The Perception of Positive Behavioral Support

Positive Behavioral Support (PBS) is a framework for developing via functional analysis an understanding of a person’s challenging behaviour, which is then used to develop effective support NHS England and Local Government Association [1]. PBS is a multi-component framework employing ethical, socially valid techniques in management of challenging behaviour and has primarily been implemented within intellectual disability [ID] services Carr [2] and schools Sugai et al. [3] as well as in other health and social care settings Allen [4] and is widely endorsed MIND, 2013, 2015; Royal College of Nursing, 2013; NHS Protect, 2014; Department of Health, 2014; National Institute for Health and Care Excellence, 2015a and b; Skills for Care and Skills for Health, 2014.


Introduction
Positive Behavioral Support (PBS) is a framework for developing via functional analysis an understanding of a person's challenging behaviour, which is then used to develop effective support NHS England and Local Government Association [1]. PBS is a multi-component framework employing ethical, socially valid techniques in management of challenging behaviour and has primarily been implemented within intellectual disability [ID] services Carr [2] and schools Sugai et al. [3] as well as in other health and social care settings Allen [4]  Any assessment or intervention implemented within the PBS framework should ideally be informed and valued via the active participation of multiple stakeholders, including the focus individual. Consideration is given by stakeholders of appropriateness of PBS interventions and their 'goodness of fit' with the focus individual and the systems supporting them Carr et al. [2]; Gore et al. [5]. Hence, PBS explicitly promotes 'an egalitarian approach towards stakeholder participation which has become a normative feature of PBS' Carr et al. [2]. The current review is aimed at understanding stakeholder's perceptions of PBS. In order to identify and determine the extent and quality of research in this domain, a systematic review and meta-synthesis of relevant qualitative research studies was conducted with a focus on studies that have reported on the views and experiences people involved in PBS across a range of organisational settings (e.g. ID services, education, juvenile justice).

Search strategy
The formal search strategy utilised two groups of search terms, namely Positive Behav* Support AND Experience* OR view* OR perception* OR perspective* OR attitude* OR opinion* OR account* OR understanding OR interpret* OR outlook* OR descri* or expla* OR qualitative OR grounded theory OR «interpretative phenomenological analysis» OR IPA OR thematic analysis. Inclusion criteria for publication included publication after 1990 in English and use of either a qualitative or mixed methodology. Exclusion criteria were duplicate articles, publication in non-peer review journals, Studies that have an exclusively quantitative methodology, studies with no primary data (e.g. direct quotations from participants), studies investigating non-organizational/non-professionalised settings (e.g. families) and all forms of 'grey literature' (e.g. review articles, commentaries, discussion pieces). Similarly, studies that did not explicitly report upon views or experiences of PBS were also excluded. The complete process is shown in (Figure 1). This process, as shown, resulted in eleven papers being identified for inclusion in the review and meta-synthesis ( Figure 1).

Quality assessment
All eleven articles employed a qualitative methodology and are summarized in Appendix A. The framework developed by Cardiff University's Support Unit for Research Evidence (SURE) [7] was used to assess the methodological quality of each study, incorporating as it does the sub-fields of CASP [8] qualitative checklist with the addition of an extra field: consideration of issues relating to potential author sponsorship / conflicts of interest. The SURE was used to assign each study a score of 0,1or 2 for each sub-field: a) 0 -This score indicates there to be no consideration given to the question posed by the quality review framework. b) 1 -This score indicates there to be partial consideration given to the question posed by the quality framework or that issues were addressed however limitations were present. c) 2 -This score indicates the article clearly addressed the question posed by the quality review framework in a clear and rigorous fashion.
d) The scores allocated to each study together with a total quality score from 0-80 are shown in Appendix B.

Synthesising Reviewed Studies
A meta-ethnographic approach Noblit and Hare [9] was employed as the basis of the meta-synthesis Ring, Ritchie, Mandava, and Jepson [10].  [11] notion of 'first-order', 'second-order' and 'third-order' constructs. First order constructs reflect participants' understandings, as reported in the original studies (e.g. direct quotations), second order constructs reflect the authors' interpretation of the participants' understanding and third order constructs reflect the subsequent interpretation of the original authors' interpretation. Typically, the 'data' or 'building blocks' of the meta-ethnographic approach are the second-order constructs within the original studies Britten et al. [11], Toye et al. [12] and this was the focus of the current synthesis.

