Introduction
The linkages between ill health and weaker employment opportunities and experiences are strong, bi-directional and well-evidenced, whether thinking about unemployment, precarious employment, sickness absence (absenteeism) or reduced productivity (absenteeism) (Eurostat, 2018). Good work is known to be beneficial to mental and physical health (Waddell and Burton, Reference Waddell and Burton2006; Dolan et al., Reference Dolan, Peasgood and White2008; Oguz et al., Reference Oguz, Merad and Snape2013) and the scale of health-related unemployment, absenteeism and presenteeism across advanced economies harms individuals, communities, employers and Exchequers alike (OECD, 2015).
It is difficult to overstate the scale of the challenge and of the support need. The percentage point difference in the employment rate of working age adults with a disability or health condition – compared to those without – is negative, sizeable and stubborn across countries (DWP-DH, 2016; WPSC, 2017; Eurostat, 2018). Of concern is not simply the scale and depth of these negative impacts but also the near complete lack of positive change following at least two decades of repeated employment support interventions across advanced economies aiming to shift this disappointing position (van Stolk et al., Reference van Stolk, Hofman, Hafner and Janta2014).
This is not to say that employment support programmes can resolve this issue alone. Changes in both the casualization and intensification of the labour market continue on the one hand to create a reality where many jobs cannot be considered to be ‘good’. On the other hand the near wholesale policy adoption and then – to varying degrees – simplification of the evidence has led in some national contexts to the hardening of welfare systems that unhelpfully threaten and coerce unemployed individuals with health conditions and disabilities into any paid employment irrespective either of its nature or the individual's need and desires, both key factors to whether paid work is in fact beneficial to wellbeing.
While it is key to recognise the wider economic and social context within which social policy interventions such as these sit, there remains nonetheless a significant need to rethink the nature of employment support for individuals with health conditions such that they deliver a step-change in support experiences and sustained employment outcomes that contribute to positive wellbeing gains and improvements in quality of life. While many interventions have demonstrably failed over recent decades, a significant body of solid evidence has developed across multiple robust evaluations around a model of employment support known as Individual Placement and Support (IPS) for individuals with severe mental health conditions within secondary mental health services (Bond et al., Reference Bond, Drake and Becker2008; Burns and Catty, Reference Burns and Catty2008; Marshall et al., Reference Marshall, Goldberg, Braude, Dougherty, Daniels, Shom Ghose, George and Delphin-Rittmon2014; van Stolk et al., Reference van Stolk, Hofman, Hafner and Janta2014; Mueser et al., Reference Mueser, Drake and Bond2016; Modini et al., Reference Modini, Tan, Brinchmann, Wang, Killackey, Glozier, Mykletun and Harvey2016; Suijkerbuijk et al., Reference Suijkerbuijk, Schaafsma, van Mechelen, Ojajarvi, Corbiere and Anema2017). The IPS model emphasises client preferences and operates a rapid place-then-train employment model in which competitive employment of well matched jobs is the goal, with individuals supported intensively by employment specialists, where the specialists have low caseloads and are integrated into mental health teams.
The effectiveness and cost-benefit of the IPS model is substantial and undeniable, even if leant towards a US evidence base (Bond et al., Reference Bond, Drake and Becker2012a). A major recent RAND evidence synthesis concludes that IPS is the leading policy option that governments should consider to support individuals with health issues into paid employment (van Stolk et al., Reference van Stolk, Hofman, Hafner and Janta2014). Summarising the evidence base across seventeen randomised control trials, Marshall et al. (Reference Marshall, Goldberg, Braude, Dougherty, Daniels, Shom Ghose, George and Delphin-Rittmon2014) conclude that IPS services show a 58 to 60 per cent employment rate compared to a 23 to 24 per cent employment rate for randomly assigned control groups – staggeringly high effects compared to those seen in evaluations of mainstream employment programmes.
