Individualised placement and support programme for people unemployed because of chronic pain: a feasibility study and the InSTEP pilot RCT

Work package 1

Work package 1 involved a sample of people with chronic health conditions (including pain) who had been unemployed for at least 24 months and had been, therefore, recruited for and attended the SJP IPS intervention in the cities of Portsmouth and Southampton in the preceding 12 months.

Clients who were potentially eligible to take part were identified by the ESWs and managers of the IPS service in the city councils. We provided the ESWs with a participant information sheet that explained the nature and purpose of our research. The leaflet explained that we were seeking willing volunteers to take part in qualitative research, what was required of participants, that their expressed views would be confidential, anonymised and non-attributable, and that they could withdraw their consent at any time. Individuals who expressed interest in participating were asked to give consent for their contact information to be passed to the qualitative researchers. Potential participants were then contacted via telephone by the researchers to arrange an interview. Interviews were conducted at a time that suited the clients and they were offered the choice of an interview by telephone or face to face (arranged in their local job centre). It was explained to each client that the interviewer was independent of the IPS programme and that they were part of a research team investigating whether or not high-quality employment services could be of benefit to people who were unemployed and affected by long-term health conditions. Clients were assured that their comments would be strictly confidential and no one from the SJP IPS programme would be present during the interview. Travel expenses incurred by the clients were remunerated and participants were sent a high street shopping voucher after their interview to thank them for their assistance.

Work package 2

Work package 2 involved ESWs currently working in the SJP IPS intervention in the cities of Portsmouth and Southampton.

All eight ESWs involved in the SJP IPS intervention in the cities of Portsmouth and Southampton were eligible to participate. Eligible individuals were given a written information sheet by their managers, which explained the purpose of the study, emphasising that participation was voluntary and that comments would be confidential, anonymised and non-attributable, and that they could withdraw their consent at any time. Semistructured interviews were arranged at the convenience of the ESW at their place of work and travel expenses incurred by ESWs were remunerated. Written, informed consent was obtained from ESWs to record their interview and for the analysis of their views.

Work package 3

Work package 3 involved qualitative focus groups with members of the primary health-care team who were familiar with chronic pain patients (i.e. PCPs). PCPs were contacted via the local Clinical Research Network, which advertised the study to research-active general practices. Practices that expressed interest in the study were then provided with written information about the purpose of the study and what was required. Any member of the practice team involved in the care of chronic pain patients was eligible to participate.

Practices that expressed interest were contacted by a member of the research team by telephone to explain what was involved in the research and, if agreeable, a suitable time was arranged to conduct a focus group within the practice premises. Focus groups were chosen to facilitate interaction between participants and enable them to bounce ideas off each other,34 and also for convenience to minimise the total clinical time of the PCPs taken up by the research. It was emphasised that all expressed views would be confidential, anonymised and non-attributable, and that they could withdraw their consent at any time. Written, informed consent to participate, for recording and for the analysis of their views was obtained from all PCPs prior to the start of the focus group. Participating general practices were remunerated so that they could backfill the time required.

Undertaking a research trial involving individualised placement and support

The PCPs were extremely positive about the idea of a trial of an employment intervention for people unemployed with chronic pain:

I think the intentions are brilliant.

PCP1

Likewise, all ESWs were very positive about the SJP IPS programme and enthusiastic about research to show its benefits. They were keen to share success stories and highlight the benefits of the intervention that they had observed.

Primary care-based health-care professionals stated that they had observed patients who were trying to RTW and find employment, and that such individuals would be interested and engaged in a trial. However, they voiced concern that some individuals might not wish to RTW and that differences in motivation might prove problematic in terms of study trial recruitment:

And it’s the individuals, there are some people who actually like to work as we would like to see that, but there are those who just wanna get the money and the money is just easier if you are just sitting on your backside sometimes.

PCP1

The PCPs expressed concern that some patients may not want to seek employment and that it would be important to identify and measure motivation as part of the study.

One client saw the benefit of conducting the proposed research trial but, similarly, expressed concerns about the purpose and how that purpose would be communicated (e.g. about the wording of the information sheet that stated that the research aim was to ‘improve the health of people who are unemployed through long-term health problems’). She felt that this could be considered derogatory, feeling that she was not ‘unemployed’ but ‘unable to work’. At the end of the interview, she voiced apprehensions about the future research being used to force individuals to work even if they were unable to do so:

I think this is a really interesting study. I think that, it’s important, but I also think in the wrong hands, it could be used to be malicious and force people with pain into work.

C5

Overall, clients were very positive about the SJP IPS programme with which they were engaged. However, we found that they had difficulty in discussing a potential research trial that is distinct from their own personal experiences of their IPS programme. Even after clarification, some clients found the concept of a research trial difficult to understand and their answers were inherently linked to their individual experience of IPS. For example, clients seemed anxious that recruitment to a trial would mean an increased workload for their ESWs, who, they perceived, already had limited time available, without them being further burdened by new clients in a research trial:

I mean at the end of the day they’ve got 25–30, erm, clients, and you’ve only got four advisors maybe five advisors. And to be totally honest, they are also doing more interviews at the job centre in [location] and [location] and that to get more people on this course. So you are limited to the amount of time you can have with them.

C2

Recruitment to a randomised controlled trial of individualised placement and support

One client expressed concern about recruitment in health-care settings:

I don’t think . . . if I got a letter like that from my GP, I think I would just shove it in the bin. Or I would . . . I wouldn’t be very happy.

C5

In contrast, other clients felt that this method of recruitment could be a positive way to identify people who might benefit from the intervention:

That sounds like a good idea, because in a way when I was put on my antidepressants the doctor who gave me the antidepressants and telling me what to do, he told me that I needed to be part of a support programme, you are not gonna be able to do this alone, you’ve got your church, you’ve got your wife but you are gonna need more than that. You need someone to steer you in the right direction, which I have done and I continue to do.