Results
The quality of the eleven studies in the current review was variable with total quality scores ranging from 50 to 72 (62.5% to 90% quality). It was considered that studies scoring over 80% were of 'high' quality, studies scoring between 70%-80% were 'mediumhigh' and studies scoring between 60%-70% were of 'medium' quality. (Figure 2) below shows the quality score for each study rounded to the nearest whole percentage number ( Figure 2). As shown, most studies (n=6) are of medium-high quality (70%-80%), two studies are of medium quality (60%-70%) and three studies are of high quality (>80%). Therefore, all eleven studies were taken forward to the next stage of the review.
In all studies, data collection methods, analyses and interpretative procedures were well described and triangulation was employed in cases to improve the validity of findings. However, most of the studies lacked a rationale for the specific methodology used, did not examine the relationship between researcher(s) and participants, did not report on data saturation and there was a general absence of information relating to research governance, for example, how the research was presented to participants or how ethical approval was obtained.
Two of the three high quality studies explored views of focus individuals who the PBS approach was implemented for. Both Davies, Mallows and Hoare [6] and Inchley-Mort and Hassiotis [13] reported themes that described the importance placed on relationships of 'understanding' within PBS, they identified similar barriers to implementation and commonly concluded that PBS is perceived to be 'valued' and 'acceptable' to most focus individuals, also pointing out that future research needs to further explore the experience of focus individuals. The other high quality study, Andreou, McIntosh, Ross and Kahn [14] reported on the importance of contextual adaptations in order to sustain PBS, in particular the fostering of environments described as 'flexible', 'creative' and containing 'foresight'.
Those studies considered to be medium-high quality were rated so due to a lack of consideration of researcher position, data-saturation and wider ethical considerations. They reported on barriers to implementing PBS being consistent with broader personal and organisational implementation patterns that need to be overcome Lohrmann [15], Lohrmann [16], systemic and resource issues (including culture, support, use of time and focus-individual involvement) that can impede or facilitate implementation. It was identified that PBS plans need to be contextually relevant, person-centred and based on available resources Hieneman and Dunlap [17]. The goals and outcomes of PBS need to be supported by key stakeholders, however the procedures required to implement PBS received less support from a social validity perspective Frey, Lee Park, Browne-Ferrigno, and Korfhage [18]. Team members' perceived the social processes of PBS to be most important and that future research needs to more fully understand the social contexts in which PBS is implemented Bambara, et al. [19].
The two medium quality studies lacked information on sampling strategy, participant selection, data saturation and comparability with other studies. Houchins, et al. [20] concluded that multiple themes relating to PBS centre around 'environmental congruence', and as such, contextual issues need to be addressed if PBS is to generalise to juvenile justice settings. Woolls, Allen and Jenkins [21] similarly concluded that themes relating to the implementation of PBS interact in ways that can affect the success of PBS interventions.

Service Contexts
Of the eleven studies reviewed, four were undertaken in community-based ID services Bambara, et al. [19] Hieneman and Dunlap [17]; Inchley-Mort and Hassiotis [13]; Woolls, Allen, and Jenkins [21], all of which were adult services with the exception of Hieneman and Dunlap's [17] study conducted in services for children with disabilities. The most common service context were primary and secondary schools or other educational settings Andreou, [14], Bambara, Nonnemacher, and Kern [19]; Frey, et al. [18]; Lohrmann, Forman, Martin, and Palmieri [22]; Lohrmann, and Patil [23] and all referred to PBS with typically-developing children with the exception of Bambara, Nonnemacher and Kern [19] that concerned children with ID. The remaining two studies were undertaken in forensic contexts, Davies, Mallows and Hoare's [24] study within a secure forensic mental health hospital and Houchins, et al. [20] study in a secure juvenile justice facility with high levels of a diagnosed mental health problems. Three studies, Davies, Mallows and Hoare [6], Inchley-Mort and Hassiotis [13] and Woolls, Allen and Jenkins [21], were British and the other eight were from the United States of America.

