Research and analysis

Individual Placement and Support - Alcohol and Drug study: main findings

Updated 26 January 2024

Applies to England

Summary

This is the first study of Individual Placement and Support (IPS) for adults in treatment for alcohol and drug dependence. It found that IPS helped more participants achieve employment in the open job market than standard employment support. There was evidence that IPS was effective in helping people with alcohol dependence and ‘other drug’ dependence find a job, but not for people with opioid dependence. Finding work in the 18 months of the study follow-up was the study’s primary outcome measure.

IPS was cost-effective for improving quality of life for people with alcohol and ‘other drug’ dependence, but not for those with opioid dependence. There was evidence that IPS was cost-effective at improving the number of days a person worked.

These promising findings informed the policy decision to make IPS available to more people with alcohol and drug dependence. It is currently being expanded across England.

This report summarises a Lancet journal article on IPS for alcohol and drug dependence. You can find more information about the study in the article.

Introduction

Background to IPS

Employment is essential in supporting people to improve their financial and social status. Job loss and unemployment are linked to poverty and ill health. Unemployed people with alcohol and drug dependence often say that employment is a recovery goal, but they are often excluded from the workplace, or struggle to get or keep a job.

Originally developed in the USA, IPS is an intensive employment support intervention to help people find and stay in work. IPS follows principles of:

• personal job-related preference
• rapid job search
• interview preparation
• in-work support

This intervention has been evaluated in populations with severe mental illness and other disabilities, and is more effective than standard employment support. But the effectiveness and the cost-effectiveness of IPS for people with alcohol and drug dependence was unknown.

The IPS-AD study

Professor Dame Carol Black’s independent review of drugs identified employment as an essential part of recovery for people with alcohol and drug dependence. With funding from the UK government’s Work and Health Unit, Public Health England (PHE) led the Individual Placement and Support - Alcohol and Drug (IPS-AD) study. After PHE was abolished in 2021, the IPS work was continued by the Office for Health Improvement and Disparities (OHID), which is part of the Department of Health and Social Care.

We ran the IPS-AD study between 2018 and 2021. It was the first multi-centre randomised controlled trial (RCT) of the effectiveness and cost-effectiveness of IPS for adults with alcohol and drug dependence. The study was initiated in 7 community treatment centres for alcohol and drug dependence in England. These centres were in:

• Birmingham
• Blackpool
• Brighton and Hove
• Derbyshire
• London Borough of Haringey
• Sheffield
• Staffordshire

Study participants:

• were adults (18 to 65 years)
• were unemployed or economically inactive for at least 6 months
• wanted to access the open job market
• had been in treatment for at least 2 weeks for alcohol dependence, opioid dependence or problems with other drugs

Study methodology

We randomly allocated participants into 2 study arms. One arm received standard employment support (or ‘treatment as usual’ (TAU)). The other arm received IPS.

We gave participants in the TAU arm information about the standard employment support services provided at the treatment centre and local area. This included support from Jobcentre Plus and any locally commissioned employment support programmes.

For participants in the IPS arm, we assigned them an IPS employment specialist who offered the IPS intervention for up to 9 months. If a participant found work, they were offered up to 4 additional months of in-work support. The total time that a participant could receive IPS in the study was 9 to 13 months.

IPS started with 4 sessions in the first month to:

• review the participant’s skills, experience and employment preferences
• offer help with writing or updating a CV
• implement a job search strategy
• prepare for interviews

After the first month, the frequency of IPS sessions was about every 2 weeks. In-work support was 4 meetings in the first month then every 2 weeks. The IPS-AD study team produced IPS training materials for alcohol and drug dependence along with specialist IPS organisations. Each employment specialist attended comprehensive IPS and practitioner skills training.

A UK Health Research Authority research ethics committee approved the study plan and materials. We pre-registered the analytical plans on the Open Science Framework and reported according to standard RCT guidelines. An independent trial steering committee and data monitoring committee oversaw the research. Both committees had particular expertise in this type of intervention and study. They included:

• experts by experience (people with experience of alcohol or drug treatment and recovery)
• experts in employment policy and support
• academics

Study implementation

Participants and what they received

After excluding people who were ineligible to take part in the study, we randomly allocated 1,687 participants to either TAU or IPS. There were 844 who received TAU and 843 who received IPS.

There were 3 clinical groups in the study:

1. People in treatment for opioid dependence, who were the largest group with 837 people.
2. People with alcohol dependence, who were the second-largest group with 610 people.
3. People who had problems with other drugs, which were mainly cannabis, cocaine and amphetamines. There were 240 people in this group.

The TAU and IPS study arms were similar on demographic and clinical characteristics. Most participants had severe dependence when they were admitted to their current treatment episode (68%).