C4

One of the core precepts of IPS is that participation is voluntary and dictated by client choice. Even so, our interviewees appeared confused about whether their enrolment in the SJP IPS programme was voluntary or mandatory. Although the ESWs explained at the outset that the programme was completely voluntary, some clients stated that they did not feel they had a choice about participation, particularly if they were referred by the job centre (which was the case for the SJP, but would not be the case in the planned pilot study). This uncertainty about their freedom of choice regarding employment services was particularly unhelpful when trying to elicit their engagement with employment-related research and when explaining the voluntary nature of research participation:

Even the advisors say it’s ‘voluntary’ it’s not voluntary . . . when you go to see the advisor at the job centre they said ‘you’re on the work activity group, you must be doing something’ so they put you on [IPS programme]. And then when you come here, they say it’s all voluntary. But it’s not voluntary; the job centre put you on this course for a reason.

C3

Clients also commented on the large volume of paperwork that arose from participation in the programme and evaluations of the IPS (i.e. evaluations of the SJP and unrelated to the current research) and that the need for this was not apparently explained to them. One client did not understand the purpose of the volume and content of questions asked in evaluations:

Yeah it’s like these [consent forms and information sheets]. Tick this thing on the computer. Tick, tick, tick. One question was did you have school dinners as a child? Yeah what’s that about?

C3

One ESW hypothesised that individuals recruited in different settings may differ from the clients who were currently being offered IPS:

. . . the referral would be different, it would be interesting to see those clients coming from a doctor’s surgery that are told to speak to someone, ’cause their mentality might be different, to people that are in the process of referrals coming often from the job centre.

ESW5

Primary care-based health-care professionals discussed recruiting eligible patients from health-care settings. They felt that it would be possible to undertake database searches using the Read codes. Read codes are the standard clinical terminology system used in general practice in the UK. 35 They provide a hierarchical clinical coding system for the purpose of reporting research decision-making and allow data to be shared reliably between different computer systems. They have been used previously for chronic pain. 36 However, the PCPs recognised that currently no Read code for unemployment exists, and so searches would identify patients with pain and certified sickness absence, but not unemployment. PCPs suggested that Read codes for chronic pain conditions could be used in conjunction with medication codes (e.g. opioids or gabapentinoids) to identify potential participants.

The PCPs also believed that they could personally identify individuals who would benefit from the programme based on regular contact through primary care services.

About the nature of the individualised placement and support intervention

Employment support workers were in agreement with offering IPS after only 3 months of unemployment, which they thought would be particularly beneficial. The clients made the same point, suggesting that the sooner after losing their job someone could enter the programme, the better it would be:

Because everybody needs that helping hand, if you think about it, if somebody’s just come out of work, for 3 months and they’ve got nothing, that’s where they’re starting to lose the point, ‘I’m unemployed, I’m signing on, I’ve gotta do this, I’ve gotta do that . . . you know where do I go next?’ We don’t want that person to go 3 months after that, 6 months unemployed and still going nowhere, they need to be somewhere where they get the support, yeah I totally agree with that.

C4

Likewise, the PCPs emphasised the importance of early intervention after unemployment:

I think the sooner you can get someone back to work, the better.

PCP4

The ESWs felt that IPS should continue to be available for longer than 1 year, as the process of preparing someone for work and securing employment could be lengthy. Clients, likewise, reported that it could take time to get someone ready for employment and into the workplace.

The pivotal role of their individual ESW in successful IPS was emphasised by the clients and, in their view, this relationship was crucial in finding clients suitable employment. On the other hand, some clients described a lack of continuity in the programme and felt that they had not developed a relationship with a single ESW, having met multiple different ones, which had negatively affected their experience of the programme:

I’ve just been moved from one advisor to another advisor, to a new one.

C2

Clients reported that ESWs asked about their potential needs and barriers to creating an individualised plan. One client recalled that the specific details of her condition had not been considered, but rather a very general discussion of her disability and illness:

So, for people with chronic pain, or anybody with mental health, or, you know, or anything that’s specific, I think what would improve this service is having an advisor who specialised in that condition. Or in a couple of things. It’s having that understanding, it bridges that gap.

C5

It seemed to the clients that personalised advice and a holistic approach were key features of the IPS intervention. Likewise, ESWs discussed the importance of tailoring IPS to each individual client. This was usually achieved by discussion with clients, as the ESW had limited knowledge about the impact and management of specific conditions:

. . . you should have advisers that have training around chronic pain and there should be a fully comprehensive directory of signposting people.

ESW4

Clients were aware of the existence of other employment interventions available for unemployed people. However, IPS was felt to be the best approach for patients with chronic pain, as other employment interventions were seen as not appropriate for individuals with complex issues:

. . . the other programme, they understood my problems but they didn’t do anything about it.

C4

The ‘control’ intervention was mentioned by the PCPs. The PCPs reviewed the draft TAU booklet that would be provided to all subjects. They reviewed the booklet positively and, in particular, highlighted the vernacular language used throughout and the reader-friendly style in which it addressed commonly asked questions without overwhelming the reader with excessive information:

I think it’s very useful, I think it will help people . . . I think patients will find it quite useful.

PCP8

The PCPs reported willingness to randomise patients to both the IPS and control interventions. One PCP explained that in previous studies participants had been unhappy if they had been randomised to a control group, and felt that it would be important to explain the benefit of both interventions:

I tend to . . . not sell it if it’s a non-intervention or control, but actually saying . . . you are actually really important in this study as well. They are part of it and they are helping. If we get to the actually nitty gritty and we are actually recruiting patients, then they’ve got that far, and we’ve talked about it, and we’ve consented them, then hopefully they are still on board.

