Scottish Government

The Independent Evaluation of 'Starting Well' Final Report

Part 3. The Process Evaluation of Starting Well

3.1 Starting Well's Initial Theory of Change

3.1.1 Introduction

Having summarised the degree to which Starting Well brought about measurable changes to health related outcomes for children and families, we turn now to more processual change arising from the project. In order to set the scene for the various components of the process evaluation undertaken by the research team it is important to understand the theoretical and programme logic that lay behind the Starting Well intervention. One approach that helps evaluators to do this is the Theories of Change (Connell and Kubisch, 1998). This section of the report briefly summarises:

• the background to a Theories of Change approach to evaluation; and
• the key elements of the Starting Well Theory of Change articulated in the first year of the project.

3.1.2 The Background to a Theories of Change Approach

The complexity of initiatives such as Starting Well, poses a challenging set of issues for evaluators (Judge and Mackenzie, 2002). These include the following:

• initiatives such as Starting Well have multiple and broad goals and are therefore not well suited to evaluation methods that rely on a small number of key outcomes;
• objectives are defined and strategies are chosen to achieve goals that often change over time - for example, interventions which aim to be locally driven require to respond to both community needs and these are not necessarily defined at the outset. Likewise, initiatives such as Starting Well frequently find themselves in the position of responding to changing policy requirements at a national level and this can militate against an approach that is consistent over the life time of a project;
• many activities and intended outcomes are difficult to measure since units of action are complex, open systems in which it is virtually impossible to control all the variables that may influence the conduct and outcome of evaluation;
• the saturation of a given community with a particular intervention limits further the potential for traditional experimental designs since the option of randomising individuals to treatment and control is not a viable one. In addition, the usefulness of identifying control areas is severely curtailed where an initiative such as Starting Well is seen as a initiative with a politically high profile and where some processes that are being introduced through the project are anticipated to be rolled out across some parts of Scotland before any formal evaluation has taken place; and,
• finally, improving child and family health outcomes that are socially determined takes longer than the lifespan of an initiative and its evaluation.

A range of theory-based approaches have been developed to partially address these challenges, with the two most common being Theories of Change (Connell and Kubisch, 1998) and Realistic Evaluation (Pawson and Tilley, 1997)

The Theory of Change approach, developed by the Aspen Institute from their evaluation of community initiatives in the US, is defined as 'a systematic and cumulative study of the links between activities, outcomes and contexts of the initiative' (Connell and Kubisch, 1998). The approach aims to gain clarity around the overall vision or theory of change of the initiative. In generating this theory, steps are taken to link the original problem or context within which the programme is located, with the activities planned to address the problem and the medium and longer-term outcomes that are predicted to follow.

The starting point is for the initiative (with the help of the evaluator) to reflect on the key aspects of the context within which it is operating. The next stage is to specify a rationale for intervening in relation to priority issues (for example, in response to a local needs assessment). This rationale should then translate into clearly defined change mechanisms each with specified targets that should form part of a logical pathway that leads in the direction of prospectively expressed, strategic goals.

The logic of the Theory of Change approach is that where such pathways are articulated in a way that is plausible (derived from an existing evidence base), doable (practically feasible given the time, financial and personnel constraints), meaningful (leading to worthwhile outcomes), and testable (are able to be shown to have taken place as anticipated) then the following benefits will accrue:

• project planning and implementation will be sharpened;
• the Theory of Change will help to identify what data need to be collected when and using what methods;
• Where prospectively specified outcomes are achieved at the predicted time and following prescribed activities then the question of causality can be partly allayed. (The argument here is that if the constituency of stakeholders are able to agree from the outset the outcomes that will constitute success then the achievement of these outcomes will be seen as a plausible means of attributing change to the intervention) ³⁶.

3.1.3 Starting Well's Theory of Change

The process of generating and articulating Starting Well's Theory of Change was an iterative one with the evaluator working with those at a strategic level over a one-year period. Participants were members of either the project steering group (a strategic management group) or the senior management team (an operational management group). Inevitably stakeholders' theories are evolving concepts and even at this early stage in the project may have moved on from the original proposal accepted by the Scottish Executive in the autumn of 2000.

As illustrated in diagram 3.1.1, the theory of change for the project, at its most general level, was that a 'step change' in the health and well being of vulnerable young children and their families could be achieved by a combination of intensive, individualised support from home visitors, a well developed community infrastructure and strong partnership working between statutory, community and voluntary services (A detailed version of this that sets out the project's targets within a three-year period is presented in Mackenzie, 2002).

Diagram 3.1.1. A simple representation of the Starting Well Theory of Change

Underlying this were a set of beliefs concerning key mechanisms. These assumptions are set out in table 3.1.1

Table 3.1.1: Key assumptions underlying the initial Starting Well Theory of Change

During the time that the project's Theory of Change was being refined, the Scottish Executive provided the Health Demonstration Projects with short-term resources to assist them in project planning and performance indicator setting. These processes resulted in the production of a revised project plan in March 2002. This is summarised in table 3.1.2

The evaluation report, based on developing project plans until January 2002, raised a number of issues concerning the Theory of Change articulated by the project, including its limited testability largely due to a lack of baseline information. This issue is just as evident in the March 2002 plans and is shared by complex initiatives across the UK (Mackenzie and Blamey, in press; Judge and Bauld, 2001). On the other hand, unlike many such initiatives, Starting Well invested in the development of a project specific monitoring database and in conducting/commissioning small pieces of evaluation that, in combination with the independent evaluation allowed many, although by no means all, of its key demonstration questions to be addressed (for a summary of these see table 3.1.3.

Table 3.1.2: A summary of Starting Well's Key Outcomes as extracted from 2002 Project Grids presented to the Scottish Executive (outcome measure)

Table 3.1.3: Starting Well's Key Priorities for Learning as extracted from the project grids submitted to the Scottish Executive in 2002

3.1.4 Conclusions

As with similar complex community initiatives, the strategic stakeholders within Starting Well struggled to articulate a Theory of Change that was wholly testable (Mackenzie, 2002; Mackenzie and Blamey, in press) but developed a relatively robust internal monitoring system to capture the implementation of its plans, and, as with the independent evaluation, adopted a strong focus on process learning.

The following sections of this part of the report now consider the more detailed findings in relation to some of the questions underlying Starting Well's Theory of Change:

• the extent to which intensive home visiting led to the development of therapeutic relationships between families and their home visitors (section 3.2);
• the implementation issues involved in developing a skill mix approach to home visiting (section 3.3); and
• the degree to which intensive home visiting at an individual family level led to improved community and strategic responses to child and family health problems (section 3.4).
• Finally, this section finishes with a discussion of strategic stakeholders' retrospective reflections of the project's initial Theory of Change.

3.2 Case Studies: Project Families and Health-Visitors

3.2.1 Introduction

The general aims of this chapter are to allow the views and voices of service-users and their health visitors to emerge in such a way that we begin to understand:

• the inter-personal processes that link inputs (intensive home-visiting) to outcomes (perceptions of benefit)
• their general perceptions of project strengths and weaknesses

In this way, we aim to move beyond more mechanistic account of impact (chapter 2) to consider whether and how the project 'works' at the level of individual families.

We begin in section 3.2.2 by providing a brief methodological context to the interviews including more detail on the characteristics of participants, the timing of interviews and a description of actual health visiting input for some of the case study families. Proceeding from a detailed analysis of fifty-nine interviews, section 3.2.3 goes on to present a process model linking intensive home visiting to a key set of perceived benefits summarised as 'enhanced support'. Section 3.2.4 describes perceptions of project strengths and weaknesses, and conclusions are drawn in 3.2.5.

3.2.2 Methodological context

Characteristics of participants

With regard to parity, age and number of children, half of the 20 mothers purposively recruited were primiparous with multiparous mothers generally having either one toddler or a number of older children. The median mother age was 29.5 (range 18-40) with around three-quarters of women having a resident partner. Around one third of households were workless. Six minority ethnic women from predominantly Scottish Asian (Pakistani) backgrounds were interviewed, two of whom required an interpreter to complete the interview. While participants were drawn equally from the two project areas, all minority ethnic women lived in the southern project area. Of the seven women not interviewed a second time, several were contacted but were either not at home or failed to attend more than one scheduled interview, whilst others could not be contacted on a reliable phone number and did not respond to a postal approach. Index health visitors had also experienced access difficulties with some non-responders although they did not obviously share other characteristics. One mother was not contacted on the advice of her health visitor due to a deteriorating home situation. All interviews were carried out in the family home apart from two which took place at the local health centre.

Around half of the health visitors interviewed were experienced workers, whilst for the other half (predominantly in the eastern project area) working on the project was their first post-qualification post, although several had experienced other forms of nursing. A mixture of part- and full-time staff were interviewed. Finally, two male health visitors were interviewed although reference to gender has been removed from quotes in order to respect confidentiality.

Timing of interviews in relation to project development

Fig. 3.2.1 shows the timing of each round of interviews which took place between early February 2002 and mid-June 2003. It is important to note that the 'skill mix' of health support-workers and nursery-nurses was constituted in the spring and summer of 2002, hence many cohort one mothers may not have had full access to this service dimension.

Fig. 3.2.1: Timing of case study interview rounds by cohort and participant type.

Actual service input

In order to describe the processes linking intensive home-visiting with perceived outcomes, it is first necessary to demonstrate that an intensive service has been delivered. Eight of the 20 women interviewed had also taken part in the quasi-experimental study and so it was possible to calculate the number of visits received up to the point of interview. At interview one, the median number of home visits over four months was 8 (range 3-14). Only six of the eight women were interviewed again and for this group, the median number of home visits between interview one and two was six (range 1-15). This compares to the one scheduled home visit (10-17 days postpartum) the women might have expected to receive from the generic service ³⁷. Whilst we have only considered a proportion of the total number of interviewees, these figures strongly indicate that a more intensive service was, in fact, being delivered.

3.2.3 The process of intensive home-visiting

This section begins with a summary description of the key benefits ('outcomes') accrued to the families as perceived by both sets of actors. The processes leading from intensive visiting input to these outcomes are then described, together with the modifying factors that explain variation in both process and outcome. Findings are summarised in model form at figure 3.2.2. In order to maintain the flow of argument, illustrative quotes are separated from the main text and displayed in a series of boxed figures.

Fig. 3.2.2. A PROCSS MODEL OF INTENSIVE HEALTH VISITING

3.2.3.1 Key perceived benefits/'outcomes'

Benefits as 'outcomes'

Both sets of actors were asked questions that were designed to elicit perceptions of benefit relating to the case, for example (to mothers) 'how do you feel the project has helped you?' and (to health visitors) 'what success do you feel you have had with this family?'. In general, health visitors found these questions easier to answer and, to the extent that they felt there had been benefit, were more likely to attribute causal significance to their work or to the project in general. Mothers' statements tended to be more circumspect and tentative and gave the impression that they were unused to reflecting on the impact of services in a cause-effect way. Moreover, whilst some causal attributions were made in response to direct questioning, many statements of benefit from mothers were indirect and implicit in either positive reactions to service components or in reflections on the project's 'good' points. It is clear then, both from the style of comments and, as we shall see, their substance, that the types of 'outcome' seen as important to mothers were quite different from those traditionally associated with quantitative research; often focussing on 'softer', more interpersonal achievements that are frequently treated as mediating factors in impact studies. This was also true of an important subset of health visitor statements. We use the term 'outcomes' therefore, to refer loosely to broad sets of identified benefits that had either psychological (mothers) or more professional (health visitors) validity for participants.

Health visitor accounts

The top half of figure 3.2.3 illustrates in their own words, the range of benefits identified by health visitors. These accounts of progress are generally consistent within-case and across time and mostly refer to the immediate and more enduring effects of an intensive period of input in the first few postnatal months. Two distinct sets of outcome emerged in addition to general preventative work, which by definition was a professional but subjective assessment of harm avoided. The first set of identified outcomes was the achievement of health-related targets, such as maintaining breastfeeding, appropriate weaning, dental registration, the delivery and use of safety equipment, and referral uptake. Active health promotion work was illustrated by at least three participants' description of how their advice had positively affected breastfeeding intention, weaning practices and opportunities for child-environment interaction. The second main set of outcomes refers to a diverse set of activities that may be summarised as the provision of general support. At its most basic, this support consisted of frequent and clear communication and the establishment and maintenance of a good relationship, however, high levels of postnatal depression amongst study mothers highlighted the specific focus of this support. The main vehicles of support here were firstly, the 'listening' visit which provides the opportunity for women to talk about their problems and secondly, referral to the GP. Finally, support was provided in the attempt to alleviate families' financial and material concerns, for example referral to debt advice, the provision of non-safety equipment (e.g. nappies, cots, toys) and writing support letters to housing associations and nurseries.

FIG 3.2.3: EXAMPLES OF PERCEIVED BENEFITS TO CASE FAMILIES

Mothers' perception of outcome

There is a strong correspondence within-case between health visitor and mother accounts of benefit, which again tend to relate to work done in the early postnatal months. Three women spoke of perceived developmental gains (either cognitive or motor) for their children as a result of taking part in the project; this was attributed to advice on parenting and interaction from health visitors and nursery nurses. Overwhelmingly, however (see fig. 3.2.3), the most dominant and clear message emerging from the mother interviews was the perception that the service offered enhanced levels of support. Statements to this effect were made in detail by virtually every participant regardless of need and background, across both cohorts and at both first and second interviews. The main source of this support at interview one was the health visitor but at interview two also included other project workers and resources. Favourable comparisons were made with previous experience of health visiting and also with friends/relatives receiving the generic service elsewhere in the city. Five dimensions of enhanced support associated with intensive health visiting could be identified:

i. Increased confidence: either in general or in connection with a specific parenting practice (e.g. breastfeeding, play).

ii.Reduced anxiety: in relation to help with common infant issues such as feeding, elimination and sleep/wake cycles and to a range of maternal physical and mental health problems.

iii. Reduced isolation: for a number of women for whom simple contact with an interested adult relieved loneliness and provided structure and variety to the day.

iv. The opportunity to confide: either chatting generally or unburdening at length about specific problems.

v. Experience of advocacy: for families experiencing childcare and/or housing problems, letters of support sent to nurseries and local housing associations gave a sense of someone 'fighting their corner' despite more mixed views as to the ultimate efficacy of such letters.

3.2.3.2 From input to outcomes

The previous section highlighted the extent to which perceptions of enhanced support were identified as key benefits for families. This was particularly the case for mothers but also for health visitors who valued support as an end in itself as well as a means to longer-term health-related goals. This section describes the processes that intervene between intensive input and perceptions of enhanced support.

