Individuals with severe mental illnesses (SMIs) experience serious social participation problems. Frequently reported issues are high unemployment rates (1–3) and difficulties participating in daytime activities such as education, unpaid employment or activities outside the home (4, 5). Participation in these activities is important because it facilitates recovery and is desired by most people with SMIs (6–8). Furthermore, competitive employment increases income and alleviates poverty (9). For instance, Levinson et al. (10) found that individuals with SMIs earned -on average- a third less than median earnings in over 19 countries in the Americas, Europe, the Middle East, Africa, Asia and New Zealand.

The low employment rates of people with SMIs not only have personal implications, but also implications for society at large. In the Netherlands and the UK, people with mental health problems represent the largest group (30–40%) of disability benefits claimants (11, 12). Therefore, increasing the social participation of individuals with SMIs may not only facilitate their recovery and improve life satisfaction, but also increase their productivity and reduce benefit payments, which is beneficial for society. Considering these problems, people with SMIs need structured supports with their social participation and this could be achieved through psychiatric rehabilitation. The aim of psychiatric rehabilitation is to “help persons with psychiatric disabilities increase their ability to function successfully and be satisfied in the environment of their choice with the least amount of ongoing professional intervention (13)”. It differs from standard care approaches with regard to its focus on personal goal attainment and (re)gaining various social and personal roles.

Several rehabilitation approaches have been developed to support individuals with SMIs with rehabilitation goals such as increasing their social participation. Some approaches target one aspect of rehabilitation, such as Individual Placement and Support (IPS) for competitive employment (14). Others, such the Strengths model or the Boston University Approach to Psychiatric Rehabilitation (BPR) (13, 15), focus on diverse rehabilitation areas.

Several trials have confirmed that BPR effectively improves social participation and functioning (16–19), and the approach has been implemented in many Dutch Mental Health Care (MHC) facilities. Nevertheless, dissemination is slow and BPR availability remains limited. An important reason for limited implementation is the fear that BPR will be time-consuming and expensive. Cost insights appear essential to further promote implementation, particularly as many MHC institutions are dealing with ongoing budget cuts.

To date, no studies on the cost-effectiveness of BPR have been conducted. Therefore, a randomized controlled trial (RCT) was conducted to investigate the cost-effectiveness and budget impact of BPR compared to an active control condition (ACC) for individuals with SMIs who have a wish for change regarding social participation. An active control condition was chosen to ensure that both conditions received equal amounts of attention and support.

It was hypothesized that BPR would be cost-effective compared to ACC because: (1) based on earlier studies, larger increases in social participation were expected combined with few additional costs, and (2) the methodology's positive effects on functioning would lower MHC costs.

All 188 participants were included in the economic evaluation. There were no significant baseline differences on any of the sociodemographic variables measured (Table 1).

Although BPR effectiveness regarding social participation has been confirmed in a number of RCT's when compared to standard care or an active control condition, dissemination and availability remain limited. This may be due to a lack of insight into the associated costs for MHC services providing this support, and into the benefits for society in terms of potential increases in quality of life and -if paid employment is obtained or kept- in productivity. Therefore, this study examined the cost effectiveness and budget impact of BPR in the area of social participation.

The main study investigating the effectiveness of BPR compared to ACC for individuals with SMIs in the area of social participation showed that both interventions effectively improved social participation after 12 months but no difference between the two interventions could be shown. Therefore, a priori expectations of BPR being cost-effective compared to ACC were low.

Results confirmed this, and BPR had a <50% probability of being cost-effective compared to ACC in the study population. Because the difference in QALYs was so small, the effects were not only statistically, but also clinically insignificant.

When costs collected over a 12 month follow-up period were divided into different types of expenditures, it became apparent that the highest total costs in both the intervention and control group were related to supported and sheltered housing, followed by inpatient care, outpatient care, and engaging in organized activities. The costs of supported and sheltered housing were considerably higher than those related to the other categories.

For the budget impact, the difference in intervention costs was small and non-significant, while differences in other healthcare costs and productivity costs were substantial, but highly uncertain. This resulted in an estimated budget impact that could range from cost savings to €190 million (see Table 3), depending on assumptions regarding uptake such as assuming participation of all eligible patients with SMIs aged 18–64, or only those who receive care from F-ACT (and variants in between). As there is no separate rate for support aimed at improving social participation, there would be no additional costs if BPR was implemented on a wider scale in mental health care institutions, from a health insurer perspective. This implies that any additional costs will have to be accommodated for by the mental healthcare organizations.

As this is the first study to investigate the cost-effectiveness of BPR in any goal area, it is difficult to compare the results to existing studies. Only studies targeted at paid employment have been conducted, and they have shown mixed results. For instance, Dixon et al. (37) compared IPS to enhanced vocational rehabilitation (EVR). The results were inconclusive as to whether IPS was more cost-effective than EVR. On the other hand, a study comparing IPS to standard vocational services (38) found that IPS had a high probability of being cost-effective. The general lack of cost effectiveness studies regarding psychiatric rehabilitation methods, and the mixed results of existing studies, clearly indicate that more research is needed in this area.