PCP4

Outcomes to be assessed in a future trial

Participants appreciated that the outcome of the programme would be entry to employment, although they indicated that a client may well not have found employment within the relatively short length of the proposed intervention (i.e. 6–12 months). At the time of interview, none of the clients on the SJP IPS programme had secured paid employment, but all were keen to discuss the progress they had made. They highlighted the skills they had gained from the programme that made them qualified for employment, with ‘job-readiness’ seen as a potential outcome for a future study:

I feel a lot better about starting now, I don’t expect anything to happen until I’ve been here a few months or so, one thing at a time, I need some solid ground to stand on.

C7

Some clients also described a boost in their confidence since starting the programme and reported that they were being more active in daily life:

. . . it gets me out of the house and that you know what I mean, well . . . I’m out and about every day.

C6

Employment support workers recognised that the likely principal outcome of IPS would be entry to employment. However, they also recognised the probability that clients may not yet have secured employment by the end of the intervention. ESWs were keen to talk about the success stories of the IPS programme, notably client improvement in several quality-of-life domains.

The ESWs highlighted the skills that clients had acquired from the programme, which had enabled them to consider a work placement in the future. ‘Job-readiness’ was therefore proposed as a potential outcome for a future study. The importance of building clients’ confidence was also noted:

The knee jerk reaction, which throws up the barriers in the first place. So you’ve got to get those down. And by doing that, when people talk about it, when you are chatting with the client and I’ll say ‘but if you are doing that, why don’t you take it. it’s the same thing but you are on a different’, and they go ‘ohhh right’. Erm, so you are getting those kind of barriers down. At the same time as doing that, ’cause you actually understand, it builds their confidence. And ‘maybe I could do that, maybe I could plan’.

ESW3

Similar to the ESWs, PCPs speculated about the importance of the intervention in building clients’ confidence. Participants also discussed changing clients’ attitudes and views about work.

The PCPs postulated a number of benefits of employment for individuals with chronic pain. For example, they suggested that employment could increase physical activity, which in turn might reduce pain levels. In addition, they highlighted the link between chronic pain and mental health, stating that improving social interaction and sense of achievement through work could improve depression. It was thought that employment could improve patients’ overall quality of life:

So then at least these sort of work based programmes are starting to tackle that, helping people to get out of the house regularly, introducing some sort of ‘maybe I could do something’, a degree of hopefulness, where there is a degree of hopelessness.

PCP8

A summary of all of the barriers to and facilitators of a future trial of IPS for people unemployed with chronic pain that were identified from WPs 1–3 is shown in Table 2.

Pre-assessment screening

All individuals who returned a reply slip were subsequently telephoned by the study co-ordinator at their convenience. This provided the opportunity to discuss the study and address any questions that the individual might have had. If agreeable, the study co-ordinator completed a brief eligibility screening form with the individual over the telephone to confirm that they:

• were able to provide written, informed consent

• had not previously been referred to, or had taken part in, the IPS programme via Southampton or Portsmouth city councils

• had chronic pain (i.e. pain continuing for > 3 months)

• had completed the diagnostic pathway for their chronic pain

• did not anticipate recovery within the coming 12 months

• wished to RTW.

Following satisfactory completion of the screening form, the individual was asked to give verbal consent to randomisation by the study co-ordinator. If agreeable, the participant was then randomly allocated by computer-generated algorithm (block 1 : 1) to either the active IPS intervention arm of the study or the TAU control arm.

Finally, the study co-ordinator informed the participant regarding the next (baseline visit) stage of the study. If randomised to the TAU arm, the research co-ordinator herself telephoned to make the baseline appointment at their general practice. If randomised to the intervention, the local ESW telephoned the participant to arrange the baseline appointment at the local city council premises.

Treatment as usual/control arm

There was no evidence-based NHS alternative to IPS available and suitable for use as a ‘control’ for the pilot trial. However, a number of services are available and are provided by the Department for Work and Pensions, local government and the voluntary sector, and individuals may self-refer to these services or can be referred by health-care professionals. Although the nature and type of services varies widely by region in the UK. It is not known what proportion of patients with chronic pain discuss their employment status with their GP or other member of the primary care team. It is also not known how comprehensively the PCPs keep themselves up to date about the availability of appropriate services in their area. A significant hindrance for this patient group is that they frequently lack confidence, skills and self-efficacy as a result of their chronic pain and comorbidities, such as depression and anxiety. Therefore, for the pilot trial, NHS TAU was supplemented with a standardised booklet to signpost participants to local employment and health-care services. We chose this approach both for ethical reasons and to encourage participation.

Two booklets were co-designed with our PPI group (i.e. one for each city involved). The resultant booklets were professional looking, informative and tailored to the particular location (i.e. Portsmouth or Southampton), with specific information about local services for pain management and job support, including voluntary and third-sector organisations and health-related advice. The PPI group suggested that pages should be included for participants to make their own notes and to construct a list of goals that was structured towards ultimately enabling gainful employment. The booklets also promoted positive messages about self-efficacy and the value of employment in enhancing health and well-being. It was envisaged that the booklet might be an appropriate vehicle to facilitate discussion with their GP about enhancing their health sufficiently for them to RTW using existing services, and provide clear guidance about local services that might help them seek employment.

Participants randomised to TAU were contacted and offered an appointment, at their convenience, with the study co-ordinator (who had a background as a health-care professional, but had no formal occupational rehabilitation training). At the appointment, any questions about the research were answered, written, informed consent for participation was obtained and the baseline questionnaire was completed (taking approximately 20 minutes). Subsequently, the study co-ordinator spent approximately 10 minutes guiding the participant through the booklet and encouraged them to take it home and read it at their leisure. Travel costs incurred were reimbursed. No further appointments were made.