Study aims
Four of the studies had very general, explorative aims concerning stakeholder perceptions of PBS within a service setting more globally, i.e. 'applicability of PBS' in a service Houchins, et al. [20] or stakeholder 'experiences' or 'perspectives' of PBS within a service Bambara, Gomez, Koger, Lohrmann-O'Rourke and Xin, [19] ; Davies, Mallows and Hoare [6], Inchley-Mort and Hassiotis [13]. Four of the studies were a little more specific in that their aim was the identification of factors relating to the efficacy of PBS. Woolls, Allen and Jenkins [21] aimed to identify 'supportive' and 'problematic' factors, Hieneman and Dunlap [17] aimed to establish 'factors that affect success', Bambara, Nonnemacher and Kern [19] aimed to investigate 'barriers' and 'facilitators' and Andreou, McIntosh, Ross and Kahn [14] aimed to explore 'factors that help and hinder' PBS. Similarly, two studies by the same lead author had aims that related to understanding pre-determined intra-staff factors in relation to PBS application i.e. 'resistance' Lohrmann, Forman, Martin and Palmieri [15] and 'buy-in' to PBS Lohrmann, Martin and Patil [16] Lastly, the study conducted by Frey, Lee Park, Browne-Ferrigno, and Korfhage [18] was different to the others in their specificity of aim: to assess the social validity of PBS.
All of the studies reviewed demonstrated that a range of individuals can hold a stake in PBS, these can be broadly organised as: individuals who are the recipients of PBS (focus individuals), the individuals directly involved in its delivery (direct-support individuals) and the individuals involved in supervising and / or training those individuals directly involved (indirect-support individuals). Typically, direct-support individuals included 'support workers', 'teachers' or 'careers' and were characterized by having a direct 'face to face' role with the focus individual and normally spend the majority of their time in direct contact. Indirect-support individuals typically included 'behaviour specialists', 'consultants' or 'administrators' and are characterized by having a direct role with the directsupport individuals and sometimes an intermittent role with focus individuals. All of the studies reviewed here investigated at least one of the above types of stakeholder (i.e. focus individual, directsupport individual or indirect-support individual), with many investigating a mixture in some form. Figure 3 illustrates the spread of participants across the three sub-types for all eleven studies reviewed here ( Figure 3).
As shown in (Figure 3), the largest sub-group investigated were direct-support individuals (n=117), with indirect-support individuals closely following (n=105). It is clear that focus individuals as a sub-type are relatively under-represented within the studies (n=16). Regarding gender of participants, five studies did not report or provided unclear information Bambara, Gomez, Koger, Lohrmann-O'Rourke and Xin [19], Frey, Lee Park, Browne-Ferrigno and Korfhage [25]; Hieneman and Dunlap [17]; Lohrmann, Forman, Martin and Palmieri [15]; Lohrmann, Martin and Patil [16]. Of the remaining six studies in which gender was clearly reported, these studies contained 71 female supportindividuals and 26 male support-individuals.
For focus individuals' gender: 15 were male and 1 was female. Whilst the sample sizes here are small, there is a skewed picture in terms of support-females (n=71) being far better represented than support-males (n=26). The opposite skew is apparent in the focus individuals (albeit an even smaller sample) whereby male focus-individuals (n=15) are far better represented than female focus individuals (n=1). The skew in support individuals is likely representative of the general picture of support individuals within the teaching and caring professions, whereby females are better represented than males. The skew of male focus-individuals may be due to the inclusion of Davies, Mallows and Hoare [6] study, which accounts for the majority of focus individual participants in the current review, sampled from male-only hospital wards.  [14] . Lastly, two studies reported that participants had experience of PBS but they did not offer any quantification of this Frey, Lee Park, Browne-Ferrigno and Korfhage [25]; Inchley-Mort and Hassiotis [13]. With regard to the studies involving focus-individuals Davies, Mallows and Hoare [6]; Inchley-Mort and Hassiotis [13] only Davies, Mallows and Hoare [6] provide information relating to the nature of their involvement, stating service users were involved in the development of plans and also provide information for each participant regarding the length of time they have had a PBS plan in place, which offers an indication of the time individuals have been 'receiving PBS'.
Some studies in the current review presented additional information to control for the efficacy or fidelity to PBS. Two studies reported that the fidelity or efficacy of PBS (as measured by the School-wide Evaluation Tool) at each site where participants were drawn as between 86%-89% Andreou, McIntosh, Ross and Kahn [14] and 80%-99% Lohrmann, Martin and Patil [16]. Three other studies made references to informal fidelity or efficacy measures, such as selecting only participants 'successful' in PBS Bambara, Gomez, Koger, Lohrmann-O'Rourke and Xin [19]; Lohrmann, Forman, Martin and Palmieri [15], the reported presence of 'key positive behaviour support characteristics' Bambara, Gomez, Koger, Lohrmann-O'Rourke and Xin [19] p.215) or components of PBS such as functional assessment Hieneman and Dunlap [17].
All studies in the current review used cross-sectional designs with qualitative methodologies to examine reported outcomes associated with the use of PBS and compare these with qualitative findings. All studies employed either individual semi-structured interviews or focus groups with participants, with one study using both Woolls, Allen and Jenkins [21]. Nine of the studies reported that data was audio-recorded and subsequently transcribed, of the remaining two studies, one employed 'detailed' note taking during focus groups Houchins, Jolivette, Wessendorf, McGlynn and Nelson [20] and the other did not report how verbatim data was recorded Hieneman and Dunlap [17].
All studies reviewed here adopted a process of coding the data in order to progressively abstract themes (or equivalent). Woolls [19] used a modified "consensual qualitative research process" to identify codes and core ideas. Every study included a process of triangulation to improve the reliability and validity of identified themes (or equivalent). All studies except Woolls, Allen and Jenkins [21] employed multiple researchers as part of the coding and theme development process, with Woolls, Allen and Jenkins [21] using a separate focus group to triangulate data. Lohrmann, Forman, Martin and Palmieri [15], in addition to the use of multiple researchers, checked codes at a later stage with participants to increase validity. Of all studies in the current review, only Woolls, Allen and Jenkins [21] proposed a theoretical model in attempt to explicitly explain the interrelation of their resultant themes, grounded in the data.