Despite its significant potential, however, traditional IPS is a niche employment support model that is – in its current configuration – ill-suited to making significant inroads into the totality of the health-related (un)employment challenge across advanced economies, even if some experimentation with IPS to new cohorts, settings, functions and larger volumes has occurred (Li-Tsang et al., Reference Li-Tsang, Li, Lam, Hui and Chan2008; Coole et al., Reference Coole, Drummond and Watson2012; Ferguson et al., Reference Ferguson, Xie and Glynn2012; Luciano et al., Reference Luciano, Drake, Bond, Becker, Carpenter-Song, Lord, Swarbrick and Swanson2014; Sveinsdottir et al., Reference Sveinsdottir, Løvvik, Fyhn, Monstad, Ludvigsen, Øverland and Reme2014; Ellison et al., Reference Ellison, Klodnick, Bond, Krzos, Kaiser, Fagan and Davis2015; Poremski et al., Reference Poremski, Rabouin and Latimer2017; Smith et al., Reference Smith, Atmatzidis, Capogreco, Lloyd-Randolfi and Seman2017). Specifically, traditional IPS is limited in four key respects:
Cohort: IPS models have supported individuals with severe mental health conditions only, despite the far larger number of individuals with low to moderate mental health issues in need of support and the frequent co-occurrence of mental and physical health issues (comorbidity);
Setting: IPS models have operated in secondary mental health services, but only a small minority of the wider cohorts identified above are (rightly) ever supported in such services;
Function: IPS services have focused exclusively on the out-of-work cohort, despite the significant support needs for those already in employment around absenteeism (e.g. sickness absence) and presenteeism (e.g. reduced productivity);
Scale: IPS services have operated at low volumes, in contrast to the far larger size of the challenge presented by mainstream low to moderate health cohorts in need of similarly effective employment support.
Alongside the interest in what we will term traditional IPS, there is therefore an understandable significant parallel policy interest and activity around: flexing the IPS model into what we term modified IPS services in the face of this combination of impressive IPS evidence; the pressing need to do much more; and the inability of the IPS model in its current form to bridge the gap. While needed and welcome, therefore, the flurry of newly emerging funding and activity around modified IPS services is at present also a source of risk to the service user experiences, outcomes and value-for-money of modified IPS services. This is due to the absence of a well-considered analytical framework within which commissioners and programme designers are enabled to reflect critically about the nature and resulting expected effectiveness of the modified IPS programmes that they are creating. The original development of that critical analytical framework is the chief task of the discussion that follows, informed by the author's embedded policy experience codesigning one such modified IPS service.
Critically reflective codesign of a large-scale modified IPS service
Since autumn 2016 the author has been seconded to work with regional policy partners leading an intensive programme of codesign work of a large-scale modified IPS trial within a UK Combined Authority – a tier of devolutionary government comprising multiple local authorities across a city-region geography.
Sponsored by the UK central government joint Work and Health Unit (WHU), the shared opportunity and challenge has been to design, commission and mobilise an innovative modified IPS trial to support around 3,750 residents with low to moderate mental health and/or physical health conditions in primary and community care settings and incorporating both those who are out-of-work and seeking employment as well as those who are in-work but off-sick or struggling at work due to their health condition. Voluntary in nature, the modified IPS trial will run between 2018 and 2021 within a randomized control trial (RCT) framework. It is to our knowledge the world's largest IPS trial and will produce a wealth of evidence of unprecedented depth and richness about what works best for whom and why (or, equally valuable, why not).
The codesign thinking took place weekly for a roughly twelve month period from Autumn 2016 to collectively develop the detailed design of our modified IPS service such that it satisfied three key criteria: firstly, that it incorporated evaluation learnings from the IPS literature around the links between fidelity, key characteristics, and performance; secondly, that it worked appropriately for the trial's differing target cohort, taking into account during the codesign process the views of service users, health practitioners and employers in a series of local stakeholder workshops; and thirdly, that it mapped effectively onto the local health systems of the five local areas involved in the trial such that it would both work effectively in live-running and be able to be sustained afterwards. Alongside the author as codesign lead, the codesign group comprised a local lead from each of the five areas involved and central government colleagues from the joint Work and Health Unit. These local leads were key to discussion and also acted as a conduit to further consultation to wider stakeholders in their local areas between codesign meetings. Local leads all worked around the work-health agenda within their organisations with a constructive mix of roles and organisations – some senior public health leads, some senior health commissioners, others employment leads.