Individualised placement and support intervention/active arm

Participants allocated to the active arm of the study attended a baseline appointment with the council-employed ESW at the local city council premises, which lasted approximately 1 hour. The ESW confirmed with the individual that they were willing to participate, addressed any further questions that they might have and then obtained their written, informed consent to take part in the study and in IPS. The participant was asked to complete the baseline questionnaire (which took approximately 20 minutes) and was also given a copy of the TAU booklet. The participant was encouraged to take the booklet home and read it at their leisure. Travel costs incurred by participants were reimbursed.

Adapting individualised placement and support for chronic pain patients

Based on the results of WP 3, the ESWs were provided with initial training by the specialist pain team before recruitment commenced. The content of this training focused on what chronic pain is, how it presents in practice, the types of medication used to manage the pain (and their side effects) and common approaches to enable functioning used by pain experts.

Furthermore, all participants recruited to the trial and randomised to IPS were seen by their assigned ESW and the community pain occupational therapist (OT) together for one of their initial appointments. Some standardised integrated documentation was developed, including those domains that both parties usually covered in their initial consultations, which it was agreed would be shared subsequently with the study team. The focus of these joint appointments was on assessment of the participants’ pain and current pain management strategies, with the possibility of specific counselling and support from the OT, signposting to other relevant services or follow-up by the pain team. The ESWs were invited to integrate as much as possible with the local pain services and attend multidisciplinary team meetings, and follow-up pain service use was available to any participant.

Active arm: individualised placement and support (subsequent follow-up appointments)

Following local procedures for IPS, and alongside people participating in the SJP, pilot trial participants met with the ESW as frequently as required to support and develop their employment plans. Fundamentally, this process fell into three stages: (1) assessing preparedness for work, (2) preparing for work and (3) choice of and allocation to, a competitive paid work placement.

Questionnaires

Questionnaires were developed for baseline and reassessment at 3, 6 and 12 months of the pilot trial. The baseline questionnaire was completed with the support of the ESWs or study co-ordinator (during attendance at the baseline visit). The three follow-up questionnaires were posted to allow participants to complete and return the form at their leisure. A pre-paid envelope accompanied the questionnaire. If no response was received after 4 weeks, a reminder letter and a further copy of the questionnaire were sent. On receipt of each completed questionnaire, a £10 shopping voucher was sent to the participant to thank them for their assistance.

Analysis

Our analysis focused on the acceptability of the questionnaire, and the rates of missing data in returned questionnaires and the attrition of participants in both arms. Another aim was to scope the most suitable outcome measures for a definitive trial and define the size of change to carry out power calculations. The questionnaires therefore asked about demographics, past employment history, current employment aims, comorbidities, health-care utilisation and health literacy. The validated tools that were included were the Brief Pain Inventory (BPI), the EuroQol-5 Dimensions, five-level version (EQ-5D-5L), health-care utilisation, the Return-to-Work Self-Efficacy Scale, the Warwick–Edinburgh Mental Wellbeing Scale (WEMWBS), Waddell’s Fear-Avoidance Beliefs Questionnaire, the Patient Health Questionnaire-9 items (PHQ-9), self-rated health and Rosenberg’s Self-Esteem Scale. At follow-up, we additionally asked about interviews that the participants may have had for jobs and any new employment that they had taken up.

Recruitment

Our original estimate was that four general practices would be needed to recruit a maximum of 80 pilot study patients. However, to recruit the final sample, we needed to widen our sampling frame to include nine general practices in Southampton and Portsmouth (covering an estimated 200,000 people).

We trialled all of the approaches to primary care recruitment suggested in WPs 1–3 to understand (1) which was most effective and (2) which was most efficient. Table 3 summarises the outcomes of the approaches we made to patients identified through the different recruitment methods.

Randomisation

Given the delays experienced with recruitment from primary care, block randomisation was compromised by the availability of the IPS intervention until a fixed time point (i.e. until the end of September 2019 and so the final recruit to the active arm was required by September 2018 to achieve the 12-month follow-up). Therefore, after random allocation initially, we subsequently allocated as many of the earlier recruits as possible to IPS until it was no longer available, and the remainder were thereafter allocated to TAU. Importantly, all those recruited from one large practice early on (n = 13) were randomly allocated 1 : 1 so that an assessment of the risk of contamination could be made. In all cases, the participants were recruited to a RCT and gave written informed consent to be allocated randomly to either arm. Nobody in either trial arm expressed dissatisfaction with randomisation or their allocation, and nobody dropped out after allocation or before their first appointment.

Contamination

Although the number of participants involved was small, the risk of contamination was assessed to be low, given that study participants viewed their employment as of limited relevance to their health care and held the same belief about the interest of their GP. GPs were not directly involved with any aspect of the trial and all trial information (including the TAU booklets) was provided by the study team directly and not distributed by practice staff members.

Characteristics of study participants from baseline questionnaires

Table 4 describes the baseline characteristics of those recruited to the pilot study, overall and by allocation. More women than men were recruited and participants were mostly white. A relatively high proportion were single/divorced (n = 21/50). The mean age at which participants left school was 16 years and approximately half attended further education or university. A small minority (n = 4) had university degrees and more than half (n = 30) had vocational qualifications. Interestingly, three participants had never held a paid job. Of the remainder, all had previously stopped working in a paid job, mainly or partly because of their health. When asked to indicate the type of health condition that had led to job loss, responses included chronic pain, but comorbid mental health/stress and ‘other’ diagnoses were also described by participants. Comorbidities were well reported by participants and revealed very high levels of anxiety/depression (n = 33 ‘yes’; n = 3 ‘not sure’). Most of those recruited (n = 41) were looking for part-time rather than full-time work and wanted to work < 24 hours per week.

