Four services were approached. Three services (n = 46 therapists and their 558 cases) participated (see below). One service manager declined, stating waiting list and caseload pressures as the reasons for non-participation.
The results are reported in four sections: the participating services and their key attributes (as the specific health service contexts may have influenced the intervention delivery, uptake, and adoption); the factors related to the intervention uptake and adoption; changes in service delivery and the study outcomes during the study period; and the costs of adopting the intervention.
The participating services
The three participating services had different attribute profiles. Services A and B had more senior therapists than Service C. Services A and B covered urban, town, and rural settlements, while Service C covered solely urban areas. Services B and C covered areas of significantly low and high deprivation, respectively (Table
1).
Table 1
Characteristics, at baseline, of the therapists; the geographical locations and populations covered by the services; and the children on the services’ caseloads
Therapists’ characteristics
| Band n(%) | 4: | 4 (16%) | 0 (0%) | 1 (20%) |
| | 5: | 0 (0%) | 0 (0%) | 0 (0%) |
| | 6: | 5 (20%) | 5a (31%) | 3 (60%) |
| | 7: | 15 (60%) | 10 (63%) | 0 (0%) |
| | 8: | 1 (4%) | 1 (6%) | 1 (20%) |
| Years as therapist (Median[IQR]) | 13 (8-17) | 15 (9-26) | 7 (4-12) |
| Years in paediatrics (Median[IQR]) | 8 (2-14) | 7 (3-16) | 6 (3-12) |
Geographical and population characteristics
| Miles required to travel to attend weekly Good Goals meetings with colleagues (mean[SD])b
| 7.0 [8.7] | 15.2 [13.6] | 0 |
| Age of the children on caseload at baseline in years and months (mean [SD]) | 5y 7m | 4y 5m | 6y 4m |
| | | [3yr 11m] | [3yr 4m] | [3yr 10m] |
| % of area in most deprived 15% in Scotland [ 53] | 13.3 | 4.8 | 29.4 |
Medical diagnoses of the children on caseloads (%)(ordered based on level of medical complexity from high to low)
| Cerebral Palsy | 23 | 41 | 18 |
| Other (e.g., global developmental delay, muscular dystrophy) | 55 | 45 | 41 |
| Autistic spectrum disorder/ Attention deficit hyperactivity disorder/ Tourette’s syndrome | 14 | 9 | 27 |
| Developmental coordination disorder/dyspraxia | 2 | 7 | 0 |
| No medical diagnosis | 20 | 16 | 29 |
While all of the services provided mainly community and outpatient care, services A and B also had some inpatients. Services A and B covered entire Health Boards while Service C covered a Community Health and Care Partnership. All the service managers described the remit of their service similarly, the essence of which is captured in the mission statement for Service A: ‘To enable children and young people to meet their highest potential in everyday life.’
The services differed in their structure and processes related to management of patient flow. Service A consisted of three clinical-speciality teams (based on diagnostic groupings) and one ‘generic’ team (Table
2). The service manager oversaw acceptance of referrals to the service, and allocated children to the teams. In the past, each team had had its own identity, norms, and caseload management processes (
e.g., ways of assessing, setting goals, and reporting), and both the manager and the therapists reported that the teams continued to have limited interaction between them:.