Meta-ethnographical synthesis
Four superordinate and nine subordinate third-order constructs were developed and entered into a grid format in order to indicate where each construct was translated from preexisting constructs from within the studies (see Table 1 below). As per Britten, N., Campbell, R., Pope, C., Donovan, J., Morgan, M., and Pill, R. (2002), each column of the grid represents a third order construct. In labelling the third order constructs, the first author used terminology relating to the most relevant and prevalent themes and concepts from each of the studies under review (Table 1).   [17]. Additionally, several participants stated that positive attitudes promoted good PBS practice. These included 'commitment' Bambara, Gomez, Koger, Lohrmann-O'Rourke and Xin [19]; Woolls, Allen and Jenkins, 2012, 'ownership' of the model Andreou, McIntosh, Ross and Kahn [14], 'empathy' Bambara, Gomez, Koger, Lohrmann-O'Rourke and Xin [19] and 'optimism' or 'energy-level' for the approach Frey, Lee Park, Browne-Ferrigno and Korfhage [25]; Hieneman and Dunlap [17].
Understanding PBS: Some studies reported that participants understanding of PBS varied. This individual understanding related to PBS as a whole and encompassed 'knowledge and understanding of PBS' Woolls, Allen and Jenkins [21], 'how I understand PBS' Davies, Mallows and Hoare [6] and 'staff not understanding PBS' Lohrmann, Martin and Patil [16] .

Organisational Factors
Congruence of organisational values with PBS: Many of the themes across the studies emphasised congruence of values between the organisation and its practitioners. This was demonstrated by reference to the 'ecological congruence' with PBS Houchins, Jolivette, Wessendorf, McGlynn and Nelson [20] , the 'fit' of PBS practices within the school context Andreou, McIntosh, Ross and Kahn [14], PBS as a broader 'world view' or 'philosophy ' [17] refered to the 'integrity of implementation'. Additionally, both studies involving focus indivudals raised similar issues around their perception of PBS plan implementation i.e. 'staff not following guidelines put in place' Inchley-Mort and Hassiotis [13], p234 and 'staff fidelity to the plan' Davies, Mallows and Hoare [6].
Collaboration / Communication: A number of studies also cited poor communication between support individuals as a barrier to the PBS process, which was variously described as 'poor internal communication' Frey, Lee Park, Browne-Ferrigno, and Korfhage [18] 'communication' Hieneman and Dunlap [17]; Houchins, Jolivette, Wessendorf, McGlynn and Nelson [20] and 'communication between staff' Woolls, Allen and Jenkins [21]. Similarly, the importance of collaboration between staff members was seen as beneficial and decribed in terms of 'networking and connections' between individuals who implement PBS Andreou, McIntosh, Ross and Kahn [14], 'collaboration among providers -support providers working together' Hieneman and Dunlap [17], 'the importance of teams supporting each other in order to support people with challenging behaviour' Bambara, Gomez, Koger, Lohrmann-O'Rourke and Xin [19] and also in the negative as 'non-collaboration' between staff being a barrier to PBS implementation Lohrmann, Forman, Martin and Palmieri [15].
The relationship between direct support and focus individuals The relationships between direct supporters and focus individuals was commneted on in several studies, being variously described as 'knowing the service user' Woolls, Allen and Jenkins [21]; 'talking about behaviour and being listened to' and 'understood' Inchley-Mort and Hassiotis [13]; 'understanding the person' and 'seeing the person as a person' Bambara, Gomez, Koger, Lohrmann-O'Rourke and Xin [19]; 'understanding me and sharing my story' Davies, Mallows and Hoare [6]; 'relationship with the individual' Hieneman and Dunlap [17] and 'family', 'student' and 'community involvment' Bambara, Nonnemacher and Kern [19] ; Frey, Lee Park, Browne-Ferrigno, and Korfhage [18].
Somewhat surprisingly, studies with relatively higher ratios of indirect support individuals Houchins, Jolivette, Wessendorf, McGlynn and Nelson [20] or those studies with samples that were exclusively indirect support individuals Lohrmann, Forman, Martin and Palmieri [15] Lohrmann, Martin and Patil [16] did not generate explicit themes concerning the relationship with the focus individual. In direct contrast, the two studies that contained focus individuals Davies, Mallows and Hoare [6]; Inchley-Mort and Hassiotis [13] both had core themes concerning the relationship between support individuals and the focus individual and specifically a need for 'understanding' between them.