Although study eligibility allowed for any period of unemployment > 3 months, we recruited people with a wide diversity of periods of unemployment (see Table 4). Twenty-eight of the 50 participants had been unemployed < 2 years and, therefore, would not have been eligible for SJP IPS. The median period of unemployment was 3.0 [interquartile range (IQR) 1.3–5.5] years, and this was quite similar in both arms (2.8 years for IPS and 3.2 years for TAU). However, we did find that there were four participants with a very long period of unemployment prior to baseline (> 20–30 years) and all four participants were allocated to TAU (Figure 1). The reason for this is likely to be the change in recruitment strategy in the last 6 months of the pilot (i.e. having opened up recruitment from pain services) because IPS became unavailable in the last 3 months. In exploring changes and distribution of outcome measures over time, this potential allocation bias needs to be borne in mind. In all other characteristics, however, the two groups seemed similar.

FIGURE 1.

Duration of unemployment (years) prior to recruitment to InSTEP pilot trial by treatment allocation.

Having identified a possible difference between participants recruited through primary care, as compared with pain services, we explored their baseline characteristics in more detail (Table 5).

We found that those participants recruited through pain services were more ethnically diverse and were very much more likely to be single (54% vs. 17%), but were somewhat more likely to have received further educational opportunities (69% vs. 46%). It appeared that those participants recruited from pain services reported somewhat lower health literacy than those recruited from general practices.

Participants showed themselves willing to provide a great deal of information about their personal circumstances, dependants, home ownership, finances and welfare benefit dependence (Table 6). Many participants (19/50) were living alone, but 16 of 46 had children aged < 18 years and the majority (n = 44) had dependants outside their household. Only 12 participants owned their home outright and most reported that they were struggling financially. Thirty of 50 participants were in receipt of welfare benefits and 41 were willing to describe their monthly income. Alcohol abuse was not reported by any participant. Fifteen participants were current cigarette smokers.

Treatment-as-usual arm

In total, 28 participants were randomised and allocated to the ‘control’ arm. All of these participants attended their appointment with the research co-ordinator, which was arranged at their convenience in their general practice. Practices were extremely helpful in identifying a room for these appointments. Participants expressed satisfaction with the appointment and the TAU booklet. No further appointments were arranged thereafter, but follow-up questionnaires were mailed as planned at 3, 6 and 12 months.

Individualised placement and support arm

In total, 22 participants were randomised and allocated to the IPS arm. For the purposes of the pilot study, IPS was provided by just two trained ESWs with experience of IPS as part of the SJP (one in Portsmouth and the other in Southampton). All 22 participants allocated to IPS attended their baseline appointments with the ESWs and provided written, informed consent to participate in the study. The baseline questionnaire was completed with the support of the ESW, if required (taking a maximum of 20 minutes of the appointment). Subsequently, all 22 IPS participants attended an appointment held jointly with the pain specialist (a single pain-trained OT took this role throughout the study) and their ESW early after recruitment, during which their pain, medication and pain management strategies were discussed. A pain management plan was constructed by all three parties. The pain OT found that participants varied widely in their needs and, therefore, a range of different interventions were recommended at this appointment. These are summarised later in this report for manualisation of IPS (see Table 14).

The ESWs found that the 22 trial participants also differed considerably in how ready and prepared they were for a work placement. As with all SJP participants, they assessed each individual based on their current needs and worked with them in aiming towards readiness for and acceptance of a supported work placement during the time available for intervention. The variation has been summarised later in this report for manualisation (see Table 14).

Withdrawals

Two participants allocated to IPS found paid work very rapidly (i.e. before the 3-month questionnaire), but the job opportunity required them to relocate (and in one case out of the UK). In consequence, both withdrew from study follow-up.

Questionnaire response rates and retention

Questionnaires were mailed out to all 48 remaining participants at 3, 6 and 12 months, with reminders posted 4 weeks later, if appropriate. In total, five participants completed only the baseline questionnaire and did not return any follow-up questionnaires. We compared the baseline characteristics of those who failed to return any questionnaires with those who returned at least one of the three questionnaires (Table 7). In particular, we were concerned with whether or not those who had not returned any questionnaires had lower levels of health literacy. The non-responders were mostly men (4/5), white (4/5) and single (4/5). They were more likely than responders to have further education (4/5). According to every parameter of health literacy, they were better off than those who completed at least one postal questionnaire.

Response rates by allocation

Four of the participants who failed to return any postal questionnaires were in the IPS arm and one was in the TAU arm (see Table 7). Overall, 19 of 28 participants in the TAU arm returned all of their questionnaires, but only 10 of 20 participants in the IPS arm returned all of their questionnaires (Table 8).

In the control arm of the trial, there were no withdrawals and all but one participant returned at least one of the three follow-up questionnaires. We hypothesised that differential response rates in the two arms might be at least partly because of the nature of the control arm. In the control arm, the study co-ordinator met with everyone to administer the TAU booklet, and during this appointment emphasised the nature of the research and explained the importance of the follow-up questionnaires. In contrast, those participating in the IPS arm had a number of appointments with the ESWs; however, these appointments were focused on the intervention and job-readiness, and, therefore, it may be that participation in the research was eclipsed in importance. Importantly, the ESWs explained that around 55% of people referred to the SJP dropped out after the assessment visit, either because they did not wish to RTW or because they felt unable to commit to the programme. It appeared, therefore, that our pilot trial participants were more likely to remain engaged with IPS than those people recruited from job centres through SJP, possibly suggesting that we were identifying people who had been unemployed for a shorter time and who were keen to get back to work, or that having their health assessed was attractive to participants in the InSTEP pilot.

