Individual placement and support for people with moderate to severe mental illnesses or substance abuse disorder: a systematic review
Systematic review
|Published
The purpose of this systematic review was to investigate the effects of IPS compared to other interventions, on the following outcomes: obtaining regular employment, time in employment, income and quality of life for people with moderate to severe mental disorders or with drug addiction.
Key message
Individual Placement and Support (IPS) is an intervention that aims to help people with mental disorders who want to work, to obtain regular employment (paid employment at a regular workplace, fulltime or part-time) as quickly as possible, with close followup. We updated a systematic review from 2017 which ex-amined the effects of IPS. We updated one comparison from this review: standard IPS vs. other interventions, for people with moderate to severe mental illnesses or drug addiction. We identified 9 new studies and one update of a previously included study. This adds up to 30 studies being summarized in this review in total. These are mainly from the U.S. and Europe, with a follow-up period from six months to six years. We com-piled the studies in meta-analyses and assessed the confidence in the results.
The results for people with moderate to serious mental disorders showed that:
• People with moderate to severe mental illnesses who receive IPS, compared to other interventions, may be twice as likely to obtain employment.
• It may be that time in employment increases somewhat.
• There is likely a small increase in income.
• For quality of life, IPS likely does not have a noticeable difference.
As for people with drug addition, we cannot draw any conclusions about the effect of IPS, as we only found one small study with these kinds of participants. There is a need for more studies that examine the effect of IPS for people with drug addiction, as well as other studies on IPS.
Summary
Introduction
Individual Placement and Support (IPS) was developed especially for people with severe mental disorders and later expanded to also include moderate mental disorders. This means that those who want to seek employment are helped as quickly as possible into regular work, with close follow-up. An underlying idea of IPS is that people themselves know what is important to them and what goals they have for their lives. Shared decision-making is an important principle and a way to ensure that these goals are heard. IPS is a comprehensive program where several criteria must be met, including that the program is integrated into treatment teams for mental health.
Objective
The purpose of this systematic review was to investigate the effects of IPS compared to other interventions, on the following outcomes: obtaining regular employment, time in employment, income and quality of life for people with moderate to severe mental disorders or with drug addiction.
Method
We updated one comparison from a previously published systematic review from the Norwegian Institute of Public Health (NIPH; 2017) on the effects of IPS: standard IPS versus other interventions (including usual care). After clarification with the client, we narrowed down the inclusion criteria for population and outcomes, but otherwise kept the same methods.
We conducted a literature search in ten databases in November 2022. Titles and abstracts, and subsequently potentially relevant full-text publications, were assessed against the inclusion criteria by two team members, independently. We assessed the risk of bias in the included studies and performed meta-analyses where possible. We assessed certainty of evidence and confidence in the results using GRADE (Grading of Recommendations Assessment, Development and Evaluation).
Results
We included 10 studies from 12 publications, of which one study report was an update of a study included in the previous review. From the original 2017 review, we retained 21 studies. We thus summarized the results from a total of 30 studies; 29 randomized controlled trials and one cohort study. In the randomized studies, between 37 and 541 participants were included, making up a total of 5 074 participants, the majority of whom were men. The cohort study, that analyzed register data included 23 369 participants. The studies included adults aged 18 to 65, except for one study which included young people aged 15-25. The population in all studies were people with mental disorders, primarily severe (schizophrenia, schizoaffective disorder, bipolar disorder, recurrent major depression, borderline personality disorder and post-traumatic stress disorder), except for one of the randomized studies that included people under methadone treatment. For participants with mental disorders, the meta-analysis on employment rates from 28 randomized trials showed a relative risk of 2.02 (95% CI 1.70 to 2.40) (low confidence). The cohort study from the Netherlands utilized vocational data from a national register, which included a sample of 513 recipients of IPS and almost 23,000 recipients of traditional vocational rehabilitation services. The authors found that, at 30 months, 47% of the participants in the IPS group were employed, compared to 39% in the matched control group.
