revised (WC-R)
The WC-R is a full length version of the SWC-R. It has 66 items and takes approximately 10 minutes to complete. As in the SWC-R, it is used to represent thoughts and actions that can be used to deal with the demands of a stressful encounter. The licence to conduct 50 and 500 paper and pencil administrations costs £59.59 and £214.51, respectively. The licence to use the online version for 50 and 500 administrations costs £71.50 and £257.42, respectively. The WC-R manual costs £23.83.
In a study that included participants with Down's syndrome only, internal consistency was found to be poor for the Passive Acceptance subscale (α = 0.53), questionable for the Stoicism subscale (α = 0.65), and acceptable for the Practical Coping, Wishful Thinking and Seeking Social Support subscales (α = 0.77–0.90) (Knussen 1992). In Hatton 1995a, 4 out of 5 subscales showed adequate levels of test-retest reliability for mothers (α > 0.6), with only the Passive Acceptance subscale failing to reach an adequate level. For fathers, all except the Stoicism subscale showed adequate levels.
In a study that included participants with Down's syndrome only, subscales resulting from factor analysis were found to be similar to those reported in earlier studies, with differences attributable to variations of personal and situational variables (Knussen 1992).
8.4.1.3. Carer stress
8.4.1.3.1. Questionnaire on Resources and Stress (QRS-F)
The QRS-F is a 52-item self-report questionnaire for families and carers, used widely with parents of children with disabilities. It assesses 4 subcomponents of parental perceptions: parent and family problems (stressful aspects of the impact of the child with disability on parents and the wider family), pessimism (parents' pessimistic beliefs about the child's future), child characteristics (features of the child that are associated with increased demands on parents), and physical incapacity (the extent to which the child is able to perform a range of typical activities). The QRS-F is a free instrument.
The 52-item version of the QRS-F has been found to have excellent internal consistency (Kuder-Richardson [KD] coefficient = 0.89-0.93) (Friedrich 1983, Scott 1989). In Honey 2005, a good level of internal consistency has been found for mothers (KD-20 = 0.85) and for both mothers and fathers (KD-20 = 0.93) of young children with autism, using a 31-item version of the QRS-F derived from factor analysis. Honey 2005 also found no significant difference between mothers' (mean = 10.67, standard deviation [SD] = 7.08) and fathers' (mean = 9.91, SD = 5.95) scores (t[42] = 1.34, p = 0.19), suggesting good inter-rater reliability with the 31-item version.
The QRS-F shows significant correlations in the expected direction with the Beck Depression Inventory and Marlowe-Crowne Social Desirability Scale, suggesting good convergent validity (Friedrich 1983). Scott 1989 successfully replicated the 4-factor solution found by Friedrich 1983. Scores have been found to vary reliably depending on the child's type of learning disability, which supports criterion validity (Scott 1989).
In a sample of participants with autism only, Honey 2005 did not find a 2- or 3-factor structure that had any resemblance to the existing QRS-F scales. Rather, the majority of the items loaded significantly onto the first factor extracted in most analyses. Adaptation (Judson scale) has been found to be significantly correlated with maternal stress (r[54] = −0.70, p <0.001) and paternal stress (r[43] = −0.46,p < 0.01), offering evidence of convergent validity (Honey 2005).
8.4.2. Health economic evidence
No studies assessing the cost effectiveness of methods used to assess and monitor the capacity of carers to support a person with a learning disability and behaviour that challenges were identified by the systematic search of the literature undertaken for this guideline. Details on the methods used for the systematic search of the economic literature are described in Chapter 3.
8.4.3. Clinical evidence statements
- For the MBI there was evidence from 2 studies demonstrating adequate internal consistency and construct validity, however there was no evidence for criterion validity, inter-rater and test-retest reliability.
- For the SWC-R there was evidence from 1 study demonstrating adequate reliability and criterion validity, however there was no evidence for inter-rater reliability and construct validity.
- For the WC-R there was evidence from 2 studies demonstrating adequate structural validity, however reliability varied and there was no available evidence for inter-rater reliability and criterion validity.
- For the QRS-F there was evidence from 3 studies demonstrating good reliability and construct validity, although there was no evidence for test-retest reliability and criterion validity.
8.4.4. Economic evidence statements
No evidence on the cost effectiveness of methods used to assess and monitor the capacity of carers to support a person with a learning disability and behaviour that challenges is available.