The functional importance of time and direct contact

Figure 3.2.4 contains illustrative quotes for this section. Both mothers and health visitors identified an increased amount of time and direct contact as important features of the intensive visiting pattern. Whilst this may seem obvious, both factors were important enablers of mother-health visitor interactions in three important ways. Firstly, they promoted the rapid formation of an informal but functional relationship that was often characterised as 'trusting', 'relaxed' and 'more like a friend'. At least four health visitors spoke of the establishment of this type of relationship as a key indicator of success, whereas several mothers went further and described it as a basic pre-requisite for getting any benefit from the service. The absence of a good relationship was felt to cause tensions, discourage women from asking questions and even prompt the foreclosing of access to the family. This relationship was generally accorded greater significance for depressed/isolated women with a manifest lack of other support. Secondly, increased time and contact led health visitors to believe that they would be able to 'follow through' particular courses of action and allowed them to tailor their approach to the particular characteristics of the mother, including needs, life circumstances and temperament. This individualisation of care was evidenced by increased visits for vulnerable families, visiting at atypical times to fit in with shift patterns and adopting a non-directive approach in response to 'strong-willed' mothers. Thirdly, more time and contact simply permitted greater surveillance; the increased opportunity to see mothers interact with their children in the home environment and to discuss/assess needs. Whilst some mothers displayed initial attitudes to health visitors as being 'nosy' and/or allied to social work, several also saw the surveillance aspect of intensive visiting as positive, either in relation to the monitoring of their own child's health, or in relation to the prevention of child abuse.

FIG 3.2.4: TIME, FLEXIBILITY & THE IMPORTANCE OF RELATIONSHIPS

Communication and awareness of need

As illustrated by quotes in figure 3.2.5, the effect of these facilitative dynamics was to co-create two phenomena that contributed to both the delivery and perception of enhanced support. The first of these was a greater and broader awareness of need as revealed by more and better quality information. Both sets of actors stated that trusting relationships encouraged disclosure ³⁸, which combined with greater surveillance, and also perhaps the multi-domain family assessment tools, resulted in the identification of a wide range of physical, psychological, environmental and material problems (see table 3.2.1).

Table 3.2.1 Type of maternal and infant problems and frequency of identification

Secondly, the processes identified in the previous section lead to the establishment of an enduring two-way dialogue between mother and health visitor. In the early days of the relationship, this was more likely to be based on face-to-face contact where health visitors gave assurances that no question or anxiety was too trivial to raise and reinforced this by handing out office or even mobile phone numbers so that contact could be maintained between visits. Even if not available, several mothers described how they had been able to obtain advice from other health visitors in the project office and this potential availability of help and advice seemed to constitute a very important psychological 'safety net' . Even at interview two, when mothers displayed increased parenting confidence and the desire to disengage from the project, frequent references were made to the likelihood of getting in touch with the health visitor should anxieties arise.

FIG 3.2.5: AWARENESS AND DIALOGUE

3.2.3.3 Explaining variation in process and outcomes: modifiers

In the last two sections, we have shown how increased levels of time and direct contact lead to a series of intermediate interpersonal achievements that are, in turn, associated with perceptions of enhanced support. However, not every interviewee benefited from the service and, of those who did, there was variation in the breadth and strength of opinions. How do we explain these variations in outcome? One major source of variation (see fig. 3.2.6 for illustrative quotes) relates to a number maternal characteristics and resources which might be summarised as receptivity to the service. Many women displayed initially negative attitudes towards health visitors relating to their perceived surveillance and/or quasi-social work (i.e. child protection) role and, whilst these attitudes were overcome for most, complaints of 'nosiness' or the inappropriateness of certain lines of questioning persisted for some and may have impeded the development of the relationship. Secondly, some women seemed less able to accept help, either because to do so would admit fallibility or because they did not agree that there was a problem to resolve. This latter point became particularly relevant when an identified problem threatened child health or safety. Thirdly, at least two experienced mothers with more negative views of the service seemed to perceive routine advice to be either patronising or a challenge to their established identities as mothers.

FIG 3.2.6: MODIFIERS [MATERNAL RECIPTIVITY, HV WORKLOAD]

Finally, in at least two cases, these factors combined with very high levels of need (related to depression and substance-abuse respectively) to produce a large number of broken appointments and very limited health visitor contact.

Health visitor characteristics too could contribute to receptivity: in one case, an already less-receptive mother cited a brusque and unfriendly manner as problematic, however, the clear majority of mothers' statements suggested workers were generally adept at presenting themselves as approachable and friendly. A more important set of modifying factors related to time pressure and its effect on the work capacity of the health visitor. Consistent with the already-established perception that more time and contact lead to benefit for families, factors that impinge on that time were perceived to reduce effectiveness. The amount of time pressure felt by health visitors varied between workers, across time and project site, but when identified, was due to a combination of: transiently low staffing levels; the extent of social and health problems in project areas; the extra training and promotional work associated with the demonstration project; and the wider range of problems disclosed as a result of more confiding relationships. These factors tended to produce caseloads that were bigger and more demanding than initially anticipated, an inevitable prioritisation according to need, and an increasing resemblance to the generic service. These very strong and consistent views must be considered to have general relevance to our understanding of process as lack of time was not identified as a problem with (or by) these specific case study families.

3.2.3.4 Summary

We noted that, although participants occasionally referred to 'harder' indicators of family-level success (e.g. recovery from postnatal depression, prolonged breastfeeding), many of their statements of benefit referred to a range of activities they felt had produced a diffuse sense of enhanced support. Participants gave a clear sense of how intensive home visiting increased the amount of time and direct contact available which enabled greater surveillance, the rapid development of a trusting relationship and the creation of an individualised care package. These factors in turn, permitted a greater awareness of disclosed and observed needs and set up an enduring two-way dialogue which were associated with perceptions of enhanced support. Occasional comparative statements from multiparous mothers suggested that their recent experience of support had been an improvement on the generic service. Finally, both the general receptivity of the mother to the service and the capacity of the health visitor to deliver intensive visiting were likely modifiers of process and hence, outcome.

This model, like all models, contains inherent oversimplifications. Analysis of participants' statements suggest the functional relationship develops over time, from the mutual ignorance of the first visit through to working and disengagement phases, and as such, temporal linearity is a valid feature. However, there will be considerable variation in the rapidity and extent to which this relationship develops (the temporal 'distance' between input and outcome), some stages may occur nearly simultaneously as opposed to sequentially (e.g. enhanced surveillance and awareness of need), and the 'end point' of enhanced support may occur at an earlier stage (e.g. early anxiety-reducing advice) and have mutually reinforcing effects on relationship quality.

3.2.4 Perceived project strengths and weaknesses

3.2.4.1 Person-centredness

Over the last ten to fifteen years, a great deal interest has been shown in moving beyond a 'medical model' of health care to more person-centred models that re-evaluate traditional roles and relationships and incorporate information on the wider determinants of health (Evans and Stoddart, 1994; Mead and Bower, 2000). Analysis of mother and health visitor statements suggests that intensive home visiting can be a particularly effective way of delivering person-centred care in two respects. Firstly, as can be seen from the quotes in figure 3.2.7, the fact that visits take place in the home setting is perceived to be convenient, reassuring and conducive to making the mother feel less 'processed' than in a health centre setting, where most routine interaction with the health visitor takes place. Convenience is particularly important for mothers who might, for a variety of reasons, have difficulties attending clinic appointments. Importantly, the fact that the mother controls access to the setting fundamentally shifts the power balance in the client-worker relationship in her favour and actively encourages both flexibility and approachability on the part of the health visitor; if one or both of these things are absent, access is more likely to be denied. The second set of factors illustrating the person-centredness of Starting Well relate to aspects of the general approach. The range of physical, mental and material concerns identified in table 3.2.1 suggest that a more social/psychosocial (as opposed to medical) model of health care was being employed in the project. In addition, it is clear that the nursing approach most often delivered by health visitors and most preferred by mothers is a non-directive, non-judgmental one that may include a clear evidence-based 'steer' but leaves final decisions on most matters unambiguously to the mother. Whilst this approach may be empowering, it is also likely to be necessary in order for the health visitor to maintain access. Finally, as we saw in the previous section, intensive visiting is associated with the rapid formation of a trusting and functional relationship.

FIG 3.2.7 EVIDENCE FOR AND LIMITS OF PERSON-CENTREDNESS

All of these factors are consonant with a high-degree of person-centredness, however a number of statements and cases illustrate important caveats to this positive conclusion. Whilst the gradual reduction of visiting was generally understood and appreciated, at least two mothers, who had previously coped well, experienced problems later on in the postnatal year (later-onset depression and bereavement respectively) and felt they required more health visitor input than they received. An allied point relates to continuity of care: several women who had had their health visitor changed at the second interview found this upsetting. Taken together, these points suggest the process of (especially abrupt) health-visitor disengagement needs to be handled carefully in order to maintain perceptions of effective support. Some statements also point to the limits of, respectively, exploring the wider determinants of health and offering a non-directive nursing approach: some women found questions relating to finance or relationships to be inappropriate and prying; whilst more directive approaches are clearly needed if child health/safety is at risk. Finally, the principle of jointly-defining health goals in the form of ' goal-setting' was not conspicuously successful: health visitors found this conceptually and practically difficult to realise with mothers, and where it was reported to occur it was not always clear whether mothers were simply following health visitor advice (e.g. the goal of 'visiting the GP' for a depressed mother).

3.2.4.2 Teamwork, training and approach (health visitors)

Moving away from the case families and concentrating on more general views of the project, three sets of strengths were identified by health visitors (see fig. 3.2.8). The first of these related to the quality of project teamwork, particularly amongst the health visitors themselves. Whilst a number of references were made to early difficulties experienced in both sites due to staff turnover and shortages, many later interviews praise the high levels of peer support available. This was particularly appreciated by newly or recently qualified staff, many of whom cited support as one of the reasons for applying for the post. Secondly, the general levels and quality of available training were praised although courses were generally taken at the expense of caseload time. Finally, specific aspects of the approach were highlighted as satisfying or useful, for example, intensive home support and community work.

3.2.4.3 Constraints on delivery (health visitors)

We have already highlighted how staffing issues, the general scale and complexity of needs and increased disclosure combined to produce larger, more demanding caseloads than anticipated and commensurate chronic time/work pressure. A number of health visitors stated that this may have affected the capacity of the teams to deliver the project as intended for all families. A number of other pressures were highlighted (see fig. 3.2.8), including: the burden of scrutiny and expectation that resulted from working on an evaluated and high profile project; the widespread resistance and scepticism of generic health-visitors and other staff, especially in the southern project area; and problems of inter-agency working, specifically the difficulty of getting hold of appropriate social work staff in crisis situations.

FIG 3.2.8: PERCEIVED STRENGTHS AND CONSTRAINTS (HVs)

3.2.4.4 Barriers to the use of community facilities (mothers)

Several health visitors described how their intended community development work had been a substantive 'victim' of increased caseload pressure. However, mothers' statements (fig. 3.2.9) suggested a number of barriers would need to be overcome in order for them to take advantage of existing, let alone enhanced, services. Women who had not considered using facilities tended to fall into two categories: those who were either intimidated by the prospect of attending on their own; and those who harboured negative views about their area or people working within it. It is notable that none of the former group had taken up the services of health support workers who may have provided the opportunity for accompanied attendance; this may point to a generalised lack of confidence with novel social situations.

The positive testimony of two women who had not considered using community facilities at interview one but were attending groups at interview two (see figure) illustrates the potential benefits of overcoming these barriers.

FIG 3.2.9: BARRIERS AND BENEFITS TO USING COMMUNITY FACILITIES (MOTHERS)

3.2.4.5 Prioritisation and targeting

As illustrated by the quotes in figure 3.2.10, mothers consistently expressed the view that the most important time for service input was in the first few months after birth, this being a time characterised by overall readjustment, physical recovery and/or postnatal depression. Intensive visiting was seen as less appropriate after this time (i.e., by interview two at 10-11 months) due to the resolution of early postnatal anxieties, greater parenting confidence and the desire for independence. Support for early intensive visiting was perhaps especially true for first-time mothers who experienced greater levels of anxiety associated with the novelty and extent of the child's demands. Support for a universal service was voiced by mothers who saw the difficulties experienced in the first postnatal months to be universal and not predicted easily by factors such as wealth or age. However, they accepted that there would always be a need within a basic universal service to prioritise to those with the most problems. In general, these views were reinforced by health visitors who supported universal provision in principle but pointed to the practical impossibility of delivering an intensive service to all given current staffing levels.

FIG 3.2.10 PRIORITISATION AND TARGETTING

3.2.5 Summary and conclusions

In our attempt to describe the interpersonal processes that putatively underpin the intervention, a number of substantive points have emerged. First, perceived benefits were more likely to be couched in terms of diffuse interpersonal achievements (e.g. better relationship, more confidence) than in terms of 'concrete' health indicators. Undoubtedly some participants (and especially mothers) had difficulty or were unused to reflecting on benefit or change in concrete terms, but the number and diversity of statements describing enhanced support point to it being a valid and valued commodity for both sets of actors.

Secondly, it was possible to describe some of the processes that linked intensive visiting to these outcomes. More visits during the first postnatal months generally equated to more time and direct contact with mothers during a period of universal need, which encouraged the rapid formation of a trusting relationship, an individualised care package and the provision of more and better quality information on needs and life circumstances. This in turn was associated with the identification of a broad range of problems and problem-solving activity and an enduring two-way (functional) dialogue between mother and health visitor. Lack of maternal receptivity to the service and health visitor caseload pressures explained variation in process and outcomes. This process model, whilst limited, draws attention to the functional importance of establishing an informal, trusting relationship. Given the mother's control of access to the home, this is probably a pre-requisite for a service to be delivered at all (McNaughton, 2000), but is also likely to be both of therapeutic value to isolated and/or depressed mothers for whom the health visitor may be the main or only support (Holden et al, 1989). The model, in combination with other statements relating to caseload pressure, also draws attention to the importance of time: for delivering basic levels of service; establishing relationships; and following through therapeutic advice.

In sum, despite resistance from associated staff, the intense scrutiny of working on a demonstration project, heavier-than-expected caseloads and variable mother receptivity, health visitors succeeded in forming generally good client relationships, identifying a wide range of problems and producing perceptions of enhanced support amongst case study families. To the extent that workload and staffing levels permit, intensive health visiting is a potentially very effective way of delivering a more person-centred, 'social' model of health that may be of particular benefit to sub-groups of women (primiparous, depressed or isolated mothers) but is generally perceived to be both acceptable and helpful.

3.3 The implementation of the home-visiting model

3.3.1 Introduction

A substantive interim report based on interviews with a purposive sample of project team and strategic level stakeholders was accepted by the Scottish Executive in June 2002 (Mackenzie, 2003). This section builds on the findings of that fieldwork, where relevant, but focuses primarily on the following two issues:

• The extent to which Starting Well model of home visiting is perceived to have acted as a vehicle for changing health visiting practice; and
• The degree to which the project's health support worker model has worked in practice.

3.3.2 Health-visiting practice

In this section of the report we consider the evidence for Starting Well as a means of changing health visiting practice in two related areas: the implementation of a standardised approach; and, progress toward skill mix. In addition we consider the implications of two distinct 'models' of Starting Well that developed in the two interventionl areas.

3.3.2.1 Standardisation

Both the interim report and the views of strategic stakeholders as summarised later in Section 3.5, highlight the difficulties in implementing a standardised approach to practice within a group of health care professionals who have traditionally identified themselves as autonomous practitioners. The means by which standardisation was encouraged were: the development of a series of evidence-based guidelines; the use of a family health plan where goals jointly identified by families and home visitors could be systematically recorded; the introduction of a core-visiting schedule that indicated when visits should take place and what information should be imparted within them; ³⁹ and, a family support scale that health visitors could use to identify families requiring more or less support than indicated within the core visiting schedule ⁴⁰.