Self-efficacy for return to work

Self-efficacy for RTW does not currently have a consensus scoring system and three different approaches have been published. 46–48 The scale developed by Brouwer and colleagues46 was particularly attractive for use in the current research project, as it included four items that specifically enquired about pain and RTW. However, subsequent to publication of this measure, Black and colleagues,49–51 attempted to incorporate the two pre-existing instruments of Brouwer and colleagues46 and Lagerveld and colleagues48 (Lagerveld and colleagues48 include items about the social aspects of returning to work) into one new measure, which they have validated. 49–51 Their final scale [i.e. the Return-to-Work Self-Efficacy Scale (RTW-SE)] includes 11 items, four of which were derived directly from Brouwer and colleagues. 46 Given this similarity and the potential importance of self-efficacy for RTW as a primary outcome for a definitive trial, we chose to include the questions for measuring both scales. The Brouwer and colleagues46 scale includes 10 items rated from 1 to 5 (‘not at all certain’ to ‘completely certain’). In contrast, the Black and colleagues49–51 RTW-SE uses 11 items, also rated from 1 to 5, but the responses are ‘strongly agree’, ‘agree’, ‘neither agree or disagree’, ‘disagree’ and ‘strongly disagree’. To maximise their utility for our research, we incorporated the additional items from Black and colleagues’ research, but did so by changing the questions and responses so that they could be included alongside the Brouwer and colleagues’ items, with the same heading and response choices. For example, the statement ‘If I resumed my work fully tomorrow, I could discuss any limitations I have to my co-workers’ with original responses graded ‘strongly agree’ to ‘strongly disagree’ became instead ‘Thinking about starting a new job, how certain are you that you would be able to explain your physical limitations to co-workers?’ with response options ‘not at all certain’ through to ‘completely certain’. In total, therefore, we asked 17 questions from the two tools. The 10 items from Brouwer and colleagues46 were analysed as a total score (scaled 0–10) and in three domains: (1) ‘supervisor’ (four items), (2) ‘pain’ (four items) and (3) ‘co-worker’ (two items), each of which were also scaled separately as a score out of 10. The RTW-SE was analysed as a total scale (i.e. including the 11 items and scaled 0–10) and also for its three different domains: (1) work completion beliefs (three items), (2) affective work beliefs (five items) and (3) work social support beliefs (three items), again each scaled 0–10. In all cases, a low score indicates poorer self-efficacy (Table 10).

These data suggested that people allocated to each arm started with different levels of RTW self-efficacy, with the scores better among those in the IPS arm. In the TAU arm, RTW self-efficacy generally declined over 12 months of follow-up. In contrast, although there were withdrawals and non-responders in the IPS arm (two of whom had attained employment), self-efficacy scores remained fairly stable in this arm and were generally substantially better (and in many cases statistically significantly better) than those in the TAU arm. Even despite our lack of data from non-responders, we believe that RTW self-efficacy could be a primary outcome for a future trial.

Self-rated health

As a general health measure, we included self-rated health. This is a widely used item that is a strong and consistent predictor of mortality in epidemiological studies. 52 Among the 50 participants, 30 reported that their self-rated health was fair or poor at baseline (Tables 11 and 12).

The data show that self-rated health scores generally declined slightly over time among all participants. This measure does not appear useful for a primary outcome.

Brief Pain Inventory pain severity scores

We included the BPI as the pain instrument53 in line with the recent IMMPACT (Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials) recommendation that this tool should be used for clinical trials involving patients with chronic pain. 52 Every participant reported pain (other than everyday kinds of pain) at baseline, one participant in the IPS arm reported no pain at 3 months, four participants (again all in the IPS arm) reported no pain at 6 months and only one participant (again in the IPS arm) reported no pain at 12 months. The number of body sites affected ranged from one (in two participants) to 13 (also in two participants). Pain severity at its best was rated at a median intensity of 4 (IQR 2–5) at baseline and at 3 months in both arms, and these scores increased in both arms to a median intensity of 5 (IQR 4–6) at 12 months. Pain severity at its worst was rated at a median intensity of 7 (IQR 6–9) at baseline and 3 months, increasing to median intensity of 8 (IQR 6–9) at 6 months and returning to a median intensity of 7 (IQR 6–9) at 12 months. Pain, on average, was rated at a median intensity of 5 (IQR 4–7) at baseline, but was rated at a median intensity of 6 (IQR 5–6) at all follow-ups. It does not appear that pain scores were altered by this intervention or that they could be recommended as a primary outcome measure.

Brief Pain Inventory pain interference scores

As part of the pain interference section of the BPI, one question enquires about ‘pain interfering with normal work’. At baseline and at all three follow-up time points, participants scored a median of 7 (IQR 5–8) for this question. There was a reduction in these scores, but the size of the effects appeared to be small, and certainly too small to recommend this as a primary outcome measure.

Rosenberg’s Self-Esteem Scale

At baseline, the median Rosenberg’s Self-Esteem Scale score54 for participants was 16 (IQR 12–21). The scores improved slightly in both study arms at 3, 6 and 12 months [the score in the IPS group at 12 months was 19 (IQR 14–21) and in the TAU arm it was 16.5 (IQR 15–18)]. This measure may be a useful secondary outcome measure for a definitive trial.

Patient Health Questionnaire-9 items depression score

At baseline, the median PHQ-955 score for depression was 10 (IQR 7–16) (score ranges from 0 to 27, with higher scores more indicative of depression). Very little change in the scores was observed during the follow-up period. If anything, scores tended to increase (but by < 1 unit) and, therefore, the difference is too small to recommend as a primary outcome measure.

Warwick–Edinburgh Mental Wellbeing Scale

The WEMWBS56 was explored as a measure of well-being. The median score at baseline for WEMWBS was 41 (IQR 37–52). Scores changed minimally over the follow-up, with a final median score of 42 (IQR 37–48). This implies that WEMWBS scores tended to improve during the trial, but the difference appears too small to recommend this as a primary outcome measure.

EuroQol-5 Dimensions, five-level version

The median EQ-5D-5L combined score was 0.6 (IQR 0.3–0.7) at baseline. Over time, these scores did not appear to change very much [with a median of 0.7 (IQR 0.2–0.8) at 12 months]. However, when we explored the visual analogue scale score of the EQ-5D-5L as a continuous measure, the median score at baseline [i.e. 50 (IQR 40–70)] appeared to increase slightly at 6 and 12 months among those in the IPS arm [a median of 54 (IQR 40–78) at 6 months and a median of 52.5 (IQR 40–80) at 12 months], whereas no changes were seen in the TAU participants. Although encouraging, these findings suggest that the EQ-5D-5L should remain a secondary outcome for assessment of quality-adjusted life-years, but not be recommended as a primary outcome measure.