Discussion
The eleven studies reviewed were of medium-to highquality, which facilitated the subsequent meta-ethnographic synthesis that revealed four super-ordinate third-order constructs: Individual factors; Organizational factors; Process factors and The relationship. The meta-ethnography suggests that individual factors such as the congruence of personal values with those of PBS, the individual's degree of 'buy-in' to the approach and their understanding of it are likely important factors in the perception of PBS. With regards to organizational factors, the meta-ethnography suggests that, similarly to the individual factors, congruence of values with those of PBS is also perceived to be important at an organizational level.
It is also suggested that more general support for PBS at an organizational level is important along with the provision of resources to support the implementation of PBS. Additionally, a number of process factors are perceived as important which include: the maintenance of fidelity to the PBS process; the need for inter-staff collaboration and communication and the importance of positive wellbeing for those direct supporters involved. Lastly, the relationship between direct supporter and focus individual is perceived to be important when implementing PBS. Most studies explore the experience and perceptions of indirect and direct support individuals. Given that nearly all studies make some thematic or conclusive reference to the importance of context and inter-relation between individuals, there is a need for research that takes a 'whole picture' approach of the multiple individuals involved in PBS, especially focus individuals, who are largely underrepresented.
In some studies, both focus-individuals and direct-support individuals perceived their relationship to be centrally important when implementing PBS and there is a need for further research to explore the nature of the relationship between focus individuals and support staff from both perspectives. This systematic review demonstrates that the current research base relating to qualitative perceptions of PBS has clear scope for further research. At present, most of the research concerns school-wide PBS or PBS within ID contexts, there is hence a clear need to explore and determine the generalisability of PBS in other contexts where PBS might take place. There was a lack of detail in the qualitative studies regarding the quality of PBS training received and the fidelity of PBS delivered by support individuals. Quality of training is likely to affect individual perceptions of PBS, as such, this needs to be addressed in future research. Similarly, future research in this field needs to be methodologically robust, with greater attention to quality assurance in several areas including specifying why certain methodologies were used or how the stance of the researchers might influence data analysis and identification of themes.
Particular implications for clinical practice include the need to include the vaules of the model within training and ensuring that training results in a good understanding of the model. This can be acheived through knowledge testing after training and supervision by experienced professionals when staff are involved in the PBS process. Fidelity can be improved with interventions such as positve monitoring (Porterfield, 1987). In addition protocals and agreement regarding the implementation of PBS should involve all levels of organizational management in the aim of promoting a shared understanding and commitment to the adoption of PBS and clear pathways for its implementation.

Conclusion
In conclusion, this systematic review has used a metaethnographic analysis to draw out themes across the qualitative research body related to PBS. This research has taken place over multiple clinical fields and contexts where the qualities of the papers were between medium and high. Meta-ethnographic synthesis revealed four super-ordinate third-order constructs: Individual factors; Organizational factors; Process factors and the relationship. Based on the review of the research to date recommendations have been made for future research and potential clinical implications were identified.