An unpublished evaluation of the use of the evidence-based guidelines by Gilhooly echoes the findings of the first stage of fieldwork and suggests that health visitors did not use guidelines in a systematic way but, instead, that their knowledge about, and use of best-practice was more tacit and implicit. Similarly, the implementation of a goal-setting approach proved to be more complex than initially anticipated. The reasons for this (as illustrated below) were three-fold. First, health visitors did not feel comfortable with the language of goal-setting:

'…it doesn't happen because, if you're having a conversation and then you say 'oh hang on' it's really hard to do that, to break the kinda, especially because you're going in and you you are trying to encourage them to kinda communicate with you verbally … I don't think any of the girls actually sit with the family and sit and say 'right I'm gonna write this, what would you like me to write? … every health visitor has done that down the line.'

Secondly, the negotiation with individual families was thought to be more important than picking off what might seem like obvious goals to health care professionals:

'…smoking, for instance, that's very often something that I feel that I'm perhaps addressing now with families, a year down the line. That you would go in and there's maybe four or five people sitting [smoking], a brand new baby and they'd all be sitting without windows open and that was something I had to ignore initially. Sometimes I would say 'could you open a window' but … it was really important to build up that relationship to do that because they could really take offence to that. … it was really things that were staring you in the face that were difficult to address.

Finally, the traditional culture of the health service was perceived to have constrained families' expectations about raising their own health needs with home visitors:

'… it's like, how a whole community perhaps perceive the roles of professionals, if you perceive that professionals are going to come in and sort me out or we'll go to the doctor and get a script or whatever. Then actually saying we're going to negotiate, empower you, to have you feel you're in charge of setting goals in your own health and so on, that's not like throwing a switch.

These findings are reflected in the health visiting literature that stresses the need for intuitive approaches to goal setting (Robinson, 1995, cited in Elkan, 2000)

In response to the realisation that goal-setting was not a straightforward process to implement, the project management team invested in staff training and in commissioning a small piece of evaluation to identify how and where joint goal-setting was operating (this work will be completed by early summer 2004). In addition those at a management level identified that, notwithstanding the difficulties identified above, change in practice was evident:

'Whereas I think a much more traditional model would have been you go in and health visitor does or directs and I think that has changed. I think the fact that they are very much more geared and attuned to the actual requirements of the family and what they need and letting them direct - the needs of the family directing actually the interventions that they give. They also challenge a lot more about why are we actually doing this.

As identified in the interim report, the use of the core-visiting schedule was also problematic as caseloads started to rise and health visitors felt pressurised to follow the schedule to the letter. Whilst those at a strategic level discussed the implicit flexibility of the schedule, this was questioned by the health visitors themselves:

'I'm not sure how much it was put across as being flexible … I don't know that we really interpreted it as being as flexible but then when I think everybody was under a lot of stress and strain, you know, management were then saying 'well, you know, this is not set in stone, it is flexible' and maybe that was to suit them a wee bit.

'…there's still a push to follow the visiting schedule.

The flexibility of the core-visiting schedule is key to the debate about targeting. Starting Well was initially conceived as a universal service targeted at vulnerable communities; yet as time went by and the realities of the financial constraints on potential roll-out became more salient, the need for some degree of targeting became accepted as a necessity. Whilst purists might be critical of this significant deviation from the original plans, some of those at a strategic level were critical of the project's failure to grasp this nettle at an earlier stage:

'I think we should have thought earlier on about was it not just a bit glib to say we won't stigmatise vulnerable families or a vulnerable community. What would that mean in practice… We don't have resources to universalise the project so clearly we are going to have to think more clearly about that. I think we should have realised that earlier on. And planned for it. And I think that we have been somewhat playing footsie under the table around that at times.'

On the other hand, the universal approach of the project allowed the development of a more preventative approach to child and family health issues:

'…there is a risk I suppose that we move to targeting the most vulnerable whereas actually what we thought we would do was targeting the less vulnerable and preventing crisis and preventing them becoming the most vulnerable.'

The home visitors themselves continued to express both a need to target to alleviate caseload pressure (this was particularly salient for one health visitor whose job remit was split over a Starting Well and a vulnerable non Starting Well caseload) and to provide the kind of support to families in the early days that would facilitate the future accessibility of the project teams:

'…it can actually frustrate me when I'm seeing that I've got to go and see this person in starting well who has lesser needs. I've got to see them weekly or fortnightly at the moment, but I'm not able to carry that out for a family who are going through horrific things.'

'…they might appear to be fine in the first couple of months and you know you might have, if you were having to make a decision at the beginning of the first couple of months, you might have said 'that family are fine, they really don't need the intensive support from me' you know, two parents, good support from the family blah, blah, blah and no concerns. But then the other side of that is that there is also a lot of families in that situation where the more contact you have with them you will build up your relationship if things do come out about other social problems within the family, like depression… I think I personally would find that very difficult to then suddenly say to somebody that you haven't seen for a year, you know, 'oh god, this is a disaster and everything is going haywire.'

What became increasingly apparent over the life of the project was that there was no simple means of assessing vulnerability since it might be chronic or acute and may or may not be predicted by a range of traditional risk factors:

'…it's not a straight sort of mathematical equation.'

'…do we identify those and there isn't a good way to do that at the moment. The best way seems to be on the basis of health visitor perceptions - that's as reliable as anything else.'

The way in which complex professional judgement came into play in determining need and labelling vulnerability is illustrated by a health visitor who talked about the relationship between a family's support scale assessment and the intensity of visiting:

'…only times it doesn't really correlate with the support scale is if you've not updated the support scale because you've just kinda said 'oh this is a wee blip, we'll just kind of work through it'. And sometimes after a few weeks you can go back, you can build yourself back up to monthly visits.'

A more targeted approach to health visiting is recommended by Hall Four (Hall, 2003) and by the Scottish Executive's response to it (Scottish Executive, 2004). Once again, the health visiting literature on assessing vulnerability as the basis for targeting highlights the importance of non-reductionist approaches (Barker, 1996; Appleton, 1994) and supports the need to view risk-factors as inherently unstable (Elkan, 2000).

3.3.2.2 Providing a skill mixed approach to home visiting - the broader project team

For many strategic level stakeholders a key outcome for Starting Well was as much to pilot a model of skill mix as to impact on child and family health. Both Hall (2003) and the Scottish Executive (2004) assume the need to base future workforce planning on a skill mix model. In this section we consider the extent to which the project teams were viewed to have worked effectively in developing a skill mix approach ⁴¹ and discuss the learning that can be gleaned from the extent to which this approach leads to empowerment or encourages dependency in families with young babies.

The initial planned make-up of the project teams was health visitors and health support workers with a bi-lingual worker in the south team. Concerns over health visitor shortages and the delay in recruiting the support workers, however, led to the development of new roles within the teams; specifically, two community nursery nurses and a community support facilitator in each area. The interim report (Mackenzie, 2003) describes the early development of these teams where, not surprisingly, a lack of role clarity was an issue for those working on the ground.

One year later there was more evidence of effective team working but issues of role clarity remained. Health visitors described and valued a range of tasks performed by nursery nurses such as undertaking routine visits, providing dental health promotion information, using Triple P, and delivering play/developmental interventions. However, despite efforts to clarify their role and, in particular, to distinguish what they did from the tasks of the health support worker, nursery nurses themselves were still concerned that duplication occurred and that they didn't feel supported as team members:

' think it's just a bit of insecurity for everybody and a role hasn't been designed. What's came down from the top is, that they don't want a defined role, that they think that that doesn't help and there's nobody prepared to say ' no that's not your role, that's your role', there's nobody prepared to take that on.'

'…you just feel like, we're here and they're there and you don't really work together, there's no real crossover.'

Much of the disquiet expressed by nursery nurses (unlike the health support workers) appeared to stem from a feeling of isolation in their new role due to their small numbers in comparison to the other professional groupings. This is an important issue for the wider NHS in light of the increasing employment of individual community nursery nurses across Glasgow that has been taking place since the beginning of Starting Well.

Given the pressure on health visitor caseloads it is perhaps surprising that there appeared to a relatively small number of families that were utilising the additional support provided by other team members other than on an ad hoc basis. Most health support workers and nursery nurses talked about seeing only a handful of families (between two and eight) on a regular basis and this tied in with health visitors' views on how many of their families used the additional support. One health support worker summarised this across one of the intervention areas as follows:

'I think we're only getting more families because there's more and more families taken on board. I think the ratio, or the amount of families that are actually getting help, hasnae went up very much - 40 or 50 out of about 400 families.'

The interim report discusses the role of health visitors in taking a community development approach to tackling ill-health in children and families and concludes that, whilst the project initially offered health visitors the opportunity to develop this aspect of their work, it was not an approach that was uniformly well-understood by staff; nor did they perceive that their caseloads allowed them to take this role (Mackenzie, 2003). One year later whilst some health visitors still expressed disappointment that this side of their role was not being developed, at least half no longer believed that this was an area in which they had expertise.

' I think I may have misread into what the project would be expected to achieve community side of things at the start because I thought it was gonnae be a lot more hands-on community development.'

'… community development, it's not really my thing as it's turned out and I thought I would be really keen.'

The devolvement of the community development role to the community support facilitators and to the bilingual worker appeared to be viewed more positively in the later fieldwork but this certainly raises questions about mainstreaming the project in the future. If health visitors in their role as public health nurses are not motivated, skilled and resourced to take a community development approach then the role of community support facilitator for child and family health needs to be build in to child primary health teams. Hall (2003) clearly states the requirement for such approaches to be embedded within health visiting practice specifically and the Scottish Executive (2004) argues for community development to form a major role of the new Community Health Partnerships.

The skill mix approach that evolved within the reduced caseloads of Starting Well allowed the provision of a range of intensive supports to be delivered to families over prolonged periods of time. A question raised at the beginning of the project was the extent to which Starting Well would enable families to become empowered or whether it would act to develop dependencies within vulnerable groups. The answer, not surprisingly, given the real life complexity of family circumstances was that it did both of these things.

The following example from a health visitor described the way in which a particular family was able to use the project as a positive support:

'… there was a mum of three who on paper doesn't appear to be a vulnerable mum, but as it turns out, as thing progressed, she was depressed, her husband was depressed. Now that we've sat down and we've spoken about it … she was able to access help for herself and her husband. She's also involved with the breast feeding initiative now whereas she felt she didn't have the confidence before … she's now looking at what employment that could lead on to … she's a volunteer now with the initiative, and she's now considering midwifery as a career. … when you've invested time with a family and the family have also invested time themselves and done all the hard work themselves and you can see it paying off. Because now she can leave her children in childcare whereas before she couldn't, she's now got the confidence to access things and there is less of a role for me now as her confidence grows, which has been great to see.'

This perception of real change is shared by the majority of health support workers:

'I havenae really had a family that hasnae sort of progressed that I've had a long time with.'

On the other hand some project team members provide examples of families who prove to be very resource intensive yet are not perceived to have changed in any long term way (predominantly because of their difficult life circumstances):

'I have a school age mum, her parents who she's living with are both on methadone scripts and she has a young baby and we've had all members of the team going in, you know, the nursery nurse going in and trying to motivate her with how to play with your child and how to stimulate her to bring on her development. We've had the health support worker, who's helping her with basic parenting skills, healthy eating, healthy cooking, trying to get her out and about to different places. Myself sorta over-seeing everything and discussing issues such as like furthering her education … social work is also going in there and basically this again is a difficult to engage family. Basically when you're going in and talking to this mum or going in and trying to do practical things with her, she's giving absolutely no response at all. She's not wanting to go down on the floor and play games with the baby, you know, with the nursery nurse and her needs are absolutely crying out at you. So, even with all this regular visiting and building up a relationship with her for about 14 or 15 months, she's had a visit at least once a fortnight from someone in the team for that length of time. And I would say that we are no further forward, we've maybe done the odd thing like physically force her down to the dentist to get her registered, get the baby registered but I mean, how's that progress. You know, you can register a baby at the dentist and they'll never go back and their teeth are still rotten because the diet's awful anyway.'

However, many health visitors viewed the question of empowerment and dependency as rather more complex than simply labelling individual families. Thus empowerment might have emerged from a period of dependency and/or the health care professionals themselves might be responsible for creating the circumstances in which dependency thrives:

'…it's maybe not the right time for them to develop because to make any changes they have to be at that level, sometimes they just want the sticking plaster that you can provide to get through the day, rather than, they don't want to look any further than today because that's the way their lifestyle is. And if you're at least supporting them in that aspect then hopefully in the future other things will fall into place.'

'you get the joke story about the girl that will not change the nappy without the health visitor. That's a sad reflection on the health visitor not the mum … I think it's the skill of the health visitor who doesn't encourage dependency and who encourages empowerment.'

In addition most staff members were positive about the process of reviewing work with families and of withdrawing specific forms of support (such as taking older children to nursery) if they were not seen to be leading to positive change.

3.3.2.3 The development of two 'models' of Starting Well

In considering the extent to which Starting Well has afforded an opportunity to change health visiting practice, it is necessary to consider the way in which the project has evolved within the context of its two Local Health Care Co-operatives (LHCC). This section summarises the differences between the intervention teams in the east and south and discusses the implications of this for future mainstreaming decisions.

From the outset the two teams have approached Starting Well differently. In the east, LHCC management accepted the view of the demonstration project as a stand-alone entity and a team was developed around a group of enthusiastic and largely inexperienced health visitors. In the south, meanwhile, the LHCC management was keen to push for integration from the outset and the health visiting team consisted of a more experienced but more antagonistic group of health visitors (a large proportion of these moved back to generic posts during the course of the project). In both teams relationships with 'generic' health visiting staff was poor. Over the course of the last three years, these individual, group and organisational differences served to crystallise into two very different 'models' of Starting Well. In the south, as a means of salvaging the project, the management team within the Health Board negotiated a contract with the LHCC to manage Starting Well locally as an intervention integrated within GP practices. This required all staff to work to a corporate caseload with considerably larger caseloads and a much more targeted approach to delivering the core-visiting schedule. An attempt to encourage the east down a similar route failed and there the project team continues to exist and operate an approach that broadly reflects the original Starting Well approach.

At the point of interviewing staff in the south, the move to integration within practices was starting to happen and health visitors who had previously not viewed themselves as 'part of a Starting Well team' (Mackenzie, 2003) were largely positive about the direction of travel. Some who were still operating a more 'traditional' Starting Well model were cautious about what the realities of increased caseloads would mean to the service they were providing:

'…if they think that … one health visitor, a health support worker and a nursery nurse then 450 families is quite acceptable , it's not and you haven't listened to anything that the people around you have been saying.'

In the east too, there was a cynical and fearful reaction to the prospect of integration using the south's model:

'… there is no way on earth with caseloads of 300 that they are doing anything like a starting well model so then you argue - what was the point in that? They've got a nursery nurse and a health support worker, they could of given everybody a nursery nurse and health support worker two years ago and saved themselves three million pounds.'