Central to the effectiveness of IPS is the twenty-five item fidelity scale shown down the left column of Table 1, against which IPS services can be measured in terms of their fidelity and quality (CMH, 2015). For traditional IPS services the fidelity scale is a powerful guide to designing and delivering high quality IPS services and performance. Given the activity, trialling and effectiveness of IPS services over the past twenty years, and given the central place of IPS fidelity to the IPS model, the IPS fidelity scale has understandably been the subject of significant attention within the academic literature in a range of ways including: the creation and evolution of the IPS-25 fidelity scale (Bond et al., Reference Bond, Becker, Drake and Vogler1997; Bond et al., Reference Bond, Campbell, Evans, Gervey, Pascaris, Tice, Del Bene and Revell2002; McGrew and Griss, Reference McGrew and Griss2005; Bond et al., Reference Bond, Peterson, Becker and Drake2012b); the link between employment outcomes with the total fidelity score (Bond et al., Reference Bond, Peterson, Becker and Drake2012b; Kim et al., Reference Kim, Bond, Becker, Swanson and Langfitt-Reese2015; Lockett et al., Reference Lockett, Waghorn, Kydd and Chant2016) and, as important but less commonly, individual fidelity items (Bond et al., Reference Bond, Peterson, Becker and Drake2012b; Kim et al., Reference Kim, Bond, Becker, Swanson and Langfitt-Reese2015; Margolies et al., Reference Margolies, Humensky, Chiang, Covell, Jewell, Broadway-Wilson, Gregory, Scannevin and Dixon2018); sub-dimensions of the fidelity scale (Bond et al., Reference Bond, Campbell, Evans, Gervey, Pascaris, Tice, Del Bene and Revell2002); differing levels or types of fidelity scale (Mowbray et al., Reference Mowbray, Holter, Teague and Bybee2003; Bond et al., Reference Bond, Becker and Drake2011); fidelity assessment and self-assessment (Bond et al., Reference Bond, Becker and Drake2011; Margolies et al., Reference Margolies, Humensky, Chiang, Covell, Broadway-Wilson, Gregory, Jewell, Scannevin, Baker and Dixon2017); and the moderating role of individual and contextual factors between IPS fidelity and employment outcomes (Campbell et al., Reference Campbell, Bond and Drake2010; Metcalfe et al., Reference Metcalfe, Drake and Bond2018).
Traditional IPS fidelity scale and our trial's modified IPS fidelity scale
Of particular relevance to the present argument, modification of IPS fidelity and services has also been a lively area of activity and discussion in the literature (Paulson et al., Reference Paulson, Post, Herinckx and Risser2002; Mowbray et al., Reference Mowbray, Holter, Teague and Bybee2003; Bond et al., Reference Bond, Becker and Drake2011; Luciano et al., Reference Luciano, Drake, Bond, Becker, Carpenter-Song, Lord, Swarbrick and Swanson2014), whether that augmentation be to add additional programme elements felt to be potentially useful – for example, a growing evidence base around the benefits of including additional cognitive remediation elements (McGurk et al., Reference McGurk, Twamley, Sitzer, McHugo and Muser2007) – or else to remove or dilute programme elements in order, for example, to reduce costs – for example, evidence that IPS time-limited to nine months rather than being time-unlimited, as per traditional IPS fidelity, may be virtually as effective but markedly more cost efficient (Burns et al., Reference Burns, Yeeles, Langford, Vazquez Montes and Burgess2015). While many wish to modify IPS to seek to harness the effectiveness of traditional IPS in new contexts, cohorts and settings, given that the traditional IPS fidelity scale is well-validated, linked to more positive outcomes and a powerful tool to drive model adherence during implementation an alternative strand of the literature rightly argues that augmentation away from the evidence-based practice of the IPS fidelity scale must proceed with caution (Bond et al., Reference Bond, Becker and Drake2011). Where cohorts, contexts, functions and volumes vary from traditional IPS services as in this policy case study, however, modification may be necessary and appropriate. The key challenge then is to effectively flex the fidelity items as required for the differing needs of their larger and wider cohorts and primary care contexts while retaining the core ethos, quality and performance of the IPS. If, why and how to augment IPS fidelity then, and with what implications for this particular policy case study?
