Research

Effectiveness

International research (randomised controlled trials (RCTs), reviews and meta-analyses) underpin the effectiveness of IPS as a method of reintegration into employment. The 23 RCTs that were carried out worldwide between 1996 and 2015 showed that IPS had found regular employment for an average of 55% of the total number of participants, compared with 23% in the control condition. In the USA, Australia and Hong Kong the percentage of placements was higher than that in Canada and Europe. This is probably associated with the differences in national regulatory frameworks (incapacity for work and state benefit systems). Varying employment markets or characteristics of the practice of IPS itself (differences between participants or degree of working in accordance with model fidelity and support in its implementation (Drake et al., 2012)) are also of influence. In the Netherlands this is 44% versus 25%. See the overview of RCTs.

Effectiveness in the long term

Studies on the maintenance of the positive effects over the long term are gradually becoming available. Has the IPS promise of staying in employment been realised? For example, in the study of Hoffmann et al. (2014) over 100 people divided into two groups, i.e. IPS support and other employment rehabilitation programmes, were followed over five years. The IPS participants did better and the positive effects seen at two years were still maintained at five years. If we look at number of people in the studies of Hoffmann et al. (2014), Salyers e.a. (2004) and Becker et al. (2007) (carried out over 5 years, 10 years and 8-12 years, respectively) who continued to work over the long term, it appears that 49% of the people who followed the IPS programme continued to work for at least half of the follow-up period. In the control conditions (ordinary employment rehabilitation) this was only 11%. See these results.

Research in the Netherlands

Eqolise

IPS is also effective in Europe, as was shown in an RCT that was carried out in diverse locations in six countries, including the Netherlands (Burns et al., 2007). Only people with psychoses took part in this trial. On average, participants in the IPS programme had regular paid work twice as often as participants in another employment rehabilitation programme (55% versus 28%). They kept their jobs for a longer period of time and worked more hours in a week. In the Netherlands (Groningen), the results were less convincing than in the other countries (33% regular paid work). These variances can be largely explained by the differences in unemployment between the regions in which the study took place, and differences in the social security systems. In the Netherlands, an implementation study had previously been carried out at four locations. The study showed that as long as a number of conditions in the organisation and financing were met, IPS is also feasible in the Netherlands (Van Erp et al., 2007; Giessen et al., 2007). These pioneers have paved the way for the wider implementation of IPS. Important conditions for success are meticulous preparation, sufficient time for the agency coordinator/programme leader and intensive IPS training for staff members.

SCION

In 2014, a Dutch RCT into IPS that had been carried out at various locations in the Netherlands was completed. SCION demonstrated that in the Netherlands too, people with severe mental illness were more often able to get regular paid employment through IPS than through standard employment rehabilitation. Within 30 months, 44% of the participants in the IPS programme had paid work compared with 25% in the control condition (Michon et al., 2014).

Meta-analyses and reviews

In a number of systematic overview articles and meta-analyses of multiple RCTs (randomised research in which a group in the IPS programme was compared with a control group that was mostly receiving other employment rehabilitation), the greater effectiveness of IPS was identified in outcomes such as getting a job, the total number of weeks worked or the hours of work of the longest period that people had a job. A recent overview containing a short profile of each article, grouped into study themes related to IPS, can be found here.

The most important recent meta-analyses come from Campbell et al. (2011) and Kinoshita et al. (2013) (containing 4 and 13 RCTs, respectively). Both of these analyses show that supported employment or IPS for people with severe mental illness delivers better results than regular employment rehabilitation. It is more difficult to identify the significant effects of secondary outcome measures such as quality of life and impact on psychological health. Kinoshita et al. (2013) report that research with a longer follow-up period is necessary in order to evaluate if the positive effects are maintained after a longer period (>2 years).

Cost effectiveness

IPS is more expensive and more intensive than other reintegration methods. Various studies have shown that the costs can be recouped. A follow-up RCT shows that after five years, IPS proved to be cost effective. Not only are the higher costs of employment rehabilitation recouped by lower expenditure on mental health costs and lower benefit payments, but people in the IPS group are also less often admitted to mental health institutions than people in the control group (Hoffmann et al., 2014). The return on investment in the IPS group was 53.5% and in the control group 18.3%.

The Netherlands: positive business case for IPS

In a social cost-benefit analysis carried out by the Erasmus University on behalf of the Ministry of Social Affairs (Schuring et al., 2016), an estimate showed that the IPS case projects followed in the study were cost-effective (albeit a cautious estimate due to all the hypotheses in the model). One example is IPS at VIP Amsterdam (agency offering early intervention in psychosis), which resulted in a saving of benefit payments by councils, a saving in care costs for the insurance companies and an increase in the client’s income from employment. This illustrates good collaboration between local council, care agency and health insurer. This practice in Amsterdam is cited as a good example by the Association of Netherlands Municipalities (VNG). A pathway to work for a person with a mental illness generates cost savings both for local councils and the health insurers. Read more about the funding of IPS.

Effects of IPS on outcomes other than work

DAccording to the overview study of Kukla & Bond (2013), RCTs do not show direct positive effects on outcomes that are not directly work-related. In a systematic review, Van Rijn et al. (2016) found a pooled, small, non-significant positive effect on both quality of life and symptoms. A few studies report an increase in self-reported health and quality of life because of IPS (Drake e.a., 2013), and a lower number of admissions to hospital (Hoffmann et al., 2014).
Furthermore, there are indirect positive effects on health and well-being from being in employment, e.g. more control over symptoms (Bush e.a., 2009; Kilian et al., 2012).

A systematic review of nine studies (Luciano et al., 2014) containing longitudinal datasets examined the influence that having employment has on non-work-related outcomes in people with severe mental illness. A number of studies found the following domains to have been positively influenced: self-respect (all studies in which this was measured), symptoms (4 of 7 studies) and satisfaction with life (3 of 9 studies). See this overview.