'…if we take away the kind of the framework of starting well there is a danger that things will just, you know, go to jelly and wobble away into corners and never be seen again. … I think there would need to be some still kinda central driving force to maintain it, because otherwise the GP focus or the LHCC priorities or just whatever other things would take over again, yes, I think we would be lost in the maze.'

In essence the tension that developed within the project was between a model in the south that encouraged the development of Starting Well skills and practice across the LHCC but that risked considerable dilution of the original approach and a model in the east that concentrated on maintaining the core principles of the Starting Well principles that then risked a degree of isolation from the rest of primary care practice. This is a classic tension between the implementation of a project and mainstreaming elements of good practice.

At a strategic level, those most closely involved in management within primary care were most closely aligned with the position that mainstreaming skill-mix approaches and the diffusion of good practice were the key objectives for the future. This is illustrated as follows:

'I don't want to have this 'fix it' team or elitist team and I would be more comfortable if we spread the skills around.'

On the other hand, those at a strategic level most closely associated with the project itself had concerns about good practice spreading so thin as to be unrecognisable:

'I have real concerns about dilution'

'I think by integrating it, it becomes a different beast. Whereas the work that's happening in the east is still to an extent the starting well idea but the more you integrate I guess, and I'm not an expert on David Olds' idea, but it becomes more and more diluted.'

Within this debate it is important to reiterate that the models themselves were artificial constructs and the two Starting Well project teams have always been 'different beasts'.

This was recognised by the project management team:

'…things have moved in the east and south differently based on different histories and the way in which things are beginning to move anyway. So we have kind of, rather than fighting against that, we've gone with that and it was really, the process really began in the south because it was a way of moving the project forward in a way that was kind of, people were making noises toward.'

However, because there are now perceived to be two different 'models' there is a danger that integration within GP practices is now viewed as the ultimate aim of the project rather than one possible means of achieving a more creative and intensive approach to addressing the health needs of children and families. This momentum is illustrated below:

'…so it was not 'the south is better than the east' it was the other way round, however, when it got to the Scottish Executive, because of their own need and value that they are placing on this integration they are saying 'ah ha, it's working much better in the south' which is very ironic because that's not where we started from at all. What we were doing in the south was saying this is not a great situation but if we are going to turn around we are going to have to move in a certain direction.'

A broader view of integration might also be considered where a skilled team delivers targeted care as per the original model but that over a much longer time period helps to generate these skills over a wider group of staff, for example:

'I suppose it depends what you mean by integration. I think integration doesn't mean well 'its adding more resources to the primary care teams and people share caseloads and are able to target within that to those with most need'. It might be that you would still have a team of people that people referred into.'

In addition, viewing the diverging approaches as 'models' masks the contextual factors underpinning the differences that have emerged and leads to implicit questions about 'which is best'. The variation in home visiting programmes is a constant theme in the evaluation of their effectiveness (Gomby et al, 1993; Powell, 1993). The reality for Starting Well was that there was little likelihood that the approach could be uniformly applied in two very different areas and furthermore, given the decentralising agenda of Community Health Partnerships, that a 'model' would emerge that could be easily delivered elsewhere:

'…you go to Govan and Pollok and it's a rip-up, you just don't impose products on them that have been working elsewhere then miles across the city. So I think the first point is why not let different models flourish if it turns out that they work in an ad hoc way for local circumstances then that's precisely to the good … what ever happens, if particular staff sites or professions are worried about their future, pissed off with it being disengaged for years and all the rest of it you're going to have problems whether you put them in the same staff room or put a note on their door saying by the way, starting well is going on, there are going to be fights in the corridor and I think one thing both models show is that you get fights. … we need to be careful that we don't turn models out of what happens to be convenient.'

This reality applies not only to the mainstreaming of Starting Well but more widely to policy roll-out. The interim report (Mackenzie, 2003) discussed the difficulties in implementing a broad public health approach within primary care. These included a lack of consistency to what the approach entails and variable commitment to its implementation. The degree to which Starting Well provided the potential for sustainable changes in health visiting practice will depend on the consistency and commitment of the organisations within which change is now located. The variation in these primary care organisations is recognised at a strategic level:

'… leadership is a key aspect of replicating either model … some are avoiding it and hiding from it and others are up confronting it and you know all sorts of variations are going on their right now. I think when we come to the next phase we will see, some settings will be at different stages of change caused by the public health roll-out. Some will be stuck in the same spot they were and so it will be even harder to them to learn these lessons but others will be so far ahead in their implementation of those that they may not see it as that difficult a thing to incorporate these ideas into their future direction.'

3.3.3 The health support worker model

The interim evaluation report produced in 2003, discussed early findings in relation to the implementation of the health support worker model. In this section we reiterate the key elements of this model and discuss stakeholders' perceptions of its success during the final year of Phase I.

For a range of practical, philosophical and economic reasons a key tenet of Starting Well was the employment of a group of lay or health support workers to augment the work of health visitors. These workers were recruited predominantly from the local areas covered by the intervention. In the first year of the project it was agreed that the health support workers should be employed directly through a voluntary sector organisation (One Plus) rather than through the NHS and that they would receive support, intensive training and direct management from a health support worker coordinator based within One Plus. Within the intervention areas, health support workers would work closely with health visitors who would retain management and legal responsibility for their individual caseloads.

Strategic stakeholders expressed a view, at the beginning of the project, that central to the successful working of the model would be: good communication between the two employing organisations at a strategic and operational level; role clarity for all team members; and, mutual respect between the health visitors and health support workers. The early perceptions of staff working on the ground indicated that, whilst good progress had been made in a relatively short space of time, there remained teething problems in each of these inter-linked areas.

We turn now to views on how these areas had progressed by 2003.

3.3.3.1 Establishing a clear role for health support workers

The interim report discussed a range of activities in which health support workers engaged. These included a range of practical tasks including child-care (within the home and in escorting children to school) and domestic tasks (such as light cleaning or shopping). Social and emotional supports to mothers were also viewed as a substantial part of the support worker's role with mothers taken to a variety of community activities or encouraged to re/engage with training and employment opportunities. A small number of health support workers spent considerable amount of time advocating on behalf of families to a range of agencies such as housing or money advice. Some had become engaged in developing community activities such as a sports club for young children in Gorbals and a men's group for fathers in Easterhouse.

One year later a very similar picture emerged with workers undertaking a wide range of tasks that might vary between families and between workers themselves; in addition, some support workers were undertaking routine visits in place of the health visitor in charge. This wide scope is encapsulated by the description of a the post by One Plus and by the summary provided by a health support worker:

'…[a support worker] is a person who is multi-skilled, able to relate quickly and easily with local people and has a knowledge of the local area. The role is holistic, it's not task-oriented, it is a person who is able to go in and identify issues, relate these back to health professionals and be an asset to the health professionals and that's how I would describe a health support worker.'

'…what I see my role as is basically providing support for the families, be that emotional support, social support, practical support and just helping them reaching a goal … if they have any problems, target that, sort that out in terms of financial advice, or housing, childcare.'

The typical way in which these roles develop within a single family are illustrated as follows:

'I think the biggest progression I had was one family where you just went in with a ROSPA pack and she was sitting and looked like she maybe needed a wee bit of help and I don't think she'd went out the house other than to do the shopping for three months. Basically we took her to one of the mother and toddler groups, a month later she was doing it herself and then the next month she was at college.'

The appropriateness of undertaking practical help within families continued to be justified as a vehicle to tackle wider health issues such as postnatal depression and there were no instances of health support workers being required by health visitors to perform tasks that they felt were inappropriate. Where they felt abused by individual families (as with the example below), the process of case review undertaken jointly with health visitors (as discussed in section 3.3.2.2) served as a means of tackling families' expectations:

'… there was a particular family that I was going into, again to help with their shopping, help her to get her money and stuff for the kids, for nine weeks. I was just going in, it's what I felt like, while she slept. I was basically going sitting with her two weans whiled she went to bed and I was like 'no, I'm no having this'. I just felt that she's pure abusing the system because I used to go in at nine o'clock to 11 and she was like 'gonnae no' come at nine cos I'm still in my bed, I don't get out my bed till ten.' So I was going in at ten, wakening her up out her bed and then she's like 'I'm dead knackered, I need to go back to bed again' and you're like 'no, I don't think so.'

The interim findings suggested that judgements about the appropriateness of different roles were made in the same way as within familial or friendship relationships. This was endorsed by the second round of fieldwork; one health visitor reflected:

'…you then hear whispers about people laying carpets for people and things like that and then I'm thinking 'oh, that wasn't an agreed goal' but I think that's just because the health support workers are there and they want to help out so, if the family is needing that done and they feel capable, they'll just get stuck in … I think it's very difficult to differentiate between befriending and becoming a friend, because if you were in your friend's house and they needed help with something that you could help them with, you would just automatically do it and I think that's where things like laying carpets and plumbing in washing machine come in.'

The success of employing local workers was partly premised on their being better able to form bonds with families than those perceived as professionals but this was perceived by health visitors to bring an added pressure to bear:

'…the health support workers, because they're from the local community, aren't getting any peace when out with their own children or out there in the shopping centre at the weekend. People are coming up to them and saying 'oh, you got so-and-so a cooker and what are you going to do for me?'

Some health support workers described the need to separate out friendship from their working relationship with families (This concords with the distinction made by Cox between friend and befriender, Cox, 1986). In some cases this happened where workers were already known to families and where they had to operate strict codes of confidentiality in relation to disclosed information; it also occurred where families crossed the lines ⁴²:

'…you've got to keep that professional boundary in your mind, you know, when you go into the family … there's even been cases where we've walked in and there's been like alcohol kind of sitting in the premises and the mum's there, she's had a glass poured out and went to take a drink. And you know, it's like, 'well I'm really sorry, I can't stay if you're gonnae drink so I'll come back and make another appointment to come back and see you.'

The role of health support worker then developed into a complex one that required a wide range of skills including flexibility and the ability to negotiate. Whilst many support workers were happy to encompass this wide range, a small number of project team members believed that two sets of skills had developed requiring two different types of individual. These are summarised below:

'…there's scope for two types of health support worker. Some people are very comfortable being like the mother type, going into a family and, moving them along for that short period … other people … just wouldn't be suited to that, it would be a whole waste of resource …initially families need the support but then they'll get to the stage where they need somebody … to pick that up and then move them forward and it's not a case of just getting by, ticking by, functioning, it's a case of getting people to believe in themselves and have the confidence to go out there and get a job.'

Whilst it may be that some health support workers were more comfortable with one or other of these sets of skills there would appear to be no a priori reason for believing that the two sets of skills cannot co-exist. Rather it would seem to be important that development of both remains central to the role of health support worker and to recognise that some of the variations in the role played by workers were perceived to stem from individual differences between health visitors as well as the support workers themselves:

'…we're quite wide in the sorta stuff we're doing, we're not sorta everybody doing exactly the same thing, it all depends on what the health visitor's asking, some of us do this, some of us do that.'

'… so it's about accepting that … jimmy will let you weigh babies, but jeanie doesnae like you to do that.'

'… the health visitors feel, you know, everybody has got their own wee niche and their different kind of strength. And they find like certain families will be suited to different health support workers.'

3.3.3.2 Implementing a dual management structure for health support workers

The rationale for employing health support workers through a voluntary organisation rather than through the NHS was that the voluntary sector in general, and One Plus in particular, had an expertise in employing, supporting and training those who have traditionally been excluded from the labour market. In addition, this management structure would allow the opportunity for a genuine partnership between statutory health organisations and the voluntary sector that might encourage a broader perspective on health issues (Mackenzie, 2003)

The interim report presented largely positive perceptions of the training provided by One Plus. Since the first round of fieldwork support workers had received accreditation in Triple P (the parenting programme implemented through Starting Well) and their on-going, modular training had been endorsed by the Scottish Qualifications Authority. The Starting Well health support workers represent the first cohort of workers to have received this health and social care qualifications and all who participated in this fieldwork expressed career aspirations in a similar field.

However, perceptions of the difficulties of implementing a dual management structure documented in the interim report were echoed, and in some instances heightened, in the final year of Phase I. Whilst for the majority of health support workers this was not a major issue (except in so far as individuals had to go through additional hoops to seek decisions), for health visitors the problems of maintaining responsibility for a caseload without direct management for the support workers had not been alleviated over the course of the project.

These problems were multi-faceted and included: a perception that communication breakdown was inevitable where there was more than one line of management; a belief that the support workers were managed by an organisation lacking in a day-to-day understanding of the work of health visitors; a sense of professional dissatisfaction over the lack of control over another group of workers' time; and a view that split management was antithetical to team working:

'…it's like Chinese whispers, 'my coordinator told me this' and that's the information you're getting from the health support workers. So then you act on that information, then it turns out that's not the information and it's just people have picked it up differently and then it's gone round like wildfire.'

'…if the health support worker came into work late continually or phoned in sick and change visits without negotiating with you, you actually have to phone somebody who has, in my opinion, no real idea of what you're doing out there.'

'…if they say they are busy you are like 'right okay' so you go and look for somebody else and you can be three health support workers down the line before you actually get somebody. Sometimes you have your doubts about just how busy is busy, kind of thing …you're not really, you're not their boss, you're not their manager so is it up to me to actually question that or do you just have to take it that they're busy?'

'…they don't know who their boss is…and I think that's been wrong … it's been 'could you do such and such' 'we don't do that' 'so, you're like ok - we're not really working as a team.'

It is disappointing that despite attempts to alleviate these types of problem that the perceptions of most health visitors had become more negative in relation to the model of employing the support workers. Only one expressed the view that the benefits of the system outweighed its costs:

'I still keep coming back to the fact that I don't think we'd have done it if we hadn't had one plus employing and managing them. I think it would have been a complete disaster … I think we need one plus's experience of supporting people, you know, from kinda like this labour market that haven't got the skills and the experience, the kinda professional qualifications, I think we would have just kind of made a complete hash of it.'

Views of those at a strategic level remained largely optimistic and supportive of the add-on value of utilising an organisation such as One Plus. This was largely related to a view that the NHS is not suited to the support of vulnerable employees:

'I think philosophically I remain entirely in favour of it and I like that link between the traditional statutory provider and the kind of the Goliath and the David out there who has particular expertise in a particular area who can feed into this process and provide a particular kind of support and a particular kind of understanding for the kind of people that we are dealing with in that employment programme. … I think the danger [if lay workers were employed within the NHS] is that we just replicate our own hierarchical structures. We replicate our own way of kind of looking at people and working with people and I am not sure that our health support workers, some of our health support workers and some of our very good ones would have survived the kind of system we had in the NHS.'

'…if there had been a problem that a health support worker had, you know, in the NHS, there would be a risk that you would just think 'oh well, we need to discipline you and sack you.'

The role that One Plus played in providing supported employment to the lay workers is highlighted as follows:

'…the stresses and strains that come with recruiting local people are without question much greater than if you were employing maybe a middle class woman who's looking for extra money to go her holidays. Now I don't mean that in a snobbish way, I just mean that's a fact of life …we've actually gone to the complete grassroots, we've given people the chance who may otherwise no' have had that chance. … I could give you a million examples, there's the girl I was talking about, having absolutely no money to get to her work, the long-term sick, you know, people who have been physically abused by partners who have alcohol problems and drug problems, you know, marriage breakdowns with husbands taking houses off people and all that kind of thing and we have actually worked through that and maintained these people in employment.'