The right hand column of Table 1 shows the fidelity scale of our codesigned modified IPS service based on its differing cohorts and needs and our wider local system. No fidelity items are added or removed and, while most remain identical to the traditional IPS fidelity scale, several are adapted for our intervention. Reflecting on the evolution of the IPS fidelity scale itself, Mowbray et al. (Reference Mowbray, Holter, Teague and Bybee2003) outline distinct empirical, practical and theoretical approaches to fidelity modification. Our approach to the modification of the IPS fidelity scale through the codesign process combines practical and theoretical approaches – a practical flexibility to bend certain items within limits to fit with the differing needs of our intervention (cohorts, functions, local health and employment system, etc.) compared to a traditional IPS service, while retaining a theoretical commitment to ensuring that the programme theory and ‘cognitive blueprint’ of IPS are retained. Given that this programme is a randomised control trial where programme fidelity, programme effectiveness, and the links between the two (at both total scale and item-level) will be evaluated, it will in time also be possible to conduct empirical assessment of these modifications.
Table 1 highlights that there is significant overlap between the two fidelity scales despite the differing cohorts – and resultant service needs – of a traditional IPS service and this codesigned modified IPS service. We term these ‘standard IPS fidelity items’ and these are shown in Table 1 as items that are – and ought appropriately to be – shared items of equal relevance across both traditional fidelity and modified fidelity scales and according to the same underlying measurement and scoring criteria. The challenge for modified IPS services is simply to deliver these items as fully as possible within their services according to exactly the same quantitative scale and scoring system as in any traditional IPS service.
For modified IPS services the key challenge is to effectively flex the fidelity items as required for the differing needs of their larger and wider cohorts and primary care contexts while retaining the core ethos, quality and performance of the IPS. The right column of Table 1 highlights in italicised text those fidelity items that our codesign work identified as in need of modification on order to function effectively for our modified IPS trial, as guided by the practical and theoretical approaches to modification outlined by Mowbray et al. (Reference Mowbray, Holter, Teague and Bybee2003). These fidelity items are of a qualitatively different nature from the shared standard fidelity items, in that while they seek to achieve the same intent and ends as those items in the traditional IPS fidelity scale the optimal means of delivering these in modified IPS services may require a reconfiguration of the item and its underlying scoring criteria given their new cohorts, needs and local systems and contexts. These items are for the first time in the literature identified as qualitatively distinct ‘modifiable IPS fidelity items’ and it is these modifiable fidelity items that are – and ought appropriately to be – open to careful configuration to deliver their desired intent and ends, dependent upon the needs, opportunities and risks of each specific programme and of the local system context that it sits within. Only then is it viable to consider how best to measure and score these modifiable fidelity items.
Within modified IPS services, however, this key translation task of those modifiable fidelity items is at present being conducted by international policy makers without clear guidance. The significant risk in this context presently is that this on-going uncritical modification of key fidelity items will – whether intentionally or, more likely, unintentionally – undermine the distinctive characteristics and effectiveness of IPS during its attempted translation for mainstream cohorts and needs. This is the analytical and policy problem that serves as a motivation for the discussion that follows. Informed by, but stepping out from, the specifics of this particular modified IPS service codesign and reflecting instead on the broader analytical considerations that emerge from that process, the discussion below offers for the first time in the literature an original and critically reflective analytical framework for the effective design of modified IPS services.