"‘…[the teams] were very much working as [separate] services… They had their own folders with their policies and procedures… [and although things have improved] we’ve still got a long way to go, and when things pop up people tend to go back to their own teams.’ (Manager, Service A)"
Table 2
The structure, demand and resources for each of the participating services
Generic | 8/10/12 | 38 | 181 | 5.75 | 38.1 |
Coordination difficulties | 11/15/12 | 49 | 127 | 3.0 | 58.7 |
Physical disabilities | 4/1/6 | 6 | 146 | 6.45 | 23.6 |
Mental health | 1/2/3 | Not available | Not available | 3.0 | Not available |
SERVICE B
|
Not available
|
123
|
344
|
11.41
|
40.9
|
Coordination difficulties | 26/21/13 | 109 | 91 | 3.0 | 57.0 |
Physical Disabilities | Special schools | | | 28 | 1.0 | |
| Pre-school | Not available | 12 | 46 | 2.367 | 26.6 |
| Team 1 | Not available | 2 | 35 | 1.0 | |
| Team 2 | Not available | 0 | 39 | 1.0 | |
| Team 3 | Not available | 0 | 58 | 1.487 | |
| Team 4 | Not available | 0 | 47 | 1.56 | |
SERVICE C
|
12/17/11
|
42
|
186
|
4.64
|
49.1
|
Service B had the lowest children-to-therapist ratio (Table
2). It was structured around: four child development teams (CDTs); two school teams; and an outpatient service (Table
2). The CDTs and the school teams saw only children with ‘complex disabilities’ and had no waiting lists. Children who did not meet the criteria to become a ‘team child’ were placed on the outpatient waiting list (Table
2). Referrals were accepted by individual therapists; the manager reported limited control over allocation of children to the teams:
"‘…if we’ve got a child that we’ve seen [at the outpatient clinic] and we think… the team should pick them up; they may not agree with that request.’ (Manager, Service B)"
The service had an operational policy for caseload management and the manager described peer pressure for everyone to adhere to it:
"‘…[the policy is] for thinking through what you would be expected to do [at assessment, treatment, and discharge]. …There is a lot of peer pressure… If somebody finds out that somebody is deviating (laugh)… they would be challenged…’ (Manager, Service B)"
However, therapists described differing motivations to adhere to the policy. Some therapists in CDTs described accepting referrals for ‘team children’ only; they reported a belief that accepting other referrals could result in increased pressure on them. Other therapists felt that accepting only ‘team children’ was de-skilling them; these therapists described a practice of discreetly taking non-team children on their caseloads.
Service C had the highest children-to-therapist ratio (Table
2). All referrals to the service were discussed in a multidisciplinary team meeting, attended by the service manager. While therapists had clinical special interests, all therapists had a responsibility to the overall service provision.
Managers for services A and C described themselves enacting leadership roles, both in general and in relation to Good Goals:
"‘[My role in general is] to have the overall plan and to gain advice and ideas from the team; make a plan and delegate who’s going to do what’ [Manager, service C]"
"‘I have said that we’re signed up to [piloting Good Goals] so therefore they will get the time and that I see this as a priority… [it is] my job to have that longer vision and take them with me.’ [Manager, service A]"
Manager for service B described her role in terms of managing the therapists and the service policy, personnel procedures, and administrative processes:
"‘…my job is about professional standards… I supervise staff and make sure they are trained, that their workload’s okay, sorting out day-to-day management issues—annual leave, recruitment…’ [Manager, Service B]"
She reported a perception that the uptake of Good Goals was likely to depend largely on individual therapists and external factors, and stressed the external pressures and lack of resources as anticipated barriers.
Comparison of the intervention uptake and adoption (see Additional file
2: Appendix for summary descriptions) between the three services indicated that the key factors related to the intervention adoption were the mode of delivery for the Good Goals intervention (underpinned by competing demands on therapists’ time), leadership by service manager and, in some instances, therapists’ perceptions of the children and families.
The mode of delivery for the Good Goals intervention was the single most influential factor in its uptake and adoption. The training sessions were well attended across all services (82% to 100% of therapists attended), and participants were observed to engage with the materials delivered within these. In contrast, for the workbooks and Good Goals weekly meetings, the number of sessions completed (mean = 9, SD = 4, per team) was considerably lower than that intended (25 sessions per team). From therapists’ reports, the main barrier to using the workbooks was unclear instructions. The main barriers to the weekly meetings were reported as lack of time, difficulties in organising meetings when a number of therapists worked part-time, and difficulties in travelling to meeting locations.
The weekly meetings were the most commonly reported challenge in adopting Good Goals (reported by 14/17 respondents in Service A; 6/7 in Service B; and 2/2 in Service C). The change techniques delivered during the weekly meetings (especially social support, encouragement and peer pressure; and modelling/demonstration of the target behaviours by others) were reported as the most important intervention ingredients:
"‘…unless you’re coming together it’s not going to achieve its aims and you could quite easily go and do your own thing the way you’ve always done it… It’s definitely about the coming together…’ (Focus Group Service A OT4)"
However, due to the reported difficulties in organising the meetings, there was an ongoing tension between the importance of holding the meetings in order to achieve sustainable change and a threat that the meetings themselves might not be sustainable:
"‘These weekly meetings… if [they] fall by the wayside, I think the quality of what the whole thing is about will go down…’ (Focus Group Service A OT5)"
In terms of the service attributes and adoption, in Service A the service manager’s actions (
e.g., providing staff with time to implement change; actively providing encouragement and positive feedback; and changing service-level processes so that they match with the intervention principles) were reported as important facilitators by the therapists (see Additional file
2: Appendix). In Service B, where the manager reported less of a leadership role than in Services A and B, some therapists explicitly commented on the lack of a service-wide approach and commitment to change (see Additional file
2: Appendix). There was no evidence of other service attributes being directly linked to adoption.