The negative views of health visitors were explained by a lack of ownership of the aspirations of the health support model and by difficulties in communicating new systems of practice. First, the overwhelming desire of health visitors to have pressures lifted by the support workers that were sold to them as a means of reducing their work load, led to disillusionment around the support required by these new workers:

'…the aim about that, at a strategic level was, you recruit people from the local area who have experienced exclusion, give them support, they can become health support workers, there's a connection between them and the families and they can bring something to it. The way it happened and the way the project was recruited, I don't think that the health visitors necessarily bought into that part of the argument. They were looking for a pair of hands. And when that pair of hands sometimes came with a little baggage on it, it wasn't something [that was viewed positively]'

'…they've never alleviated workload for the health visitors as was originally envisaged they would do. Just never have done and I don't think they ever will do … they are always gonnae need supervising, monitoring and feeding back and all those things that really take quite a bit of time.'

Secondly, project management messages about the way in which health visitors were to manage their caseload and the day-to-day work of their project teams were not perceived to have been effective:

'…we can't get the message through. Now that says something about the way in which we need to prepare G grade nurses and I don't think we do prepare them very well and I don't think we can move their models and experience from hospitals out into the community.'

'…it should be reiterated that it is supposed to be a nonstigmatised team working approach and we don't have this hierarchy. I think the health visitors need to be reassured that they are the caseload holders, they are the managers of their families, you know, they're legally responsible. But I'm quite sure the health support workers are blue in the face hearing this because I've told them a million times.'

The move from autonomous, GP attached practitioner to caseload manager has therefore required considerably more time and energy than initially anticipated by the project.

Whilst the majority of those at a strategic level viewed the involvement of One Plus as important in the future employment of support workers (and indeed the organisation is now involved in providing a variation of the Starting Well model in other Local Authority areas), there was a view expressed that, in the longer term, it will not be acceptable for statutory organisations to delegate the employment of vulnerable individuals to voluntary sector organisations and that the capacity to support these workers will require to be developed in-house:

'I do think one plus was a good solid partner but I think the partner relationship should henceforth lie elsewhere … I think that the NHS does need to consider the issues because if it is going to get anywhere around its current labour force shortages, it is going to have to dig much further beyond the sort of jobseeker allowance into people who are long-term unemployed, on incapacity and invalidity with addictions and all sorts of other problems … it comes from a more general perspective that the NHS with thousands of vacancies in Glasgow in the next few years and not all of which are consultant level but form lab techs and porters and sorts of entry level jobs, has to sort this thinking through anyway, about how we can get a labour force of any sort, because they're not coming from Bearsden to do those jobs and how we can maximise the social inclusion aspects of what it does for the mass of Glaswegians who don't have work.'

3.3.3.3 Health support workers and health visitors: a relationship based on respect?

The third requirement for the health support model as expressed by strategic stakeholders at the outset of the project was that health visitors and support workers should treat each other with respect. The first round of fieldwork found that, whilst relationships were largely positive in one of the project teams, there was some evidence of hostility in the other that centred on a lack of respect (Mackenzie, 2003).

The recent fieldwork found that most health visitors in both areas were enthusiastic about the role played by support workers:

'God if I was in a generic post or another post I would actually miss having, knowing that they're there'. Alright, I might not use them all the time but it's just knowing that they're there and knowing that there is somebody else that you can refer the family onto for these practical things. And I'm sure I'd be kicking myself in another post and think 'god, I wish I had a health support worker now.'

'…please don't take them away, you couldn't do it, I mean you couldn't it would be really unfortunate if you had to move away to a caseload with no skill mix now, you know, because they're just, they're a godsend, you know. They're able to get in, they're able to follow up stuff, they're able to kinda keep an eye on people that you kinda think, that would be great if there was three of me. And, in a way, although there's not, there's still three sets of eyes … I'm their number one fan.'

Likewise, health support workers were positive about the way in which they were treated by health visitors:

'I think they've put their trust in us right away because, at the end of the day, it's like anything you put your trust in and if it goes, it's downhill. The trust has got stronger and stronger and stronger … there's not a single health visitor that I've had any issues with, not one'

'…we have got absolutely no problems between health visitors and health support workers. I'm treated with respect'

However, an issue raised with some frequency particularly within one project team was the extent to which particular individuals with personal problems were providing inappropriate support to families, crossing the boundaries between a professional and non-professional relationship, providing families with their home numbers, and, inappropriately liasing with social services without communicating with the relevant health visitor. The following quotation expresses the concern stated by a number of health visitors that vulnerability within employees may be problematic when dealing with similarly vulnerable families:

'I would be looking for somebody who's had life experience, somebody who has been able to deal with their life experiences … not somebody who's carrying baggage around with them … I mean, everybody carries baggage but that shouldn't be their agenda and why they're doing the work.'

Interestingly, this perception was shared by one health support worker who felt alienated from her co-workers:

'…the majority of the people I work with have more problems than the families they visit … I think we've got to remember we're dealing with real families and real people and I think if you scratch the surface you've got to be there for that open wound … if you're no' solving your own problems, how can you detach yourself and say to somebody else 'well this is what I would do? … I'm not saying that people that don't work shouldnae have a chance but I …don't think people with major, major problems are right to go into other families.'

This view highlights again the question of whether the social inclusion aims of the health support model were fully endorsed by hard-pushed health visitors working on the ground. (In addition, it does not take cognisance of the existence of vulnerability within other professional groupings. Domestic abuse, for example, is as prevalent within middle class women as it is for those living in deprived areas.)

Some health visitors reflected upon individual differences between health support workers in an attempt to learn lessons for future recruitment:

'…certainly some of the health support workers have very much fitted more than others and part of that is around, I mean we've been trying to analyse what it is about the ones that are busy that makes them different to the other ones and it kind of boils down to very simple things like their time-keeping is good, they will do what's asked of them and come back and tell you it's done. They are reliable in that they will always do what they say they're going to do or, if not, they will make sure you know they haven't done it and why. They are conscientious, you know, they're very kind of just, personality things almost, you know, kind of characteristics of personality that make them fit into the system. Whereas some of the other ones are a little more laissez-faire about things and 'och, what's time?'

Others offered judgments that did not appear to have been based on personal experience, for example:

'…you will get ones that will look for work to do and you've got other ones you probably put Richard and Judy on in the corner while the baby is sleeping. And I'm not saying that I think that happens with any of the ones we work with, but I think it's just human nature, you just get people who think 'anything for an easy life.'

It is perhaps inevitable that a new group of workers is open to greater scrutiny than an established professional grouping and, that instances of negative working practice unfairly generalise to the wider group:

'… there is a couple of characters who, for whatever reason, things don't work as well as they do with the majority of them. And I think although there are only two of them particularly, it kind of colours the view of everybody else … I think it's that kind of little bit of, you know, a little bit of black ink in with the white and it just makes everything a bit grey.'

'…every health visitor you speak to says 'well I've got a really good relationship' so you think, 'where are these stories coming from.'

This should not detract from the huge progress that has been made in introducing and supporting a potentially vulnerable group of workers to a profession under some considerable change.

3.3.4 Conclusions

In a relatively short space of time Starting Well developed two project teams to implement its home visiting model. These teams travelled a long distance in this time in which they encountered and negotiated a number of barriers.

The health visitors within the teams were encouraged to adopt an evidence-based and standardised approach to their practice. In reality whilst many developed a broader understanding of the health and social issues affecting the families that they were dealing with, the extent to which a consistent approach to the quantity and quality of visiting that was undertaken is less clear. Health visitors continued to argue for the need to practise in an unstandardised, intuitive manner that cannot be entirely reconciled with a more managerial perspective (Robinson, 1995). This has wider implications for the use of standardised family health plans as recommended by Hall (2003).

Over time the project acted as an advance demonstration of the kind of targeted intensive support within the context of universal provision that is recommended by Hall (2003). Once again, the need for flexible approaches to identify vulnerability was recognised over the course of the project.

In considering the success of the two project teams in influencing their wider LHCCs, context is everything. The projects developed in very different ways as a result of their individual team members, the dynamics between them and the organisational culture within which they were located. Those charged with mainstreaming Starting Well practice require to be alert to the potency of context. In addition, the degree to which the Starting Well model will be 'diluted' through integration with GP practices will be dependent on the project team members, the GP practices and the wider LHCC. These same factors will impact on whether a more discrete team of skilled workers manages to broaden the practice of other staff within an LHCC.

In relation to skill mix with nursery nurses and community support facilitators, the project illustrated some of the possible ways in which different professionals can work together to tackle need. However, a lack of role clarity remained within the project teams by the final year of Phase I. More explicit support mechanisms for nursery nurses were required. Health visitors believed that they had done far less of a community support nature that they had originally anticipated but many were not concerned by this and had come to the view that this was not an area of practice that they had a particular wish to develop. This has implications for ensuring that community development approaches to child and family health become part of routine primary care (Hall, 2003).

A key element of the Starting Well home visiting model was the employment of health support workers from the local communities. These workers developed very diffuse and flexible roles that varied between the individuals; different health visitors also deployed them in different ways. The lack of role clarity did not appear to be problematic for the support workers themselves although some health visitors found these variations difficult to manage. A specific concern arose about the extent to which some health support workers were able to maintain professional boundaries in their relationships with service users as a result of their access to the real life complexity of families' lives. Reflective practice and supervision should perhaps be an explicit and continuous part of the support that such workers receive.

In general health visitors were positive about the role played by these additional workers and, in particular, the part played by those support workers that they worked most closely with. Nonetheless, stories of negative experiences with health support workers were told. It is important that, whilst lessons are learned from critical analysis of incidents that go wrong, individual stories are not generalised to a wider group of workers.

A further layer of innovation in the health support worker model was that the staff were employed, trained and managed by One Plus, a voluntary sector organisation. This offered the workers support from an organisation with expertise in employing those re/entering the labour market and was part of the project's strategy for tackling social inclusion.

Perhaps inevitably this dual management structure led to some difficulties at an operational level with staff perceiving there to be a distinct lack of clarity over the day-to-day management of caseloads. This became a major stressor for many of the health visitors and senior managers attempted throughout the project to resolve this issue. However, those at a strategic level acknowledged that the close involvement of a voluntary sector organisation allowed a supportive work culture for the local workers that would have been inconceivable within the NHS. The problem remained however that health visitors had viewed the support workers as a means of alleviating their caseloads and instead found themselves with a significant supervisory role. More work at the beginning and throughout the project in ensuring health visitor ownership of the social inclusion aims of the support worker model may have been valuable. This issue will become increasingly salient for public sector organisations as they devise ways of expanding the natural constituency of their workforces.

3.4 Delivering change at community and strategic levels

3.4.1 Introduction

Starting Well was conceived as a project that would impact not only at the level of individual child and family health but also at the level of community and organisational capacity to respond to the health needs of local families. In assessing its impact at these levels this section of the report considers two sets of issues: the degree to which community infrastructures and supports have been strengthened through Starting Well activity; and the extent to which the project is perceived to have impacted on the strategic planning and provision of child and family health services.

As outlined in Sections 1.1 and 3.1, a rational model of change was assumed whereby:

• Step 1 - the process of intervening intensively with individual families would result in an understanding within project teams of key community health needs;
• Step 2 - a shared understanding of these needs would feed into local implementation groups; which
• Step 3 - would develop and support local, community solutions; that,
• Step 4 - if unable to be resolved locally would be referred to the project steering group; which
• Step 5 - would act as an advocate for more strategic, Glasgow-wide solutions.

The various elements of this model of change are considered in turn.

3.4.2 Community Support

3.4.2.1 Introduction

In this section we look at the processes that were established to encourage the identification of community level health needs and describe the types of action that were taken to address these.

The project teams were able to describe a range of health issues that they believed to be salient at a community as well as an individual family level, a full list of which is provided below in Table 3.4.1.

Table 3.4.1. Identified community health needs

Two of these, head lice and domestic abuse, are used as illustration throughout this section. These two issues were selected to provide examples across a spectrum from a time limited local level initiative to a more endemic social determinant of health and well-being. Tables 3.4.2a and 3.4.2b below provide a background to these two very different issues.

Table 3.4.2: The background to the selected health issues

3.4.2.2 The Process of identifying community health needs

Members of the home visiting teams described four distinct ways by which needs were identified:

• Through home visiting individual families;
• By communication (both formal and informal) within the team;
• In collaboration with other agencies; and
• Through a pre-existing broader knowledge of community needs.

The home visiting teams identified community health needs through their engagement with individual families and the recognition that problems experienced were not unique :

'As a team we acknowledged … it wasn't just one individual.'

The subsequent communication of these needs within the teams was then important as it was through this that issues from an individual's caseload were shared with other team members. Neither of the home visiting teams used a formal written system to report needs and communication appeared to be largely oral. A pro forma had been designed but had been utilised so infrequently that its use had fallen away:

'I think that system only lasted a month or two and was used two or three times. I think what actually that enabled us to do was to kind of bypass having to write it on a piece of paper. And it's now happening verbally that the staff, instead of writing it on a piece of paper, give to me or [the Community Support Facilitator] or just come and say to me or [the Community Support Facilitator].'

Members of one home visiting team discussed their caseload informally with colleagues and realised that the needs they encountered were also present in their colleague's caseloads. They raised issues with the health visitor co-ordinator as part of their regular clinical supervision and during Starting Well and home visiting team meetings:

' So we've tried to kind of do it so that once a month there is a whole team [meeting] and then the other point in the month, it's just kind of health visiting business.'

In the other team the situation was different, with regular meetings having been disbanded:

'We used to have team meetings every two weeks, which was fantastic, and I was able to tell what's going on, because there was always new stuff on the agenda and so forth. They've stopped unfortunately.'

The ensuing lack of communication within the team was described by one health visitor who gave an example of not knowing that a particular group had been set up:

Health Visitor 1: 'I mean there was an Asian Woman's group started up…I was one of the Health Visitors in the project and didn't even know anything about it starting up. Even as a team there is people taking their own wee chunk and not saying, 'Well I am doing this, what does everybody think of it?' because then they want to buy into it.'

Health Visitor 2: 'But that has always been a problem. We have all been doing wee bits of work...'

Health Visitor 1: 'People get very protective of their bit.'

This protectionism of roles was reaffirmed when a health visitor spoke of how she did not want to pass things to the Community Support Facilitator (CSF) for fear of losing her input into them:

'I was just like saying that weaning is one of the things that I can do and I don't want anybody else to take that away.'

This protectionism was symptomatic of an early lack of team cohesion in one of the teams.

In addition to those family health needs identified through their own case-work, some community needs were communicated to the Home Visiting Teams through their liaisons with other agencies:

'…it was not just within the health visiting team, it was through networking with other agencies and everything.'