From modified IPS to modified IPSs: developing an analytical framework to guide future policy design
The task that policy makers face with these modifiable fidelity items is to configure the service in the most appropriate ways in order to deliver the intent and ends of each of these items in maximally effective and efficient ways. The modifiable fidelity items highlighted in Table 1 fall into two categories: one set of items that relate to the focus and intensity of the core employment specialists (target cohort, caseload size, and time-limited support) and, once defined, a second set of items that relate to the qualitative (re)configuration of the mode of programme delivery to maximise service user experiences and employment outcomes (IPS employment specialist role in providing whole-person support, integration into health team, employer contact frequency and employer engagement, collaboration with public employment service, senior health system support for IPS and open employment for patients).
Let us turn first to those modifiable fidelity items relating to the focus and intensity of the core employment offer – target cohort, caseload size, and time-limited support. The shift to modified IPS often involves opening out the target cohort to differing cohorts compared to the traditional IPS emphasis on severe mental health only, as in our trial's focus on low to moderate mental and/or physical health as well as to individuals who may be in-work and off-sick or struggling as well as out-of-work and seeking to move into employment.
Once the target cohort has been defined, a second decision concerns the desired live caseload for each IPS employment specialist – a key driver of core service intensity and also cost. Traditional IPS seeks a highly ambitious caseload of twenty within its fidelity scale. As with traditional IPS fidelity the task within any modified IPS service remains to maximise fidelity on this item via low caseloads. This can however be considered as a modifiable fidelity item so as to signal that for these wider and larger cohorts some slippage in target caseload size (to maximum thirty-five in our codesigned trial for instance) is possible if not expected. To provide some benchmarking context, typical caseload size in the UK context is around one hundred per employment advisor (Considine et al., Reference Considine, Lewis, O'Sullivan and Sol2015; WPSC, 2016: 33).
The third modifiable item in this first set relates to time-limited support. Traditional IPS models are in principle time-unlimited, though in practice IPS seek to exit service users from service once their employment transitions are stable. Recent trial evidence suggests that time-limiting IPS services to nine months’ support delivers similar employment outcomes as a time-unlimited service but in a significantly more cost-efficient manner (Burns et al., Reference Burns, Yeeles, Langford, Vazquez Montes and Burgess2015). The expectation is that many modified IPS services will be time-limited in this way in order to seek to optimally balance these financial and performance considerations.
Although these fidelity items may require reconfiguration for modified IPS services their reworking relates more to choices around priority cohorts and budgets that set the parameters within which the more complex design choices – and performance risks – of the modifiable fidelity items occur. Turning to that second set of modifiable fidelity items, of central importance to their effective reconfiguration is the recognition that programme designers can seek to deliver the intent and ends of these items in two radically different ways. One approach can be considered a discrete service approach in which the IPS service and employment specialist have primary responsibility for their delivery within a relatively stand-alone service. In contrast, an alternative model takes a networked service approach in which the IPS service and employment specialist are instead embedded into and required to co-ordinate the utilisation of the wider services, partners and networks of the local health and employment system.
A resulting framework can be presented visually as in Figure 1 to frame analytically a comparative approach to assessing modified IPS services. Along the vertical axis, services must strive to reach as high a score as possible on their modified fidelity scale in terms of the quality of their IPS service and its likely effectiveness, just as in traditional IPS. Along the horizontal axis, however, there is recognition that this can be done in qualitatively distinct ways dependent upon the extent of discrete or networked approach in its reconfigured delivery.
Mapping the nature and quality of alternative modified IPSs
Two significant implications fall out. Firstly, local context matters, for what is optimal will vary depending upon a range of programme specific factors – cohort, volumes, priorities, partners, leadership, ambition, existing employment and health context, budgets, and so on. Secondly, and related, there is as a consequence no one ‘true’ modified IPS model but, instead, potential for multiple different but equal modified IPSs in terms of their fidelity and effectiveness, dependent upon programmatic and contextual specificity – in contrast to the impression of one ‘true’ high quality model towards which all services should aim in the case of traditional IPS.