In the TAU arm, 66% of participants received standard employment support. Of the participants in the IPS arm, 95% received at least one IPS session.

Among the 843 participants who received IPS, there was an average of 14.5 contacts with the employment specialist. On average, participants received IPS for 30 weeks.

Evaluating the delivery of IPS

After initiation, independent practitioners from the Centre for Mental Health and IPS Grow evaluated the quality of IPS delivered by each of the 7 centres at 5 to 7 months and 15 to 18 months. This included using standardised tools, such as the IPS fidelity scale. This 25-item tool assesses how closely the IPS being delivered sticks to the IPS model and its delivery principles. Reviewers assessed:

• staffing, including caseload size
• organisation, such as support offered from other team members
• service delivery issues, such as the type of jobs and employers found

The reviewers also offered written recommendations to each centre on how they could improve. The first fidelity review rated IPS delivery in all 7 centres as ‘fair’. But all services improved after 18 months, with 2 maintaining their ‘fair’ grade and 5 centres achieving ‘good’ fidelity.

The study also included an independent, external qualitative evaluation undertaken by RAND Europe and the Centre for Mental Health. This evaluation involved a series of interviews with:

• study participants
• IPS specialists
• treatment service staff
• other stakeholders, such as Jobcentre Plus staff and employers

Outcome measures

Primary outcome measure

The primary outcome measure to evaluate IPS effectiveness was whether the participant found work during the 18 months of study follow-up. We defined this as working in the open job market for at least one day.

Secondary outcome measures

There were also 12 secondary outcome measures, made up of 5 vocational outcomes and 7 health outcomes relating to alcohol and drug treatment.

The 5 vocational outcomes were:

• number of days of employment
• number of days to first job from involvement in study
• number of jobs during follow-up period
• length of longest-held job
• working in a single job for at least 13 weeks

The 7 health outcomes relating to alcohol and drug treatment were:

• alcohol consumption (for the alcohol group)
• opioid use and drug injecting (for the opioid group)
• main drugs used (for the ‘other drugs’ group)
• substance use dependence status (for all groups)
• number of days enrolled in treatment (for all groups)
• number of treatment episodes (for all groups)
• treatment status at end of follow-up (for all groups)

Effect of the COVID-19 pandemic and other employment services

Recruitment to the study was during the COVID-19 pandemic, so we assessed whether the impact of the pandemic had an effect on the probability of getting a job.

Also, since 67% of the IPS group also had at least one contact with Jobcentre Plus, we assessed the impact of IPS only and IPS with Jobcentre Plus support.

Measuring vocational and health outcomes

To measure vocational outcomes, we linked the participant’s National Insurance number (or their name, date of birth and postcode) to databases operated by the Department for Work and Pensions (DWP) and HM Revenue and Customs (HMRC). The study received data for 18 months before and after the date the participant enrolled in the study.

To measure health outcomes relating to alcohol and drug treatment, we linked information from the National Drug Treatment Monitoring System (NDTMS) using the same procedure. NDTMS is a database that collects information about drug and alcohol treatment in England.

Of the 1,687 participants taking part in the study, HMRC verified the employment records of 1,403 (83%) participants. We estimated the missing data for the remaining participants using a statistical procedure called multiple imputation.   

Cost-effectiveness outcomes

There were 2 cost-effectiveness outcome measures:

• a quality of life health outcome expressed as a quality-adjusted life year (QALY) (where one QALY is equal to one year of life in perfect health, measured in terms of the person’s ability to carry out daily life activities free from pain or mental ill health)
• the number of additional days a person receiving IPS worked during follow-up (compared with the TAU group)

We also estimated the cost of IPS, alcohol and drug treatment, and health services participants received.

Estimating cost-effectiveness

We estimated cost-effectiveness using a threshold value for an economic concept called willingness to pay (WTP). A WTP threshold value is the amount of money that a consumer is willing to pay for an improvement in an outcome. In this study, we evaluated each additional QALY using the HM Treasury’s WTP threshold of £70,000 for interventions like IPS. Following a previous IPS study, we evaluated each additional day in employment using a WTP threshold of £200.

Study findings

Primary outcome

For the primary outcome of participants finding work, HMRC confirmed that 207 of 697 participants in the IPS arm (30%) found work. In the TAU arm, HMRC confirmed that 175 of 706 participants (25%) found work. In our analysis using multiple imputation, this was a statistically significant effect for IPS.

While there was evidence that IPS was effective in helping participants in the alcohol and ‘other drug’ groups find work, it was ineffective for the opioid group. The alcohol and ‘other drug’ groups were respectively 48% and 45% more likely to find work.

There was no evidence that the COVID-19 restrictions reduced the IPS intervention effect.