Finally, in the questionnaire data, some therapists (5/17 respondents in Service A; 3/7 in Service B; and 0/2 in Service C) reported difficulties in carrying out the target behaviours with particular families (e.g., parents with whom therapists had difficult interactions) or children ( e.g., with complex conditions or of younger age). However, there was no evidence from the case note data analysis post-intervention that therapists were identifying and agreeing goals or evaluating progress differently due to children's age or complexity of condition. Further analysis of the focus group data and the researcher’s observations indicated that therapists’ expressions about difficulty of carrying out the target behaviours were often linked to that individual therapist’s beliefs and values. For example, the following quote illustrates how one therapist’s perception about difficulty in identifying treatment goals with some children was linked to her belief about the content of acceptable treatment goals:
"‘[some children]…come up with absolutely ridiculous goals. Two little ones, both in wheelchairs, who wanted to play football. …you say ‘you can’t do that… you can maybe get ball skills in a different setting’ but no, this little one wants to play with his brothers…’ (Focus Group Service B OT5)"
Changes observed in the study outcomes
The changes that service managers and therapists reported related to the adoption of the intervention were similar across all three services. The intervention was reported to improve equity of care through ensuring a shared rationale for decisions by 54% [14/26] of the questionnaire respondents. This was also reflected in the focus group discussion:
"‘It’s made a much more equitable service… it’s really helped us to be doing similar things with patients, which we weren’t doing before.’ (Focus Group Service A OT5)"
It was reported to increase therapists’ clarity on role, resource use, and intervention provision by 42% [11/26] of the questionnaire respondents. This was similarly reflected in the focus group discussion:
"‘I think we’ve changed quite considerably since the introduction of Good Goals … we’re much more goals focused … which then really guides us to what’s important for the child … It used to be a standard battery of assessments regardless of what was wrong with the child and what the child and parent wanted …’ (Focus Group Service A OT4)"
Finally, the intervention was reported to improve therapists’ interactions with families and schools by 38% of the questionnaire respondents. This was reflected in the focus group:
"‘It’s definitely changed the focus and [we are] asking a lot more questions. I think it empowers the kids to make a decision about what it is they want to work on (…)’ (Focus Group Service C OT3)"
During the study period, there was a measurable increase in the target behaviours across the three services (see Table
3). Estimated odds ratios (95% confidence intervals) comparing pre-intervention to post-intervention were: identifying goals, 2.4 (95% CI 1.5 to 3.8); agreeing goals, 3.5 (2.4 to 5.1); evaluating progress, 2.0 (1.1 to 3.5). LoT decreased by two months [95% CI −8 to +4 months] across all sites during the study period, adjusted for clustering at therapist-level and for the child’s diagnoses and age.
Table 3
Number and proportion of cases where there was evidence of the performance of the three target behaviours at baseline and follow-up
Service A¥
| 51 | 39 | 46 | 41 | 23 | 18 | 40 | 36 | 32 | 24 | 23 | 21 |
Service B¥
| 32 | 27 | 74 | 59 | 13 | 11 | 40 | 32 | 14 | 12 | 45 | 36 |
Service C¥
| 7 | 21 | 17 | 46 | 6 | 18 | 12 | 32 | 1 | 3 | 9 | 24 |
Total
|
90
|
32
|
137
|
50
|
42
|
15
|
92
|
34
|
47
|
17
|
77
|
28
|
In terms of contextual factors, the therapists reported, and the researchers observed, that managerial leadership was important for achieving changes in service-level processes that facilitated sustainable, long-term change (see Additional file
2: Appendix). No patterns emerged between the other service attributes assessed and the changes in the target behaviours. For example, the two services in which the largest increases in the target behaviours were observed had the therapists with most and least experience, the lowest and highest demand-to-resource ratios, and the most extreme geographical and population characteristics.