The next section discusses the principal options and considerations for policy designers across these key modifiable fidelity items that are at the heart of critically reflective and context specific qualitative translation of IPS principles and performance into modified IPS services.
In or out? Networked vs discrete reconfiguration of modifiable fidelity items
By way of illustration to these ideas, Table 2 summarises high-level differences between what alternative discrete and networked approaches to the reconfiguration of these fidelity items. The following sub-sections discuss each in turn in further detail.
Discrete and networked approaches to delivering modifiable fidelity items
Items 10 and 11: Senior support for IPS and for the principle of open employment for the health and disability cohort
The traditional IPS fidelity scale rightly acknowledges the importance of strategic, policy and financial support at senior levels of the health systems that run them. Similar senior support is equally important inside modified IPS services but is more demanding in ideally being required across a wider set of partner organisations, recognising the reach and interaction of modified IPS services to wider elements of the local health and employment systems – health partners still, but also local and regional government partners, public employment services (PES), and third sector too, for example.
Yet if it is essential to have this senior support across organisations, the question remains as to how best to configure it? In the utopian dream of locally integrated employment ‘ecosystems’ these senior partners might approach their own collaboration in this same networked approach with integrated governance, strategies, commissioning, budgets, targets, and so on. While in principle highly attractive and beneficial, the barriers to this approach are multiple and significant given the reality of differing organizational priorities, activities and governance. Alternatively, therefore, relevant senior partners might formally sign up to supporting the project and agree to collaborate as required on a task and finish basis to overcome blockages, commit necessary operational resources and ensure that the project can be a success, but might not commit to the formulation of shared strategies, policies, governance, budgeting and so on beyond this.
Item 6: Collaboration with the Public Employment Service (PES)
Traditional IPS services within secondary mental health services often involve relatively limited interactions with the PES. A new need in modified IPS services is to strengthen the relationship between the IPS service and the PES in terms both of cohort flows (both referrals and exit handovers) and, with varying importance dependent upon the nature of the national benefits system, ensuring any benefits conditionality requirements of individuals on the modified IPS services are met.
While traditional IPS services emphasise referrals from the same secondary mental health teams in which IPS employment specialists are co-located, in modified IPS services the wider system is instead key to referrals. Discussed below, wider primary care health services are of relevance. So too however are the nation's PES given that many eligible participants will be in receipt of social security benefits and attending meetings at their PES to check on-going benefit eligibility and/or receive employment support. It is sensible in modified IPS services therefore to encourage and enable referrals from PES to modified IPS services, and similarly to facilitate warm handovers with assessment and intervention notes from the IPS service back to PES as appropriate where employment outcomes are not achieved and a return to PES occurs.
In terms of conditionality, there is a key need to ensure that individuals inside these voluntary modified IPS schemes continue to satisfy any PES conditionality requirements such that they avoid sanctions and continue to be eligible for social security benefits while participating in modified IPS rather than PES employment support. These conditionality needs require close attention within modified IPS services, given that a significant part of their wider mainstream cohorts will continue to face such conditionality requirements, opening up potentially significant financial risks for participating low income service users. The importance of this need will inevitably vary by context, dependent both upon the cohort's level of social security receipt and the nature of the conditionality and sanctions regimes in operation in that country or locality. Whether this need can be handled in a networked way via seamless data flows, or – as in the UK context – instead via new clerical processes of evidence creation and sharing between the modified IPS service and PES, will be case-specific and depend on the legal, technical and relational contexts of particular services.
Item four: Integration into health teams
Integration remains a critical feature to the effectiveness of IPS programmes but requires systemic reconfiguration in modified IPS services in relation both to appropriate host health teams and referral sources. Likely host teams for IPS employment specialists in modified services include primary care services for mental health, physio and pain management teams, community hubs and GP practices. In a discrete approach to modifying this item, the host primary care team would be the health ‘universe’ in which the individual's support was considered and provided. A networked approach would in contrast recognise the frequent comorbidity of different health conditions for a significant proportion of any health cohort and establish expectations and processes to enable IPS employment specialist integrated into one team relating to the primary health condition of the individual – a primary care mental health team, for instance – to work effectively with another primary care team – pain management, for instance – as required, to offer whole-person healthcare support to their caseload. Alongside PES, all such primary care teams into which IPS employment specialists are integrated are naturally expected and enabled to refer into the modified IPS service, alongside wider parts of the health system potentially (such as pharmacies, community health teams and social prescribing services, for example).