The study with 45 people under methadone treatment examined the effect of IPS for individuals with moderate to severe opioid abuse. The participants were randomly allocated to receive IPS plus usual treatment or to a waiting list plus usual treatment. After 12 months, the authors found that 11 out of 22 in the intervention group, versus 5 out of 23 in the waiting list group, had had regular employment (RR 2.30 (95% CI 0.95 to 5.55)).
For the outcomes time employed, income and quality of life, the results were respectively: SMD 0.63 (95% CI 0.36 to 0.89), SMD 0.25 (95% CI 0.12 to 0.37) and 0.16 (95% CI 0.06 to 0.26). The summary table below shows the results for each outcome and the degree of confidence after we assessed each domain according to GRADE.
|
Summary of findings: |
|||||
|
IPS compared to other intervention for persons with moderate to severe mental illnesses |
|||||
|
Patient or population: Persons with moderate to serious mental illnesses Setting: Psychiatry or social service Intervention: Individual placement and support (IPS) Comparison: Other intervention |
|||||
|
Outcomes |
Anticipated absolute effects* (95% CI) |
Relative effect |
№ of participants |
Certainty of the evidence |
|
|
Risk with control |
Risk with IPS |
||||
|
Employment |
298 per 1 000 |
613 per 1 000 |
RR 2.02 |
4736 |
⨁⨁◯◯ |
|
Time in work |
- |
SMD 0.63 higher |
- |
3324 |
⨁⨁◯◯ |
|
Income |
- |
SMD 0.25 higher |
- |
2133 |
⨁⨁⨁◯ |
|
Quality of life |
- |
SMD 0.16 SD higher |
- |
1330 |
⨁⨁⨁◯ |
|
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio; SMD: standardised mean difference. Explanations: a. Several studies with unclear risk of bias (unclear allocation procedure and lack of blinding of participants) b. Inconsistency/Heterogeneity 80-90% |
|||||
Discussion
We investigated the effect of standard IPS versus other interventions for people with moderate to severe mental disorders or with drug addiction. For people with mental disorders, IPS may be twice as likely to result in regular employment, and the time in employment may increase somewhat, when compared to receiving other services. This probably leads to a small increase in income. Although the effect on quality of life was also in favor of IPS, it is likely too small to make a meaningful impact. As for people with drug addiction, we cannot draw any conclusions about the effect of IPS, because we only identified one small study with these kinds of participants.
Our results correspond to a large extent with the previous reviews in this field. However, it is important to be aware that there was great variation among the effect estimates, both for obtaining employment and time in employment. We therefore have low confidence in the effect estimates for these two outcomes. When there is a large variation in the results between studies, it becomes difficult to predict the effect of each new IPS intervention that is initiated. Furthermore, it varied between studies how the outcome of obtaining employment was measured. In some studies, the amount of work required to be counted as paid employment during follow-up was very small (e.g. one day's paid work). The outcome of employment, therefore, does not say anything about the workload during the follow-up period, but only shows the relationship between the groups. Some authors did not report what the criteria were for being defined as being employed, and those who obtained data from registers did not report what the registers' criteria were for being registered as being employed.
The evidence base in this review is about effects of IPS versus other interventions for a population with moderate to serious mental illnesses, but notably only on selected outcomes. Our objective did not include effects on outcomes such as hospital admissions, mental symptoms, or costs. Neither did the population include other groups apart from persons with mental disorders or drug addiction. The intervention was also limited to standard IPS. We excluded all studies which we judged as being about IPS with various types of reinforcements or improvements or interventions that could be labelled IPS-inspired. The evidence base in this review is, thus, not suitable for drawing conclusions about such kinds of interventions.
Conclusion
IPS, as a vocational rehabilitation that is part of an interdisciplinary treatment, has the potential to help people with moderate to severe mental disorders to work. It is also possible that the intervention may lead to a somewhat longer time in normal work and probably to a slightly higher income. For quality of life, the effect of IPS is probably so small that it will not make any noticeable difference for those who participate in an IPS intervention. The effect of IPS on helping people with drug addiction to work is uncertain.