From additional analysis, we found there was evidence that the odds of finding a job were greater for participants who received both IPS and standard employment support than for those who only received IPS.

Secondary vocational outcomes

Looking at secondary vocational outcomes among participants in the IPS and TAU arms who found work, there was no evidence that IPS was associated with:

• a reduction in the number of days to first starting a job
• whether participants had 2 or more jobs
• an increase in the total days employed
• the length of longest job held
• working in a single job for at least 13 weeks

Alcohol and drug dependence treatment outcomes

Looking at alcohol and drug dependence treatment outcomes, there was no evidence that IPS was associated with reduced alcohol consumption. There was also no evidence of a change in use of opioids, cannabis, cocaine and amphetamines.

There was no evidence that IPS was associated with improved rates of recovery from dependence or successful completion of treatment.

Cost-effectiveness

Overall, IPS was associated with a 1% probability that it was cost-effective at improving participants’ quality of life. This non-significant effect reflected the poor outcome among participants with opioid dependence. However, there was a 52% probability that IPS was cost-effective for participants with alcohol dependence. And there was 97% probability of cost-effectiveness for the ‘other drug’ dependence group.

For the outcome about additional days worked during follow-up, there was a 61% probability of cost-effectiveness for IPS. Over the 18-month follow-up period, a spend of £1,153 resulted in an additional 7 days of employment per IPS participant. The main cause of this outcome was the alcohol dependence group, which had a 99% probability of cost-effectiveness.

Independent qualitative evaluation

During the early implementation phase of IPS, the independent evaluators identified several positive findings of IPS for people with alcohol and drug dependence.

Positive findings the evaluators identified included the:

• flexible and individualised nature of IPS to address the complex and changing needs of people with alcohol and drug dependence
• emerging strategies used by employment specialists to engage with employers without disclosing their participant’s treatment status
• successful implementation of IPS in the study areas and improvement in fidelity scores during the study period
• use of learning from the study to enhance implementation and delivery of the IPS intervention in new areas

The evaluators also identified some areas for development, including:

• more frequent employer engagement
• more effective employer engagement
• the consistent engagement of people in effective in-work support

Expanding IPS for people with alcohol and drug dependence

Rationale for expansion

The government’s 10-year drug strategy sets out the commitment to improving employment opportunities and integration of treatment and recovery. The promising findings from the IPS-AD study informed the policy decision to make IPS available to more people with alcohol and drug dependence. It is currently being expanded across England.

Factors that informed this decision were as follows.

1. Evidence from Professor Dame Carol Black’s independent review of drugs that employment is an essential part of recovery for people with alcohol and drug dependence.
2. The challenges faced by mainstream employment support services in meeting the needs of people with alcohol and drug dependence.
3. The extensive evidence that IPS is more effective than standard employment support for people with complex health conditions such as mental ill health.
4. The finding from the IPS-AD study that IPS helped more participants get employment in the open job market than standard employment support. And a positive return on investment for the ‘alcohol’ and ‘other drug’ groups.

Applying the experience of the study is supporting the expansion of IPS, so that new services are able to reach good fidelity and high effectiveness more quickly than the initial study sites.

Monitoring post-study employment outcomes

We are closely monitoring the performance of the IPS expansion, mainly through verifying IPS clients’ employment outcomes by 6-monthly data linkage with HMRC. There have been 2 rounds of linking IPS and HMRC data for the people who received IPS interventions in the expansion after the study finished. This data linkage showed a considerable improvement in the primary outcome. Approximately half of the clients had at least one day of employment (compared with 30% of IPS-AD study IPS participants).

We have also seen improvements in secondary outcomes compared with the IPS-AD study findings, including:

• higher rates of work in a single job for at least 13 weeks
• longer time employed
• shorter time to get a first job

Improvements in vocational outcomes after the IPS-AD study demonstrate that well-established services can achieve better outcomes. And the learning from the study can be used to enhance newer services’ outcomes.

We have started development work, in collaboration with DWP, to explore a range of analytical options to enable us to further evaluate effectiveness.

DWP and OHID will continue to expand the IPS intervention across England so there can be an IPS service in every local authority. The expansion is happening alongside additional investment in treatment and recovery services, which was not in place at the time of the study.

The OHID IPS team continue to identify and use lessons learned from the study in the roll-out.

Acknowledgements

The authors would like to thank the IPS-AD study contributors, including:

• the participants
• the principal investigators
• IPS specialists
• centre management

We are grateful to the chairs and members of the trial steering committee and data monitoring committee for their support and guidance.

We also kindly acknowledge research funding from the government’s Work and Health Unit and support from:

• PHE
• Social Finance
• the Centre for Mental Health
• DWP