8.5. Recommendations and link to evidence
8.5.1. The assessment process
| Recommendations |
|
|---|
8.5.2. Initial assessment of behaviour that challenges
| Recommendations |
|
|---|
8.5.3. Risk assessment
| Recommendations |
|
|---|
8.5.4. Further assessment of behaviour that challenges
| Recommendations |
|
|---|
8.5.5. Functional assessment of behaviour
| Recommendations |
|
|---|
8.5.6. After further assessment
| Recommendations |
|
|---|
8.5.7. Behaviour support plan
| Recommendations |
|
|---|
8.5.8. Interventions for coexisting health problems
| Recommendations |
|
|---|
8.5.9. Link to evidence across all topics
| Relative values of different outcomes | The GDG decided that clinical utility (including the key components of assessment, sensitivity and specificity, reliability and reliability) was the critical outcome. |
|---|---|
| Trade-off between clinical benefits and harms | The GDG decided to adopt a graduated approach to assessment. This was because, in their expert opinion and experience, in a number of circumstances only limited assessment was necessary. The GDG recognised that while this is less intrusive and less consuming of resources, it does increase the risk that more complex factors contributing to the behavioural problem may not be identified. |
| Trade-off between net health benefits and resource use | Effective assessment and monitoring of carers' capacity in supporting people with a learning disability and behaviour that challenges has important clinical and resource implications for the carers, in terms of intervention costs and the carers' coping and HRQoL; it has also important clinical and resource implications for people with a learning disability, as it enables carers to assess and monitor them most effectively, which, in turn, contributes to the effective and cost-effective anticipation and management of behaviour that challenges. It is therefore likely that costs of assessment and monitoring may be offset, at least partially, by savings associated with earlier and more effective management of behaviour that challenges. |
| Quality of evidence | There was very limited evidence on the structure and content of assessment. There was moderate to low-quality evidence on the psychometric properties of a number of measures reviewed. |
| Other considerations | In the absence of evidence on the structure, content and validity of the assessment process, the GDG used informal consensus methods to arrive at the recommendations related to this topic in this chapter. The GDG also drew on the evidence in the chapter on experience of care (which provided evidence of service users' and carers' experience of the assessment process) and the chapter on psychosocial interventions, which identified functional assessment as a moderator of treatment effectiveness. The GDG decided first that a phased approach to assessment was needed to balance the burden of assessment with the need to understand the drivers behind any behavioural problem. They judged that this should start with an initial assessment, including a risk assessment, followed by further assessment if the behaviour is severe or complex, or has not responded to the behaviour support plan. To ensure that the assessment is fully informed and that any plan that emerged has full service user and carer involvement, the GDG judged that both service users and carers should be fully involved in all stages of the assessment. The evidence drawn from the chapter on psychosocial interventions that functional assessment is an important moderator of a good outcome led the GDG to recommend this as an integral part of a further assessment. Formal rating scales (for which there was evidence for their reliability and validity, including behaviour that challenges, mental state and quality of life) were also considered to be of use in informing the assessment and providing reliable data on the impact of any interventions. The GDG was aware that any assessment or intervention that focused on behaviour that challenges could increase risk and so recommended that a risk assessment be an integral part of any assessment. The GDG also bore in mind the reactive nature of many interventions for behaviour that challenges and decided that wherever possible all interventions should be contained within a behavioural support plan, which emphasises proactive as well as reactive strategies. Finally, where the assessment indicated a coexisting mental or physical health problem, the GDG agreed that it would be good practice to offer an appropriate intervention in line with relevant NICE guidance, but the nature, content and delivery should be adjusted to take account of the severity and impact of the person's learning disability and behaviour that challenges. |
- Introduction
- Review question: In people with a learning disability, what are the key components of, and the most effective structure for, an assessment of the behaviour that challenges across a range of settings?
- Review question: In people with a learning disability and behaviour that challenges, what is the utility of methods and tools for assessment?
- Review question: In carers of people with a learning disability and behaviour that challenges, what is the utility of methods used to assess and monitor their capacity to support the person?
- Recommendations and link to evidence
- Assessment - Challenging Behaviour and Learning DisabilitiesAssessment - Challenging Behaviour and Learning Disabilities
Your browsing activity is empty.
Activity recording is turned off.
See more...