A number of strong points are associated with our study. First, an active control condition was used in which practitioners were explicitly instructed to offer support with rehabilitation goals. Many studies that investigate the effectiveness of protocolled rehabilitation methods use a control condition in which standard care but no active support is offered, which may overestimate the effect of protocolled rehabilitation methods. Second, studies on the cost-effectiveness of psychiatric rehabilitation programs often do not measure and analyze costs from a societal perspective, whereas we did. Including costs within as well as outside of the healthcare sector is important to providing appropriate incentives to take into account when making decisions on allocation or reimbursement of health-related resources (39, 40).Third, as far as we know, this is the first study to investigate the cost-effectiveness of a psychiatric rehabilitation approach focused on attaining a wide range of social participation goals as opposed to just competitive employment. Not all people with SMIs are willing to engage or capable of engaging in paid employment. It is therefore important to incorporate other forms of social activities in this type of study, such as unpaid employment. Fourth, the attrition rate was very low compared to other studies involving people with SMIs.

A limitation was the collection of cost data through interviews with service users, which could be considered less reliable than administratively recorded data or data obtained from public registries. However, production loss involving paid or unpaid employment is also difficult to calculate from registries. Furthermore, a study by Patel et al. (41) showed congruence between data obtained with self-reports and case records, and the TIC-P is a widely used self -report measure for the assessment of costs (42). A second limitation was the follow-up period of 12 months. A previous study on BPR effectiveness by a member of the current study group showed that rehabilitation goal attainment almost doubled in BPR from 12 to 24 months after enrolment (16). It should be noted that this study already found better results for BPR compared to ACC at 12 months. Perhaps more time was needed for psychiatric rehabilitation support, and perhaps BPR may need more time to show its full potential. RCTs in mental healthcare often have short follow-up periods for financial reasons. This may interfere with a reliable assessment of costs. Sharfstein and Clark (43) emphasize the importance of assessing costs over a sufficiently long period of time in order to include both short- and long-term outcomes. This may particularly apply to psychiatric rehabilitation methods as their effects may not be limited to the short-term but may become apparent with some delay. The National Health Care Institute even recommends a life-long time horizon for economic evaluations in healthcare (21). On the other hand, it could also be argued that generic health status is a relatively stable construct, which is not easily influenced by an increase in social participation. As data on this subject are lacking, it is difficult to predict what the effect of a longer follow-up period would be. A final limitation is that QALYs were derived from the SF-6D, which measures generic health and is not particularly tailored to individuals with SMIs (44, 45). In a study by Mulhern et al. (46), the SF-6D and the EuroQol-5D were found to be reliable in populations with common mental disorders, but possibly less sensitive to change for use among individuals with schizophrenia. On the other hand, a generic measure of quality of life allows broad comparison of results to other interventions and target groups.

Cost-effectiveness studies on psychiatric rehabilitation programs are scarce, while information on their effectiveness is needed in order to make decisions on further implementation. This is particularly important since MHC organizations continue to be faced with financial constraints. Future studies on psychiatric rehabilitation programs should therefore incorporate cost-effectiveness and budget impact analyses within a sufficiently long time horizon to fully capture the relevant costs and benefits. Future research should also focus on further exploring differences between subgroups to identify specific subgroups that may particularly benefit from BPR. Examples of such subgroup comparisons are unemployed or working in sheltered workshops compared to competitive employment or supported and sheltered housing compared to Independent living. Furthermore, the approach of “social return on investment” as the ratio of benefits (clients' work earnings) to total investment for each client, expressed as a percentage could be an interesting angle to explore.

The current study showed that BPR was not cost-effective compared to ACC. People in both conditions considerably increased their social participation, however at comparable costs. It would be unfair to discard the value of BPR based on a single cost-effectiveness study in a specific goal area. There were many factors that influenced the effectiveness of BPR compared to ACC such as the fact that ACC has become increasingly focused on rehabilitation in the past few years, and that most participants in our study sample wanted to obtain paid employment, which is the goal that proved the most difficult to obtain. These and other factors are extensively discussed in the main article on BPR effectiveness from the same study population (23). A process evaluation could have shed more light on the active components of successful psychiatric rehabilitation, and would be recommended for future studies.

As this was the first study to investigate the cost-effectiveness of BPR, and because we limited BPR to the area of social participation for which it was not specifically designed, more studies are needed before definite conclusions can be drawn on the cost-effectiveness of the method as a whole.

The datasets presented in this article are not readily available because restrictions apply to the availability of these data, which were used under license for this study. The data that support the findings of this study are available from the author WS, senior researcher, with the permission of Altrecht Institute for Mental Health Care. Requests to access the datasets should be directed to w.swildens@altrecht.nl

The studies involving human participants were reviewed and approved by University Medical Center Groningen Ethical approval reference number 2013/70. The patients/participants provided their written informed consent to participate in this study.

SS, TF, WS, JvB, JvW, and TvA contributed to the study conception and design and commented on previous versions of the manuscript. SS, WS, JvB, TF, and TvA performed material preparation and data collection and analysis. SS wrote the first draft of the manuscript. All authors contributed to the article and approved the submitted version.

This research was supported by a grant from the Netherlands Organization for Health Research and Development (ZonMw) (project number 837002006).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.