Employment outcomes

Among the TAU group, 14 (50%) of 28 participants returned all three follow-up questionnaires and never entered a paid job (Table 13). Three participants started a paid job during the year, but stopped working again before the final questionnaire. Another three participants reported that they were not working at the time of their last completed questionnaire. Two participants reported that they were in paid work at every follow-up point and another six reported that they were working in the last questionnaire that they returned. Therefore, at best, 8 of 28 participants of the TAU group achieved sustained work during the study and, at worst, it was 3 of 28 participants. It did seem, however, that participants became less likely to return questionnaires once they were working. This may be because they perceive that they were no longer relevant to the research or that they had more limited time for completing a questionnaire. Either way, this is an important consideration to address in any definitive trial.

As stated above, 2 of the 22 participants who were allocated to IPS obtained jobs within the first 3 months; however, in both cases the jobs required that they move away and so they decided to withdraw from follow-up in the research. Another four participants in the IPS arm engaged well with the ESWs and the IPS intervention, but failed to return any follow-up questionnaires. As described above, we considered that this might be because of poorer health literacy, but our analyses suggested that this was not the case and that, in fact, the non-responders were more likely to be male, single and have a high level of educational attainment than those who did return questionnaires. Among the remaining 16 participants, five returned all three questionnaires and were never in a paid job, and two obtained work at one follow-up time point, but reported that they were not working at 12 months (Table 14). A further five participants were not working at the last time point at which they responded. Overall, two participants were in paid work at all three follow-up time points, one participant obtained work by 6 months (which was sustained at 12 months) and one participant was in paid work when they last responded (at 6 months).

Employment outcomes: data from employment support workers

Additional data about employment outcomes were available for the IPS participants from the records made by the ESWs. According to these data, 8 of the 22 IPS participants attained a job by the end of the study (including the two participants who moved away to work), a further one participant was doing voluntary work, four participants were in vocational training and seven participants were actively job-seeking (with one participant having received a job offer).

This indicates a clear difficulty with this research: the questionnaire retention was poor among those receiving the IPS intervention. As a result, the follow-up questionnaire data provided an incomplete picture of the changes in actual achievement of a competitive paid job. This is a major limitation of the current pilot study and would need careful consideration for any future trial.

Job applications and job interviews

Given the rates of attrition of follow-up questionnaires, it was also difficult to comment on whether or not participants in either arm were making more job applications or attending more interviews for jobs.

Untoward incident

One individual contacted the Research Governance Office at the University of Southampton (the sponsor for the research) directly to complain about having received the study pack. The complaint was left as an answering machine message, as nobody was in the office. Despite numerous attempts to contact the person and having left numerous telephone messages, the University of Southampton research governance team were not able to speak with the individual to discuss the complaint. The general practice from which the pack had been mailed confirmed that they had not received a complaint from any individual with regard to the study. We were unable to follow this up any further, as we had no way of contacting the relevant complainant.

Study participants

Disappointingly, only 4 of the 48 trial participants expressed interest in participating in WP 5 (three participants from the IPS arm and one participant from the TAU arm). All of these participants provided contact details, but, despite numerous attempts, only one participant (from the IPS arm) actually proved contactable and was available to attend an interview with the qualitative researcher.

This one interview was carried out and analysed; however, the participant was nervous and rather unforthcoming, once again demonstrating some difficulty in understanding that the research objectives are separate from those of the employment intervention.

The themes that emerged from comments on the returned questionnaires covered participation in the study, content of the questionnaires, effective and less effective aspects of the IPS programme and the realities of living and working with chronic pain. The comments also highlighted the role that health-care professionals can play in worsening disability.

Participation in the study

Participants in both arms expressed gratitude for the opportunity provided within the study:

. . . thanks for your help for myself, thanks for your assistance to me. I happy had this time for my health. Thanks so much.

IPS

It has been a pleasure and hopefully I have contributed to your studies.

TAU

Questionnaire content

Two participants reflected that some questions in the questionnaires were difficult for them to answer:

There were questions I answered which I’d have answered differently if it hadn’t been a check box. There were answers I left as I didn’t know the answer.

IPS

Some of the questions are difficult to answer as they do not accurately deal with my illness.

TAU

Aspects of individualised placement and support

Positive progress was reported by an IPS participant who also reflected on the importance of the ESW’s confidence and support in that progress:

I am moving forwards slowly. When I see ESW I feel more optimistic in gaining some paid employment, however, I wave/wobble on my own. I have applied for some voluntary work. I have socialised more than over the past 3 months which has improved confidence.

IPS

Another IPS participant reflected on the role that IPS had played in their life and that the same could not have been achieved by TAU:

I couldn’t have turned myself around and regained my shattered/crushed confidence/self-esteem fundamental to a return to paid employment without your support and having seen you over the course of the last year. A paper questionnaire approach alone from the InSTEP study, would have totally failed with my type of personality/temperament. With your support I view the year’s InSTEP study and work a total success!

IPS

However, another IPS participant expressed dissatisfaction with their IPS work placement:

From the start of the programme, I have had one job through the scheme which was at a coffee shop. After 2 months it came to light that there never was going to be a permanent position for me. I get the impression that companies sign up for this scheme only use it for cheap/free labour.

IPS

Living and trying to work with their chronic pain condition

One participant from the IPS arm reflected that their pain was worsening:

I have had physio [physiotherapy] several times but it hasn’t done much good. I have also had steroid injections in my hands that also didn’t work. I shall make an appointment with the doctor as some of the pain seems to be getting worse at times.