Items 2 and 17: IPS services staff role in ensuring whole-person support and employer engagement
In modified IPS services the same three key functions as in traditional IPS services still need to be delivered: personalised and intensive employability support to individuals; delivery of health and wider (e.g. housing, debt, family, travel) support needs; and personalised employer engagement to enable effective job matching, entry and sustainment for individuals and employers. While traditional IPS services themselves deliver all three, and while IPS employment specialists in modified IPS services naturally retain responsibility for delivery of personalised employment support to individuals, there is both a capacity consideration and a system opportunity (alignment, effectiveness, efficiency, simplicity) to consideration of a networked approach to the delivery of employment engagement and wider support needs.
On the supply-side individuals will require supports for both health conditions and a range of wider support needs (e.g. housing, finances, skills, family). In terms of health supports, existing mental health and physical health primary care services are key, but are not always most appropriate for every individual's needs and are facing significant demand pressures in many contexts that affect eligibility and waiting times. One possible expansion of IPS employment specialist's roles in modified IPS services would be to become trained in the types of low-level mental and physical health needs that will be common across their caseloads – cognitive behavioural therapy (CBT), physio and pain management, for example. When conceived across a team of IPS employment specialists this range of skill sets, combined with clear training and guidance around how to practice and when it is appropriate to instead refer up to more specialist and/or intensive services, could be a helpful addition to ensuring timely and appropriate health support to service users. The risk would be to divert the time and expertise of IPS specialists away from the dedicated employment focus as in a traditional IPS model and it would therefore be important to assess empirically the effectiveness of any such role bend.
There is similarly a question about how best to deliver an individual's wider support needs in a context where needs vary, appropriate local services already exist, and caseloads are by financial necessity inevitably somewhat higher than in traditional IPS services. A discrete approach in which IPS employment specialists in modified IPS services are expected to deliver these wider support services themselves may be neither realistic nor desirable. Instead, networked approaches that draw on existing wider expertise and resources across the wider local system outside of the modified IPS service itself may be attractive. Indeed, local areas could help modified IPS services to deliver this function through formalised governance mechanisms to co-ordinate such wider support services – Local Integration Boards as they are coming to be known in several UK city-regions.
Similar considerations are apparent on the demand-side where it is possible to separate out key employer engagement functions around identification and proactive cultivation of suitable vacancies from individual-level discussion and brokerage with employers and service users to build relationships, understand needs, and support job transitions, adaptations and on-going sustainment. Within traditional IPS models the IPS employment specialist is responsible for both functions and while this may be replicated in modified IPS services more networked approaches may again be more realistic and/or desirable. It is possible that the initial function of large-scale aggregate initial employer engagement to source opportunities be led by existing (e.g. local authority) employer engagement teams outside of the modified IPS service itself but with an eye to the needs of its service users. These wider employer engagement teams could then communicate opportunities to the modified IPS service and broker introductions between employers and IPS workers. All individual-level work with and between employers and service users can then continue to be delivered by IPS employment specialists, as in traditional IPS models. Unlike the more aggregate-level functions around identifying initial opportunities, it is these individual-level relationally based functional needs (around understanding job roles and needs, brokering job matches and work transitions, arranging workplace and/or role adaptation and providing on-going support in work) where focusing the time and energy of IPS employment specialists in modified services can really add value.