This contact was either through attending multi-agency fora, such as the Easterhouse Family Forum or the South East Glasgow Domestic Abuse Forum, through multi-disciplinary working on specific projects with agencies such as nurseries, LHCC personnel, and other community groups concerned with child and family health issues. One CSF, in fact, felt that most of the needs she was working to address had come to her through other agencies:

'When I was working with those three or four agencies, we had a development day in Toryglen and one of the priorities that came through from that was that we had to have a Children and Families Development Worker… we put together a bid to the Toryglen SIP and we were successful in getting 37,500. So as of today we now have a Children and Families Development Worker working in Toryglen which is great.'

In addition, the CSFs appeared to have made good links with other agencies in the community as well as within the LHCC, for example, the Public Health Practitioner.

Those home visiting team staff that had either previously worked in the project areas or were resident there drew on this experience to inform them in identifying community health needs. Health support workers in particular felt that, as many of them lived locally, this enabled them to have specific knowledge of the issues facing the community. This supports the view that a key strength of the health support worker model was the local knowledge of the employees. One health visitor had over two decades experience working in the area and commented on the long standing health needs she was aware of:

'The issues that we have spoke about in Easterhouse are the same issues that were here 20 year ago when I worked in the community, so there hasnae been a lot of changes, the only changes I can see are in the housing.'

Once individual level needs had been aggregated into community health needs the communication of these by the home visiting teams differed across the two areas. This appeared largely to be influenced by the communication structures in place and the relationship between the CSF and the home visiting team members.

The first port of call for the south home visiting team members was either the health visitor co-ordinator or their local LHCC based Public Health Practitioner. They were aware of the latter's plans for community support work and wished to avoid duplication of work. In the east their first point of contact was generally the CSF or health visitor co-ordinator, unless they felt it was an issue they could deal with independently, for example, setting up a mother and toddler group. The nature of the need influenced the person to whom it was communicated. It appeared to both health visitor co-ordinators that if a need was perceived by the health visitor or home support worker to fall into a more medical definition of health it may be more likely to be passed to them, for example mental health, whereas more social health needs would be passed to the CSF or Public Health Practitioner. Viewing mental health as a medical issue, however, runs counter to the social model of health that Starting Well was seeking to promote.

One home visiting team did not appear to have as much communication with their CSF. The team, for example, spoke of not having been consulted on community health needs:

Health Visitor 1: 'Health Visitors have got local knowledge, we should know and we've got good links but I don't think she's ever spoken to us.'

Health Visitor 2: 'She goes on about the LIG and 'Come to that' but it's a cop out.'

Here the CSF felt that the home visiting team was too busy with the demands of their caseloads to pass on community needs:

'I'm not bombarded with needs because they are too busy just getting on with their caseloads and just doing the health visiting.'

The problem of communication between the CSF and the home visiting team was a two way problem, the CSF felt that few needs were passed to her and thus successfully focussed her work on developing wider community involvement.

In the other area the situation was different with team members saying that most people had communicated a need to the CSF:

'Well [the CSF] chairs the LIG … I think we have all given her topics to raise at the LIG, that they can then discuss among themselves and decide if there's something, from their point of view, that they can take forward.'

3.4.2.3 Local implementation group

In this section we look at the views of the home visiting teams, local implementation group (LIG) members and strategic stakeholders on the impact of these groups in addressing local child and family health issues.

The focus of the community capacity building element of Starting Well was the LIG. This was a group with membership drawn from local community agencies and its role was to discuss community issues, support the development of community infrastructure and had a ring-fenced budget to provide small amounts of funding to applicants from affiliated local community organisations. The LIGs were initially intended to have input from statutory agencies such as the Social Work and Housing departments although there was no representation provided. All local community agencies providing services to children and families were invited to apply for affiliated status and encouraged to apply for funding. They had a role in influencing strategic change, passing information about aggregated community needs on to other higher-level groups such as the Starting Well Steering Group or other more specific strategic agencies such as the Greater Easterhouse Community Mental Health Forum.

The community support element of Starting Well had the aim of strengthening community links and capacity building through both local and strategic level development. Whilst the area of community development is a long contested area with many differing opinions on what its aims are and what activities can be considered meaningful community development, it is generally accepted that the key aspect is the involvement of community members in the planning, development and delivery of services. Using the models of community work as defined by Popple (2001) the closest definition to the Starting Well approach is that of 'community organisation'. This entails improving co-ordination predominantly between service providers with the aim of avoiding duplication and tends to be service and project orientated. In contrast Popple's (2001) definition of a 'Community Development' approach centres around assisting groups to acquire skills to enhance members' quality of life and to aid their participation. Sullivan and Skelcher (2002) describe the requirement of full involvement of service users as an essential ingredient of community collaboration and capacity building. The element of citizen or even service user involvement in this aspect of Starting Well had not met initial expectations with participants talking of the dwindling interest of parents in the strategic structures put in place. Only one of the LIGs had a parent representative; it was felt that meetings were not appealing to parents who had fallen away as the groups had progressed. The members were confident in their own abilities to identify family and community health needs and bring them to the table for discussion without the presence of parents.

As the LIG was seen as being the focus for tackling community health needs, the perceptions of the home visiting teams and how they communicated with it was an important factor. The home visiting teams appeared to have a good understanding of the function and purpose of the LIG, particularly in one area where they were positive about it as a forum to which community needs might be taken:

'I think probably having the LIG we are probably in quite a good position in that there is actually somewhere that you can take things like that to. Because if it wasn't for that you would probably be a bit of a lone voice.'

Despite this the health visitor co-ordinator in this area was less positive about the group, seeming unsure of its role:

'I've never been particularly clear myself exactly what should go there and what to expect from that group, so it's kind of difficult to say how I feel it's gone.'

The LIG was most closely aligned with the CSFs in both areas. Both had chaired the group at some point and were responsible for feeding back identified needs from the home visiting teams to the group. The CSFs both saw the LIG as developing more in the direction of an awareness raising and campaigning role as opposed to merely a funding body:

'I do think it's fulfilled its potential, but I think we can still do more …We are not quite sure, are we a funding body just to identify funding? I think it should be a cocktail actually, to be looking at wee bits of funding, to be campaigning or raising awareness around issues that don't get [raised] anywhere else.'

The other CSF had a similar view but saw greater links with the Steering Group:

'We are there to integrate the home visiting element into the project, we are also there to integrate the agencies that are out there into Starting Well. We are also there as a lobbying body, if you like, for the Steering Group, we are also there to fund local projects.'

The health visitor co-ordinator in this area felt that continuity within the LIG had been difficult to establish due to changing personnel:

'I think it's probably a group that does need to continue and evolve and to attract new members to be affiliated with it… initially we did consult with all the pre-five agencies and pre-five type services. But as they change or as personnel change, it's quite difficult to kind of maintain that.'

It was a different situation in the other area where a core group had emerged:

'…In terms of working relationships within the group, it's a good group and there is a fairly steady kind of core membership that attend it and I think that's been good.' (Health visitor co-ordinator)

The views of LIG members were also explored and the issue of funding and community links raised. One member felt that the group was serving a useful purpose by having a budget to distribute to local agencies:

'I've not gone to every meeting, but the meetings that I've gone to, I thought have been reasonably effective in terms of passing out the money and the funds to needy organisations, local organisations.'

The fact that people felt unable to attend all the meetings was commonly described by members of the LIG who felt that they had many commitments attending other fora, often as well as casework:

'I've been a member of the LIG but I actually have seldom attended and that's just been due to other priorities and other commitments.'

The difficulties health visitors had attending was also remarked upon by several participants, echoing the comments of the home visiting team:

'When I first started the first two meetings the first two Child and family sub groups there were health visitors and health support workers at them but there has not been any attendance since then. So they were reporting from their end but I don't know whether they've been too busy to attend.'

LIG members generally agreed that they would bring community needs to the group but not exclusively, as they often sat on other fora and would communicate needs to whatever forum they attended, as did the CSFs who had linked in with other structures in the communities, for example, the SIPs. LIG members seemed largely unaware of where needs from the LIG would be passed on to, seeing themselves as predominantly an operational group who could initiate and fund pieces of work or use their membership of the group for information sharing.

'See it's quite operational this group but I don't know… I mean I would imagine it [a need] would be passed on.'

'I think groups tend to work in partnership with one another and I feel if there's something that you want to bring forward you can bring it to the group and if there's anybody there can help out or bring in some other information, well that's kind of the way that I've been using it.'

The membership from local community organisations had led to a perception of strengthened links horizontally across the community organisations:

'Because they've met more often and at least know who each other are, sitting across the table.'

A survey of community agencies found that over half of the respondents across both areas felt that Starting Well had helped develop multi-disciplinary working and increased liaison and networking which is reinforced by the opinions of the LIG members. Despite these successful local links, there was little vertical communication with minimal awareness of the connection with the Starting Well Steering Group among the LIG members not employed by the project.

The perceptions held by more strategic level stakeholders were also explored and the following issues raised: the extent to which the LIG had linked in with other stakeholders; and, its operational rather than strategic focus. The intended role of the LIGs was described by one participant who had a strategic role both within the Local Authority and Starting Well:

'…With the LIGs the hope was they might be the reformative base for some general linkage in children's services planning terms between the local community, parents and local workers and the broader picture. That theoretically has happened but I don't think it's right, I think in the way issues immediately come to mind when you raise that question.'

As with the LIG members' views of the group, the strategic level participants also saw the LIG as having a more operational than strategic focus:

'I think the implementation group was a forum that was started by Starting Well and **** in particular has taken it forward. And that has in itself become a strong forum, [it] has developed and taken forward a number of issues and made links. So it's been a bit of both, it's been about yes Starting Well linking in, but also being able to do stuff itself.''

The LIG was seen by the strategic level personnel at the highest strategic level, the Scottish Executive, as being a forum for communication and capacity building between local agencies and not primarily as a funding body:

'My understanding is that it's not a huge budget which is quite good. Because I think often things that affect communications do not take a lot of money is more about this perception that money is needed… it's generally about the will the will to make things work and often that comes from the very fact of bringing people together. I think the money side of it, if there is money, it can take away from it. The positive side of it is that it means people don't feel they have to come with a pot of money, they can come and look at the best way of achieving something, the negative side of it that can be the only incentive for people to participate.'

Not having partnerships over-reliant on funding is described by Stewart et al (2003) as fostering a 'can do' spirit, with additional resources being seen as encouraging complacency. Strategically the gathering of individual members of organisations together was seen as the most important factor in generating capacity for agency involvement.

3.5.2.4. Actions taken to tackle community health needs

Overall it appears that, notwithstanding the issues raised earlier, the LIGs made their mark within the local community with regard to membership and providing funding and support to member agencies. The strategic work that had been carried out had been very much on a local level with the examples of lobbying which mainstream services in the east and entering into partnership and collaborative working with wider agencies and professionals, for example, the Public Health Practitioner in the south. In the east, action was taken to influence a public transport provider when this had emerged as a problem for families in accessing healthcare at hospitals and travelling on buses with pushchairs. Additionally the CSF had approached the developers of a new shopping centre to raise the issue of family friendly facilities being included in the plans. In the south the CSF and bilingual worker had liased with personnel within the LHCCs on initiatives such as postnatal depression screening for minority ethnic women and promoting breastfeeding. The survey of community organisations (Berzins et al, 2004) found that 70% of respondents felt that Starting Well had had a positive impact on the community.

The impact of the project within the community was seen a wide range of stakeholders as having achieved three aims: taking health needs from parents to the LIG; providing resources and support for existing community organisations; and entering into partnership with them to take forward new initiatives:

'Yes I mean certainly what Starting Well has brought is an addition in terms of resources, in terms of input it's a more intensive approach to working. It's one of its great strengths I think is the way that it's been able to engage with and involve the local community, involve local parents both through the home support workers but also by from being based within the Ruchazie Family Centre and that whole development, I think that the community feels a sense of ownership around that.'

There was a sense that collaboration and successful partnership working had been achieved in some instances at a local level within the community organisations.

Taking the example of head lice, see Diagram 3.5.1, the LIG had passed the issue on to their pre-five sub group, which takes forward short term working and which co-ordinated the 'Bug busting' programme (this method aims for a systematic removal of head lice by combing hair to remove lice). The initiative was funded by the LIG in partnership with nurseries, to take the information into the nursery schools and groups where parents were already meeting and provide training for community nurses and other personnel. It provided information leaflets, posters, an educational video and distributed fine-tooth combs for the identification and removal of head lice. The aim was to make to the delivery of information non-threatening so that parents did not feel they were being judged or lectured.

Figure 3.4.3 Bug busting initiative pathway

It aimed to encourage collaboration across agencies and disciplines. The CSF felt that this was an issue that only Starting Well would have raised:

'We are raising issues that nobody else is, nobody, no other groups are interested in head lice.'

The sustainability of the initiative was described by a nursery head viewed the initial resources from Starting Well as pump-priming:

'I feel that Starting Well gave us the impetus to go and do it and we will just carry it on, so we've got the basis there, all we need to do is top up with the combs.'

The nursery felt that the education they had provided to parents would stay with them and they would continue to use the resources with the new intake.

Providing a co-ordinated campaign across all agencies was perceived by some to have had the effect of changing the practice of mainstream agencies. However, as a professional with responsibility for joining up services and providing an integrated approach, one strategic level participant had concerns about the fact that the 'Bug busting' was a time limited initiative and as can be seen in Diagram 1 did not 'join-up' with existing initiatives:

'…it needed to be sustained and in terms of where does it link with the community pharmacy head lice initiative that's happened… how do we kind of tie up all that kind of stuff really? It was a January thing and that's it finished…. '

A further Glasgow-wide initiative originated by the Health Board is to follow this later in the year providing access to medicated shampoos and lotions for children and families. This is a different approach to 'Bug busting' and it remains to be seen whether this conflicting advice about how to deal with the problem causes confusion with parents' understanding of how to tackle it. It seems that although the initiative was deemed a success by those most closely involved, there was more scepticism from those within the wider service community concerning integration within mainstream provision.

From the other end of the unmet need spectrum, figure 3.4.4 illustrates the approaches taken and barriers encountered in tackling domestic abuse within the project.

Figure 3.4.4 Domestic abuse strategic pathway

Despite recognising the issue as a problem and, despite training in recognising and dealing with domestic abuse, Health visitors often felt unable to cope with it within their caseloads. Instead they felt that they should refer women to Social Work but perceived that this did not meet women's needs:

'It's strange though because you feel because if you go to Social Work about domestic violence then they haven't got the capacity to take that now. So all they are doing is handing a leaflet that we have already handed this girl and saying 'Well there is counselling services there is you can access them but they have got a waiting list of about six months, so it's really frustrating…'

The home visiting team, therefore, did not all feel confident to deal with women that were experiencing domestic abuse and some did not feel that training that they had received was an adequate response to the problems they were encountering. Many felt that they would benefit from further training and, in particular, multi-agency training (Womens' Health Team, unpublished Starting Well Domestic Violence Study). Some health visitors believed that there should be specialist services that they could refer on to. This challenges the conclusions of Abbott and Wallace (1990) concerning the current readiness of this group of professionals to recognise and address domestic abuse.

On the other hand there was a specific response at project level to the needs of women from the minority ethnic communities experiencing abuse. The bilingual worker arranged an ad hoc counselling and support service from a specialist agency for these women and made close links with a Muslim-faith based agency that provided support for Muslim women as well as joint training for agency staff.

