Cost-effectiveness analysis of the Un Chez Soi d’Abord randomized controlled trial: a Markov model

Abstract

Objectives Providing permanent supported housing (rent subsidies), integrating clinical and social services to homeless people may represent a financial barrier for building a healthier European public policy. Our objective was to conduct a cost-effectiveness analysis of the French Housing First program. Design and methods Between 2011 and 2014, homeless people with severe mental illness from 4 French major cities were randomized to Housing First (HF, n = 353) or Treatment as Usual (TAU, n = 350) and followed over 24 months. Housing stability and survival were assessed. Costs incurred by health (ED visits, hospital admissions, physician consultations), social (shelters, supported accommodations, other residential structures), justice services (court appearances, detention and penitentiary centres), and welfare benefits were collected over 24 months. Markov models of 6-month transition probabilities were compared among HF group (offering an independent house and support from outpatient mental health and social team) and TAU group (receiving pre-existing dedicated homeless-targeted programs and services). Three Markov states were defined: stable housing (at least 90 nights in an independent housing), unstable housing (i.e. inverse) and death. A societal perspective and a life-time horizon were chosen. Robust sensitivity analyses were performed, including Tornado diagram and Monte-Carlo simulations. Results There is evidence of changes (P<.001) in housing stability (246% more days housed) for the HF group compared with TAU group. From the societal perspective, HF was €10,163 (3%) more costly than usual care. ICER was €5.32 per additional night spent in independent housing. In sensitivity analyses, HF was less costly and more effective in 28% of iterations. Conclusions HF proves to be cost-effective in homeless people with mental disorders and may assist decision-makers in defining conditions for further implementation of these innovations in Europe. Key messages Housing First model proves to be cost-effective in homeless people with mental disorders. Housing First model was less costly and more effective in 28% of ieration compared to treatment as usual.

Cost‐effectiveness of a ‘Housing First’ programme implemented in Spain: An evaluation based on a randomised controlled trial

“…For the evaluation of the H abitat programme, the aim was to provide evidence on the usefulness and viability of this example of HF in Spain. Based on the results, our research is in keeping with the majority of evaluations that find HF programmes cost-effective (Basu et al, 2012;Blood et al, 2017;Bretherton & Pleace, 2015;Chalmers McLaughlin, 2011;Goering et al, 2014;Latimer et al, 2019;Lemoine et al, 2019;Martinez & Burt, 2006;Mason & Grimbeek, 2013;Perlman & Parvensky, 2006;Srebnik et al, 2013). It has been verified that H abitat provides a large positive difference compared with its alternatives in terms of the aspects evaluated, both subjective (quality of life) and objective (avoid homeless situations or providing institutional accommodation).…”

“…Although the sample does not match the size of the largest studies carried out internationally (Rosenheck et al, 2003: N = 460;Goering et al, 2014Goering et al, : N = 2148Lemoine et al, 2019: N = 703), it far exceeds those of the studies carried out in Spain (Bernad et al, 2016: N = 62;Housing First Guide Europe, 2016: N = 38). One of the characteristics of homeless population studies is the difficulty of tracking people over time (traceability).…”

“…For the second line of research, it has been found that people who participate in HF programmes make less use of resources such as emergency accommodation or emergency medical services and are less likely to be arrested. This translates into a reduction in non‐specific costs and greater efficiency compared with the TAU (Basu et al, 2012; Bretherton & Pleace, 2015; Busch‐Geertsema, 2012; Busch‐Geertsema, 2013; Culhane, 2008; Larimer et al, 2009; Lemoine et al, 2019; Srebnik et al, 2013; Tsemberis, 2010). Another relevant aspect is the cost savings in adults with mental disabilities and/or chronic mental illnesses due to less use of social services (Brown et al, 1991; Gulcur et al, 2003; Stefancic & Tsemberis, 2007).…”

The effect of “Housing First” on mobile and digital media usage by people experiencing homelessness: A program evaluation based on a randomized controlled trial in Spain

“…This is exacerbated in some items, including some ICT-related ones, through a reduced response rate. Although the sample does not match the size of the largest studies carried out internationally (Goering et al, 2014; n = 2,148; Lemoine et al, 2019: n = 703; Rosenheck et al, 2003: n = 460), it far exceeds those of the studies carried out in Spain (Bernad et al, 2016: n = 62; Pleace, 2016: n = 38).…”

Evaluación de un programa Housing First: buscando una solución viable para transformar la vida de las personas sin hogar

“…Así, se han señalado efectos económicos positivos relevantes e incluso se ha constatado un menor coste que servicios de asistencia tradicionales (Gilmer et al, 2010;Ly y Latimer 2015;Pleace y Bretherton, 2019). También se comprueba el ahorro de costes en servicios no específicos de personas sin hogar, las denominadas externalidades (Mangano, 2009) y el carácter coste-efectivo del modelo (Wright y Peasgood, 2018;Latimer et al, 2019;Lemoine et al, 2019). Más recientemente, en nuestro país, se ha descrito la obtención de ahorros netos unidos a la mejora de la calidad de vida en personas usuarias de recursos específicos de la red Housing First (Sánchez et al, 2019).…”