Stepping back and moving forwards: realising the potential of modified IPS
Transforming the employment support experiences and outcomes for individuals with health conditions and disabilities is an essential part of a broader set of changes in societies and labour markets if nations are to begin to make meaningful inroads into the entrenched disability employment gaps that span advanced economies. IPS services have a proven track record of effectiveness with a severe mental health cohort in secondary mental health settings; and modified IPS services present significant potential in the larger, wider health cohorts in primary and community care settings that will be key to narrowing the gap. Based on the validated and well-evidenced principles, fidelity and performance of traditional IPS services internationally, this article has argued that well-considered, critically reflexive modification of IPS can and should offer a solid foundation on which to build more effective employment programmes for other wider types of health conditions and severities and, indeed, cohorts more broadly.
The article outlines however that a significant translation gap from traditional to modified IPS models exists at present and that this creates on-going barriers and risks around the design and resultant performance of such modified IPS services. This article's critically reflexive discussion of IPS modification, and its presentation for the first time in the international literature of an original analytical framework to guide policy makers through a well-considered process of IPS modification, offer powerful contributions in this context. Four specific contributions can be identified.
Firstly, the article differentiates for the first time in the literature between, on the one hand, what we term standard fidelity items that programme designers of modified IPS services ought simply to maximise in the same manner as traditional IPS services and, on the other hand, modifiable fidelity items whose configurations might appropriately require qualitatively reconfiguration in order to deliver optimally the desired intent and ends of the fidelity item.
Secondly, the discussion highlights that most items of the traditional IPS fidelity scale can indeed be understood as standard fidelity items that apply equally to larger mainstreamed modified IPS services for a broader health cohort. This is an important finding in terms of the flexible potential of modified IPS model to effectively support a far larger share of the total health-related employment challenge across advanced economies. Central to why is the recognition that the IPS fidelity scale and principles outlined above are in significant part expressions simply of what good employment support could and should look like for everybody, irrespective of whether they have a health condition. This is in contrast to the frequent view of IPS as some sort of different and special employment model for particular health cohorts and, critically, indicates that modified IPS models have significant wider potential beyond the health cohort to which discussion is normally exclusively attached. Yet, beyond the small and self-contained bubble of traditional IPS, this is often not what mainstream employment support for individuals with health conditions looks like, across the advanced economies. Indeed, countries such as the UK display a ‘low road’ employment support model comprising a highly limited employment offer to individuals with health conditions and disabilities (in terms of its coverage, intensity and resources) combined with an emphasis on low-cost behaviouralistic mandation and sanctions (Eleveld, Reference Eleveld2017; Fletcher and Wright, Reference Fletcher and Wright2017) – an approach which presents the near complete antithesis of the approach (from a solid evidence base) around what has worked to grow levels of employment, incomes and wellbeing for individuals with health conditions and disabilities.
Thirdly, the necessary reconfiguration of several key modifiable fidelity items can be approached in what we term either a networked or discrete manner, with decisions needing to be considered carefully based on the nature of the wider local services and system as well as the needs, priorities and resources of the modified IPS service. As a consequence, and quite unlike the situation with traditional IPS services, variability in the design of modified IPS services dependent upon programme and local context is both possible and perfectly appropriate, meaning no one ‘true’ modified IPS service exists for all contexts.
Finally, the article brings these ideas together within the presentation of an original analytical framework to help policy makers to develop effective modified IPS services in a structured approach that is well-considered and critically reflective in their particular programmes and local contexts. It is hoped that this will help analysts and policy colleagues to make more effective design decisions in their particular modified IPS programmes by filling the current translation gap and associated programme risks around the appropriate sharing and modification of IPS principles and fidelity items. It remains the case that such modified IPS interventions will require the same attention to the evaluation of their fidelity and performance as seen in the traditional IPS literature, in order that we can grow a similarly robust evidence base about their performance and optimisation. For if we are to make the much-needed step-change in employment support experiences and outcomes across advanced economies it is important that scholars and policy makers alike engage seriously with the solid evidence of IPS services and the growing evidence and analytical thinking around the potential of well-considered modified IPS services: evidence and thinking to which this present article has sought to contribute. Only then can we hope to design the types of employment support programmes that will give us the transformative change in outcomes that we want and need, and the types of support that our citizens deserve.