For issues of abuse within non-minority ethnic women the home visiting teams relied on a strategic network already in place and were represented on the local domestic abuse fora.

[there is] 'A domestic violence forum that we go to with that.'

Members of the home visiting team were, however, dissatisfied with the information coming back from these fora:

'What was the outcome of domestic violence forum? There is NO communication about it as well because you don't know what's going on, there might be something going on but we don't know about it.

I feel, I don't know, I feel it's quite often that there are a lot of groups working on these things and then there is no end to it.'

The forum was one of a number that covered the city of Glasgow. The chair of each forum attended a citywide group which in turn fed into a Scottish Executive forum hypothetically providing a link from the local community to the Scottish Executive. The local fora themselves had proved problematic; one forum co-ordinator described the lack of consistent attendance by members due to the priority service users were given over attending meetings or the high turn over of staff within the member organisations. This lack of attendance of forum members meant that the co-ordinator felt this could be off-putting to those who did attend due to the repetitive nature of the meetings. The actual activities of the Domestic Abuse fora were perceived to be limited:

'It's been more of a talking shop really.'

The groups had carried out information sharing exercises about local services and produced leaflets about what was available for women locally but was not perceived to have influenced strategic decisions. The co-ordinator of the forum at Glasgow South East level was asked how much influence she felt it had had over policy either at city level or Scottish Executive level and whether its members knew of the strategic pathway the forum provided:

'Yes they have been informed about the link in fact … I think there was a wee chart that the Violence Against Women Partnership did and I gave out to everybody.'

In this case the actual structures were in place for ground level professionals to pass needs in their communities upwards as far as the Scottish Executive. Diagram 3.5.2, however, shows how this process had become fractured at the first level due to the problems in the attendance and membership of the fora with the result that community needs were not effectively communicated. Hague (1998) found that nationally multi-agency fora for domestic abuse were at risk at becoming a smokescreen for any meaningful activity and their usefulness in information sharing and providing directories of services were time limited; there is some support for this analysis within the Glasgow example.

In general, apart from those directly involved in Starting Well strategic stakeholders struggled to identify what the project had achieved at a community level and what, if any, community health needs had been communicated, Often saying that they were 'too far away' or that they could not remember. Only one strategic level participant could talk about something new she had learned from the project. She felt she had gained some insight into provisions for minority ethnic groups:

'What people had thought was 'provide a clinic or a centre for them to come to and it will be fine, they can come to the centre and get everything'. But actually the centres weren't getting used, that wasn't working.'

In general when strategic level personnel were asked about what information they had gained from the project it was vague and focussed very much on the home visiting element and the 'hard' outcomes such as the six-month findings. The community element was seen as something that they knew was going on and that there had been problems with implementation due to the early imbalance between home visiting and community development.

For those working at a local level but not within the project itself, the general consensus was that the project had not raised any issues that were not already known about:

'To be honest with you I don't think Starting Well from where I sit has uncovered anything that you wouldn't have known through common sense anyway. That some of the vulnerable families have huge problems with parenting and childcare, and if we did something about it it's probably a good idea.'

This feeling was echoed by a participant and LIG member from a local voluntary organisation:

'I mean in some ways I don't think that Starting Well has done anything radically new to be honest, I think what they've done is good. I don't think that they've invented something that didn't exist before, I think they've taken good ideas from a number of different areas and they've put into practice within one project.'

It was, however, seen as having the potential to provide evidence to reinforce these 'common sense' opinions:

'I think something like Starting Well does kind of cement that approach, particularly if it proves to be effective, then it enables us to defend the social model which people challenge all the time.'

3.4.3 Strategic Change Beyond the Local Level

3.4.3.1 Introduction

Having considered the types of community health needs identified through Starting Well and the processes adopted to tackle them at a local level, this section now turns to the role of the steering group as a driver for strategic change and considers the future potential of the project to influence planning for child and family health.

3.4.3.2 The role and effectiveness of the project steering group

As we discuss later in Section 3.5, which discusses stakeholders' views of the extent to which Starting Well met its original aspirations, partnership working at the steering group level was identified as the most disappointing aspect of the project. The steering group was perceived to lack role clarity, ownership and the will to exercise its strategic muscle:

'… we were never very clear what we wanted that group to do.'

'…you see the same people sitting round the table there with actually key responsibilities somewhere else. They should have been able to align these, you know, very powerfully around the starting well table and they didn't. It was a committee meeting that they attended.'

'…now I feel that the steering group might have had a little bit more muscle and I was disappointed that they didn't, … I feel that the people that are there or should be there at the table, are the people who can flex muscles and get things done.'

During its lifespan it discussed the transport issues and lack of baby-friendly feeding facilities raised by the east LIG (neither of which could be satisfactorily progressed due to the lack of statutory accountability of the transport company or the Easterhouse Shopping Centre). Furthermore, although a short-term solution to the issue of women's mental health services raised by the project teams at the beginning of the project was put in place, this was not sustained (Mackenzie, 2003). The literature on strategic change emphasises the importance of vertical connections between levels of planning and implementation (Wilkinson and Applebee, 1999; Chapman, 2002); there was little evidence of this within the structures that evolved.

At the same time the number of issues that came to the group from the LIGs was scant and some stakeholders reflected that the predicted model of issues filtering through the various structures was perhaps unrealistic and suggested that changes were happening locally in ways that did not touch either the steering group or high level individuals within its represented organisations:

'…that aspiration is not being met at all about kind of aggregating need and feeding all that back in, at least it wasn't manifest for us but I'm not sure that it isn't happening at all, in kind of divergent ways out there in the community. But it didn't come through a system that we set up - that's a classic - setting up a system and expecting life to work according to our system …sometimes I'm not sure … throwing a stone into a pond, what do the ripples look like somewhere further down … and we don't always know what those are so we might have had impacts that I'm not particularly aware of.'

3.4.3.3 Working behind the scenes

This notion of progress occurring opportunistically outside the structures established by the project is evidenced by the work undertaken to develop a money advice project in both the east and the south and to expand the full employment initiative already operating in part of Greater Easterhouse to cover the geographical boundaries of the project in the east. The first of these, emerging from the work of the CSFs in conjunction with Glasgow Healthy Cities, resulted in the securing of Scottish Executive monies for welfare rights advice to be delivered within the home as referred by members of the project team. This represents a good example of a community health need addressed within the local areas within the life-span of the project.

The second example on the other hand, the Full Employment Initiative, emerged through a more circuitous route that encountered barriers at the local level, did not come to the steering group for discussion but which, nonetheless, developed into a project that demonstrates the kinds of synergies that can happen when links are made between initiatives. Its genesis is described below:

'…we've always had some extra money floating about … well, I was at a meeting and heard people talk about the full employment programme - it was on the border of where we are in the east but it wasn't in our area … and I simply thought 'wait a minute this is really interesting' … I took it to [csf] and said 'take this [to the local implementation group] because I don't want to make that decision. It isn't my money, I want the community to make that decision' … well, she took it to the group of all professionals who said 'no, not a good idea, we really need to spend this money in this area on mental health' so, I had to go back to [csf] and say 'this money is for this or nothing. It cannot be used for anything else. It's not a pot of money that we are trying to find a home for. It's a specific pot of money and this is what we can do with it. Well, then there was a big hue and cry because it was felt as if I, me, the project was centrally pushing through something that the community didn't want. But it wasn't the community, it was the professionals sitting around'

On the other hand, whilst strategic links were made across projects, a concern is raised by some that the project did not make the kinds of links with mainstream services that would have been required to lead to long-term change. For example:

'…with the women's health team and the children's health team being just across the corridor you might have expected some cross fertilisation about the project … that approach has only been one way and that has been from us to starting well in relation to the piece of work that we have been doing, looking at the response to domestic violence …not enough work has been done to think about, what is the strategic landscape look like and how does starting well fit into it.'

In the last months of Phase I, the project started to move toward a different type of planning whereby Starting Well and Sure Start would develop an integrated approach to tackling the health of vulnerable families. A group entitled Ensuring a Good Start, chaired by Starting Well's existing project manager would be established and linked more explicitly into children's planning services at a Glasgow level. Some, closely involved with this process, believed that this approach would be productive since partnership working around children's planning was viewed to be effective:

'…we've had to work together and agree joint priorities, the children's services planning process has been I think very positive in Glasgow city.'

Others, also closely involved in children's planning were much more sceptical about the degree of drive and partnership working associated with this strategic level planning and viewed Glasgow as illustrative of the crisis in tackling childhood vulnerability identified by For Scotland's Children (Scottish Executive, 2001):

'…the general state of children's services, … as assessed by the Scottish executive through an action committee, is close to crisis … the picture is one of lack of integration, lack of co-ordination and a lack of ownership … so, I think we're not exempt from that in Glasgow … so that's the first thing. The second is then that in the absence of top-level strategic drive 'let's get this right, let's take an integrated approach, let's make Sure Start really help families with hardship, let's break the cycle', in the absence of that, then you have the blockages which is that people only know the boundaries in their organisations and the limits in their thinking that goes along with that.'

3.4.3.4 Starting Well and the Community Health Partnership agenda

As discussed, some of the difficulty for Starting Well in influencing change was the disjuncture between the LIGs and the project steering group. The advent of Community Health Partnerships (CHP) provides a potential opportunity for projects such as Starting Well to bridge this gap and to influence strategic planning at a local level. Their development was seen by stakeholders as an opportunity to impact on the breadth of responses to community health needs and to shift the focus of local health care management away from a narrow GP model:

'CHPs for me are the once in a generation opportunity for the NHS to create structures that are much more directly linked to communities and much more socially orientated and deal[ing] with causes as well as effects. So I'm terribly optimistic about CHPs...'

'LHCCs were never owned by their local communities because they were imposed by the NHS because basically GPs decided how they wanted to group together. So for me CHPs are about more than just LHCCs and health promotion, they're about a real step change in the influence that communities have on the NHS. And actually organising and structuring that in a way that really influences resource and local decision making…So you look at something like child health I would expect each CHP to have a major focus on child health as a kind of core issue.

The potential for Starting Well to influence this agenda was stated by one Health Board participant as follows:

'I think Starting Well should be influencing these big major strategy things in future, what practices look like and I suppose as part of our child health team we should be looking at that… if the whole planning for that community's health including its children's and families health is at that community health partnership level and they are working hand in glove with local authorities with education with social work, there is a real potential and I think it would be daft if Starting Well is not part of that.'

However, one strategic stakeholder was not aware of the project currently capitalising on the opportunities to influence this process:

'I think then Starting Well does have a future across the city but to the best of my knowledge currently Starting Well is not at the forefront of any thinking around development of CHPs.'

3.4.4 Conclusions

There is some evidence that the process of working more intensively with families generated within health visitors a greater understanding of complex health needs that needed to be addressed at a community as well as an individual level. The greater awareness of domestic abuse is a good example of this. Health visitors also developed their picture of community health needs through their own experience of the areas they worked in, through their collaboration with other team members, and through their links with other organisations.

At the same time, it is not surprising that the kinds of health issues that were uncovered were not 'new' in the sense that they were generally issues that were already recognised within the community and the agencies working in the areas.

The processes of discussing shared perceptions of community level need were relatively haphazard within the two intervention areas and their success appeared to be less the function of the existence of formal systems of communication and more about levels of collaborative working within the project teams. The burden on health visiting caseloads, the subsequent delay in establishing the community support arm of the project, and the lack of early clarity in the role of the CSF (Mackenzie, 2003) also served to push this further down the agenda for health visitors.

However, through both the work of the LIGs and of the CSFs in conjunction with other multi-agency fora, there is evidence that Starting Well made an impact on local organisations. The survey of community organisations undertaken jointly by the independent and internal evaluation teams, for example, demonstrated a relatively high level of awareness of Starting Well and its activities.

In addition, the development funds were used to support small scale and short-term activity within a variety of settings (such as nursery schools and play groups). More significantly, in terms of sustainable work, the community support facilitators were instrumental in securing funding from the Scottish Executive for the Money Advice Project in each area whilst the bi-lingual worker made significant advances in developing support for minority ethnic women suffering from domestic abuse and raised this issue within the LHCC ⁴³.

However, there is little evidence that the project steering group acted as advocates for Starting Well. Whilst it was probably naïve, in retrospect, to expect the LIGs to fully find their feet, identify needs and filter those that needed to be dealt with more strategically through to the steering group, those needs that were identified in this way were not progressed significantly.

Key stakeholders and members of the steering group acknowledged that the group suffered from a lack of clarity as to its role, a lack of ownership and a luke-warm approach to partnership working. These features seriously compromise the potential for mainstream developments (Stewart et al, 1999; Sullivan and Skelcher, 2002; Rummery and Coleman, 2003)). Where strategic change began to take place this was behind the scenes rather than around the steering group table.

These issues remain salient if Starting Well is to influence the policy and practice of child and family health within the new CHPs.

3.5 Revisiting the Theory of Change - Strategic Level Stakeholders' Perceptions of Success

3.5.1 Introduction

Having discussed the extent to which Starting Well made an impact on a number of key processes, this section now turns to the question of whether the same group of individuals who articulated the project's initial Theory of Change, viewed the project as having been successful within Phase I (in other words, was the Theory of Change doable in retrospect?). We then consider their reflections on the meaningfulness of the original project plans.

3.5.2 Retrospective Reflections on the doability of Starting Well's Theory of Change

Stakeholders were asked to consider, at the most general level, the objectives set by the project in 2002. Most perceived these to have been met.

'…I'd be surprised if they haven't achieved most of [them]. I know that the community development side was not as quick off the mark as it ought to have been but that was a question of resources and, when the resources were brought to speed, then I think they picked up in that.'

'…if I speak in general terms, I don't think the actual remit and objectives is in jeopardy.'

'…I don't think there is any in the territory of really, you know, … falling off the rails'

The ways in which participants supported their views of the project's success varied depending on their proximity to Starting Well. For those most intimately involved, success was evidenced with performance monitoring data on specific objectives such as dental registrations or direct experience of managing particular components of the project:

'…Well if you look at these for example forty percentage registered with community dentists - we've done all that. In fact we're well in advance of that.'

'…it's been fascinating for me seeing a bunch of paper and outlined plans turn into real people and real initiatives and so on. So my overall assessment is that many of the things we hoped might come to fruition have done so in terms of staffing, teams, relationships, interactions.'

Others at greater distance from the project provided more impressionistic and qualified responses:

'…Um. Well my sense of it is, and there are limits to this because of how close I am to it, my sense of it is that it's met its promise - my sense of it is that in large part it's done what it set out to do. Its reach is evidenced by the fact that people refer to it a lot. So people are aware of it within health and beyond health as being a different way of doing things. So I think on that basis, that's an indicator [but] it's like a half life - once it's radiated so far from the core of experience, you don't know what the value of that visibility is, but from where I sit and what I pick up from social worker colleagues … it has positioned that service differently from what happens elsewhere.'