IPS

A TAU participant described their situation at 12 months’ follow-up as follows:

I’m getting no support and have just left university because of this and another illness has made it too difficult to study or work. I can’t prove either so getting work or income support is extremely difficult.

TAU

Similar themes were expressed by another TAU participant at 12 months:

I have been given an appointment at the orthopaedic department and this can’t come soon enough. My quality of life is really low as I have no money, no transport and my mobility is severely affected. I don’t have much quality of life, sometimes I feel like I just want to go to sleep and never wake up. How desperate is that?

TAU

A further TAU participant reported:

I’m now looking for full time work which is proving difficult. Having not worked for 13 years hasn’t helped. At the moment I’m working 2 to 3 days a week as a forklift driver. The agency I’m with only found me this job which I started 3 weeks ago. Having now come off benefits is now a shock to the system. A wake-up call, more financial pressure.

TAU

The financial issues were also highlighted by an IPS participant:

I’m still waiting for court appeal to get back my benefits so do not have any income!

IPS

Role of health care in creating disability

An individual in the IPS arm reflected on their perceptions of the role of health care in their pain and disability:

I still struggle with the fact that for the best part of 3 years I struggled to cope with the pain and belief that I needed a total knee replacement when everything was cleared in February 2018 with an injection. These 3 years did an awful lot of damage to me psychologically – I ‘lost’ 3 years of my life to excruciating pain.

IPS

Health-care professionals, pain clinic recruiters and employment support workers

Six focus groups took place and included nine PCPs who were involved in recruiting to the pilot trial in primary care and one PCP in the community pain services who recruited for the InSTEP pilot. One-to-one qualitative interviews took place with both of the ESWs half-way through the pilot trial and again at the end, and the same process was carried out with the pain service OT.

Thematic analysis led to the following headings: recruitment (the process, barriers and facilitators); intervention [IPS for people with chronic pain, setting up the IPS intervention (multidisciplinary working, providing information about the programme, practicalities, training of the ESWs)]; IPS activities (gathering information from participants, roles within IPS, intervention and tools, outcomes and role of IPS); the TAU booklet as the control; and feedback to health-care professionals.

Motivation of the employer

All three employers expressed the importance to them of being involved with work placements, describing their involvement altruistically as ‘giving back to the community’ (employer B). Interestingly, one employer explained that the directors of the company were particularly keen and supportive of this InSTEP placement because a family relative also suffered a chronic pain condition.

Benefits to the employer of providing these work placements

One employer commented that it had proved easier to obtain an individual prepared to work part-time hours (which was what the company needed) through this mechanism than on the conventional job market ‘where most people want full time’ (employer B). An employer raised that they perceived that senior management had persuaded them to be involved in the programme on the basis that it was ‘free labour for 6 months’ (employer A), but that they did not view it as such; rather they were willing to give a person unemployed with a health condition a chance. They acknowledged that it was a leap of faith and an unknown quantity when they offered the placement, but stated that it had been well worth the gamble in their opinion and they ‘would not hesitate to offer future work placements if funding became available again. It was brilliant’ (employer A). They observed the job satisfaction experienced by the individuals (this company had placed two individuals through SJP). The employers appeared to be genuinely pleased to see their employee flourishing at work. They did not, however, consider that they had done anything out of the ordinary or special in terms of offering work placements.

Importance of flexibility

The employers acknowledged that the employees on these work placements tended to have more need for flexibility (e.g. more sick days), but perceived this as being straightforward to accommodate and that flexibility was shown by all colleagues in the organisation. Good communication regularly between the employee and their manager was considered essential. The employers also talked about the fluctuating nature of pain conditions and recognised that, once they had understood this, they were able to accommodate it. The employers suggested that involving everyone in the organisation in supporting their employee had been ‘good for the business’ (employer B) and they did not perceive any problems with relationships with other staff. They felt that the employees appreciated the support and flexibility (e.g. with working hours and tasks) that had allowed them to settle in. Employer A reported that a worker had been overwhelmed by the job initially, but with the support of the manager had been helped ‘to break the job down into manageable tasks’, which had enabled this worker to cope and, eventually, to compile a ‘how to’ booklet to help himself and other new recruits to perform the work effectively.

Need to match the individual to the work placement

The employers were not obliged to take anybody on through this placement scheme, but instead were able to interview applicants interested in working with them. In both cases, they felt that this was important for the success of the placement and that they would offer such placements only if they could make this selection. One employer explained that other candidates could not be offered a work placement with this SME (e.g. for communication issues and because of a physical disability that simply could not be accommodated in that employment). Another point of this interview valued by the employers was the chance to assess an individual’s motivation for the work. They drew a parallel with workers who had been sent via Job Centre Plus who were coming for an interview simply because they felt that they had no choice and did not want the job at all. Motivation to work was seen as pivotal to success of the work placement. It was the employer’s view that both recruits placed with them had demonstrated a keen enthusiasm and desire to work, which had led to them both being taken on rapidly after starting the placement. The managers identified this motivation as a necessity if the placements were to be successful.

Support of the employment support worker

Slightly different views were expressed about the ESW support. Interestingly, one employer had decided to recruit their work placement individual very quickly after they had started (i.e. within 4–6 weeks) and they found that the ESW support was then rather quickly withdrawn. In contrast, the other employer noted that their ongoing support and engagement throughout the placement had been helpful on issues such as benefits to both the employer and the employee (who had also ultimately been taken on by the organisation).

Benefits for health of being at work

All three interviewees reported that they were absolutely certain that the employment placement had benefited the health of their employee. In one case, the employer said that their employee had advised them that ‘he had come off all his pain medication within weeks’ (employer A). One commented that work gave ‘a sense of worth’ and kept the mind occupied (employer B). One commented that the employee had themselves expressed how much the work helps and had stated that they ‘look forward to coming to work’ (employer B).