. 'I ought to be able to answer that question fairly easily. And yet I continue to find that with the exception of your interim independent evaluation, I find it very difficult to get to grips with the areas in which the project has been more or less successful and therefore for that very reason I find it very difficult to answer your question. … (I have) concerns that some of what we are hearing is the policy rhetoric that the project feels is appropriate to feedback to more external stakeholders.'

Stakeholders were then asked to discuss the extent to which the project had met its objectives in the areas of home visiting, community support and partnership working. These are discussed in turn.

Echoing the data reported in section 3.2, the home visiting model employed by the project was believed to have been largely successful at engaging families, developing productive, therapeutic relationships with families and in changing professional practice:

'I see some of the huge difficulties that is attached to actually getting people to change their kind of frames of reference and the way they have always done things, so they have managed to put together a project that does deliver different things and challenges a lot of people's assumptions about the way things have been done in the past and the way that they need to be done in the future and I think it has also managed to engage parents very successfully in that process so I think that the recruitments rates are, you know very, very good and most families have taken part in it.'

'When you go to Starting Well events of any sort and you talk to parents and talk to health visitors you get an amazing, real depth and richness of the anecdotal evidence of what you might call case studies, but which I simply call narrative of people's lives almost transformed by that relationship and the transformative aspects may be little more than the health visitor saying 'go on hen and I'll accompany you to the debt advice centre or the alcoholics anonymous or anything'. I think it might not be major stuff but it has transformative effects because someone took an interest, someone had the information, someone maybe geed them up a little bit to go along, arranged the childcare, arranged the meeting, gave them the confidence, made sure they bloody well kept their appointment and went through with it. The next thing you know they're on their own, they don't need the health visitor any more.'

'I think that as a starter for ten, it's quite useful that the team has produced documents as recommended by Nursing for Health.'

Two issues were highlighted by those most closely involved in the project's development in relation to operationalising the project's plan. These were: the difficulties in establishing a structured and standardised model of health visiting practice; and, the vast underestimation of the time and effort required to integrate a new workforce into an existing professional grouping (again, these confirm the findings discussed in section 3.3):

'I think the biggest frustration in health visiting (and I'm generalising) is they are not really that interested - or there is a distrust of evidence based practice or of evidence. The biggest evidence for them is their experience. And whenever we try and look at standardising or guidelines or protocols, the discussions comes back to, 'but we have to respond to the individual family's needs and that's our professional skill and that's what we think is right'. Particularly we're seeing it, looking at the parenting programme where we have said 'well the evidence is very strong elsewhere for triple P, we want to see that programme used more widely' and they will say 'well that's only relevant for some families - some families prefer another programme or I'm more used to another programme'. And there maybe is, um I suppose but there will be rights and wrongs in both of these. But the sort of cultural move in health visiting to evidence based practice is behind where probably it is in medicine for example.'

'…we have done some interesting things in terms of bring in the health support workers and if we consider the original proposal that was actually a couple of sentences. And it was a much more complicated process to try and do, in terms of beginning and simply trying to get a contract in place. You may recall it took a long time because of going back and forth and legal departments etc but it eventually got there. I think that that programme is beginning now to reach what potentially it could offer … and again it was a huge cultural change to bring people who are non-professionals into a professionally oriented culture very hierarchical culture em and into the middle of a project where people who were supposed to be leading them, the health visitors, were trying to find their own feet at the same time. So it was a big bang approach to try to change services and I think that it is beginning to pay off.'

In relation to the community support aims of the project stakeholders were generally happy that these had been met despite the project's early concentration on home visiting:

'Community support. I think that again through the community support facilitator we have moved a long direction in that … you know, the right road of travel and there are all kinds of things going in the local communities and all kind of local community organisations being brought together and I think this whole use, or this whole development of a local implementation group has been very useful and a specific group that looks at pre five services in a local community - there are other groups, there might have been other ways to do that, when I look back on it but I think that was useful for our purposes. ...having someone with a specific focus on the pre five agenda has proved really valuable.'

Securing of funding for a money advice project through the Scottish Executive was cited as an example of good practice as was the development of a Full Employment Initiative in Easterhouse. On the other hand, for one stakeholder based at a local level, the implementation group, established to tackle child and family health needs, was not believed to have integrated with existing primary care structures:

'If I am being honest I don't think it has been terribly successful. …I don't think there is any issues that they have … brought into my table. But I think that is one of the difficulties with the way project is run - it is seen as separate.'

As discussed earlier in section 3.4, the element of the project that was viewed almost uniformly as having missed the mark in terms of the project's aims was partnership working. The project steering group, which was initially viewed as the means through which child and family health issues that could not be tackled at a local level would be addressed, was disbanded in 2003 and a new structure incorporating Starting Well and SureStart ('Ensuring a Better Start in Life') developed. The critiques offered by stakeholders cover the general utility of steering groups as a model of governance and the engagement of key partners. These are key themes in the literature of partnership working (Sullivan & Skelcher, 2003):

'The partnership stuff, I think that's the area that still requires the most development. And I think we moved down a certain road but …I think we have to question ourselves, not just this project, I think projects generally, the health service generally, social services, we create these steering groups and I'm not sure that - are they good value? I haven't got another model in my head, may I say - I don't know how else we do that. But what … we have here is exactly what we didn't want to have, where we have a range of people, duplication from agencies, not always the right people, people without investment or real ownership of the project. And we have to take responsibility for that as well. Peopleweren't asked often to make very important decisions around the project or if they were, the direction of travel and that decision was so clear that really it did become kind of rubber stamping in some way. So, that didn't really work and evidence of that is that eventually it kind of wound down although want we want to do now is to pick it up under a new banner under the Ensuring a Good Start in Life Group and link it into SureStart.'

'I think coming from a local authority background, my feeling is that the local authority's involvement has been disappointing as well. On two fronts - first of all by its own lights at times I think it has just stood away and said 'ah well loads of Executive money going to an NHS project - that's fine, good luck to you. We haven't got that money to invest how could we? You're lucky - good on you'. I think there has been that, which is a form of institutional miserabilism and which is questionable anyway because the local authority is the housing for a lot of new Executive money anyway under Changing Children's Services and pre-fives and all the rest of it. Secondly, I do feel … because of the concentration of the project particularly in work force issues around a particular area i.e. health visitors, that broader perspective was lost earlier on and my colleagues in the local authority were not as engaged as they should have been earlier on. And I think we have probably lost opportunities to integrate better with children's services planning which we are only now beginning to address.

A further issue raised in relation to both the partnership working and community support aspects of the project concern the extent to which the objectives around these processes were realistic or thought about carefully enough in the planning stages:

'The only bit where it was unrealistic because it was going into unchartered territories was … it was a quite vague on the aim of partnership and collaboration and community development because that begs that bigger question of what is the framework for community development and integrated service delivery. …. So that is like how far you can actually change the world from looking at one corner of it. That's unrealistic.'

3.5.3 Retrospective Reflections on the meaningfulness of Starting Well's Theory of Change

Having explored the extent to which strategic stakeholders believe Starting Well to have met its original objectives for Phase I, we now turn to a discussion of their views on the role of different types of evidence in supporting their perceptions of success ( testability), and the extent to which they were meaningful in relation to the long-term goal of achieving a 'step change' in child and family health.

Despite the fact that the project was encouraged to set itself objectives that included processes as well as harder indicators and outcomes, and that both the independent and external evaluation employed a multi-method approach, a number of stakeholders expressed a concern that, the dominant paradigm for judging success would be that of the medical model:

'I think it's one of my areas of regret that a lot of people have come to perceive Starting Well as an evaluation that is a … clinical trial rather than as a demonstration project. …you still hear the phrase in and around Starting Well, of people observing and then commenting on it 'when does the evaluation come in'. As if there is a single one and that is a dominant deception I would say in classic publichealth, that all the rest is essentially window dressing. … we could end up, and I use the phrase intendedly I guess, throwing the baby out of the bath water, that's my biggest concern. '

In addition, (confused) expectations from project commissioners were perceived to feed into a skewed implementation focus on the objectives that were more readily measurable:

'…in terms of what the Scottish Executive has been expecting um, at times … you've got messages from the CMO … 'you have got to be clear, you've got to stick to these, you've got to get really hard evidence about this, that and the other' and then it's 'no, more process stuff'. Aye well ok - I think again [the project management team] must have been thinking 'I really have got to look at oral health, dental care, and all that kind of stuff. As for community development, well it's airy fairy, it's hard to just sort of explain, … while folk are bursting for the hard stuff'.'

Early evidence from the quasi-experimental arm of the independent evaluation (section 2) that indicated a difference between the intervention and control areas on only a small number of variables (including maternal mental health and dental registration) was discussed by stakeholders. For some the lack of effect size was neither surprising nor disappointing given long-term nature of inequalities and poor health outcomes.

'…you might actually help stabilise a family through intensive input but, in terms of any outcome measures, there may be none. They might not become drug free or you might not help get the kids off the child protection register, or whatever.'

'I think it takes generations to change outcomes and I think that you understand that you only get funding for certain things, so you put outcomes in to get the funding. I don't know if that is realistic given the ambition of what it is trying to change … I think that's more to do with the framework of how we actually develop initiatives'.'

One stakeholder expressed an uncertainty about what it would mean for the project's underlying Theory of Change if there were no evidence of the project's impact as demonstrated by quasi-experimental data.

'I don't think we know yet what that means. Does that mean it's [the 6 month data] too early? Or does that mean, it could be a number of other things. I don't think we know. I don't - maybe I don't have a huge sense of optimism that we are going tosee big differences even in the 18 month outcomes…. I think there will be a number of reasons for that. I think that there are a lot of initiatives going on anyway in terms of the control area. I think it has taken us time to get the project up and running well. And we still haven't got everything perfect but certainly in that first year much of when the cohort was recruited from things were still at a very early stage. Um some of it will be, these are the most vulnerable families that will need a lot of help over and above what we are able to offer here. Maybe some of it is we have got the wrong measurements. I suppose there is a lot unknown here that we need to discuss when we see the results. …So I guess the answer is I'm not absolutely sure yet. But I don't think we are going to see such amazingly fantastic results that everybody says well we've got to put lots more money into following a Starting Well model which is what we need to think about when we look at the next phase'.

The supremacy of experimental evidence in convincing future funders and planners of comprehensive, child health services can, however, be contested. One planner, for example, indicated that the evidence required might be less rigorous than anticipated:

'What I'm relying on is a sell from this project.'

Another stakeholder reflected on the power of 'human narrative' over performance monitoring data in persuading policymakers of impact:

'…it is the sort of stuff where I know it doesn't necessarily evaluate always rigorously well, it is the sort of stuff if you were to say to the minister 'in a nutshell this is what it has done' - 'oh, what have you done' - 'turn lives around'. 'Brilliant. What more do you want? Would you rather have reduced dental caries?'

This leads then to the discussion of how meaningful the project's plans were believed to have been in retrospect. Whilst some stakeholders believed the plans to have been appropriately ambitious statements of intent, some were less convinced that they represented significant inroads into the task of producing a 'step change' in child and family health. One, for example, discussed the lack of early planning around the way in which the project would tackle major issues impacting on population health:

' think some of those, some things I think we could have anticipated and because we could have anticipated, hadn't done as much as I'd liked and I'm disappointed we haven't. And that's I think around what I will call the iceberg effect of finding out much more about women's mental health and domestic abuse issues that has come through the health visitors. Now we damn well should have known that. Could have guessed it and if we hadn't guessed it - there were enough reports out there from Glasgow and elsewhere to have thought this is what we are going to find … we should have maybe just sat and taken some time and thought our way through the implications of what happens if all of a sudden people turn round and say um there is an awful problem about mental health issues and black and ethnic minorities, domestic abuse or whatever.... we should have somehow got culturally into the mind set at the beginning where there will be a certain fix set of stuff that this will be based on … but a lot of it should have been left open … It doesn't have to have a structure plan or Venn diagrams either, you just need to say, as a matter of culture and openness, this is what we will try and do.'

Part of this difference in viewpoint was due to differing perspectives of the scope of the Starting Well project (as one stakeholder noted 'I go back to basic principles …and again, you'll know everybody in the team will have slightly different or a dramatically different take on what they are'). Another component of the disagreement about the meaningfulness of the project's plan is the tension around whether performance targets and objectives are liberating or constraining to projects that seek to adapt to complex policy and organisational contexts (Blalock, 2000). Whilst some stakeholders believed that reaching targets was less important than setting out bold aspirations, others viewed targets as a statement of 'can-do's':

'…some of the stuff around oral health and weaning they are all well and good, but it's a bit flat. … I think that there are other areas where I would have wanted maybe to have set, well what I would regard as more meaningful indicators or challenges, we may have met only half of them or none them and then have had an argument about it but I would rather we did that …I would much rather you kind of aimed for stars and got as far as the moon as opposed to just jumped up and down on the spot. I think there are aspects there where we haven't developed as far as we could have done because we weren't spurring ourselves with that particular perspective.'

'…we [had to] try to be very careful about the size of this stick we were building. …So what we try to do is find something that is achievable, a little bit difficult but achievable that we can work towards … but it's not going to rebound in a way that would be just unhelpful.'

This latter stakeholder then goes on to describe how the relatively unambitious targets within the project plan, whilst not perceived as meaningful in relation to the project's ultimate goals, can become prompts to implementers for activity that would not otherwise have happened:

'I think we could go back to a project and ask what use is it having a three year project plan written - I would say very little because things are going to emerge all the time and having this written down in this kind of concrete way is not probably that helpful. There is another approach however. So I think it is a painful process, probably worth using but not reifying into something other than what it is. It is a kind of guide. But then lets not get carried away with this. It is going to suggest a direction of travel but we might move, based on our experience. We have to be alive to the possibility that a million things are going to emerge here. In the same way, if we don't have these kinds of targets we can go in all kinds of direction so as a means of kind of concentrating thinking, they are probably quite useful. As a means of ensuring some kind of outcome in long term health I am not sure they are particularly useful but what they give if we have to think about ensuring that people get to the dentist we have to think about this whole infrastructure of activity and we have to make links with all kinds of people we wouldn't have made if we didn't have to think about that target. So the value is probably not the value for which it is intended.

For others the energy that was expended on the more easily measurable inevitably detracted from other more difficult objectives:

'…what does a step change actually look like, which criteria do you use, again you're constrained by what you can actually measure and what is considered to be a robust instrument rather than what other things you'd ideally like to measure.'

3.5.4 Conclusions

In relation to the retrospective doability of the project's programme plan, strategic stakeholders were relatively confident that success could be evidenced for much of the home visiting model and, to a lesser degree, the community support element of the project that had experienced delays in its implementation. The achievement (and indeed realism) of the partnership working objectives of the project was much less evident to stakeholders. These findings echo those discussed throughout Part 2 of this report. On the other hand, the findings from the quasi-experimental study of the project suggest that a more tentative view of success is warranted.

The extent to which the achievement of the stated objectives could be viewed as meaningful in producing a longer term 'step change' in child and family health was contested by stakeholders. Short-term health indicators, for example, were viewed as both constraining and useful to project development. This tension is found in many complex initiatives (Mackenzie et al, forthcoming) and signals the need for clearer ownership of and commitment to a model of project planning that acknowledges the need for both monitoring and more creative evolution.