Discussion

A restricted data-set encompassing a small number of units and patients, all from one NHS Trust with a predominantly urban catchment area, inevitably limits the generalisability of the findings from this study.

The results offer some support to the Joint Commissioning Panel for Mental Health's⁵ updated guidance on the typology of rehabilitation wards, with a newly introduced division of HDUs (previously for between 1 and 3 years) into long-term high-dependency (high-support) units (up to 5 years) and high-dependency (high-support) in-patient rehabilitation units (up to 1 year). Furthermore, the Joint Commissioning Panel for Mental Health's advice⁵ on CRUs has increased the length of stay from 1 year to 1–2 years, as well as that for CCUs, from ‘several years’ to a more exacting but perhaps longer period of 5–10 years. In addition, where a local rehabilitation pathway provides either a high dependency (high support) in-patient rehabilitation unit or a community rehabilitation unit (rather than both), the expected length of stay may be up to 3 years.

Nevertheless, the data from this study suggest that the lengths of stay proposed for each type of unit possibly underestimates the time patients currently spend on rehabilitation wards. Two-fifths of patients were in hospital at the follow-up or the time of their death, and the median length of stay for those who were discharged from the HDUs was almost 10 years; this is in accordance with the cohort of very-long-stay psychiatric in-patients highlighted in the Centre for Mental Health briefing on rehabilitation,^{Reference Wright10} but seemingly at odds with the most recent guidance on the various types of rehabilitation wards by the Royal College of Psychiatrists,¹¹ which no longer includes CCUs.

Indeed, the recent National Institute for Health and Care Excellence guideline on ‘Rehabilitation for Adults with Complex Psychosis’¹² draws attention to the need for more research on people who spend very long periods of time in longer-term rehabilitation services. It may well be clinically appropriate for a minority of patients to remain in hospital for many years, but it is necessary to then ask whether units accommodating them, especially where locked and restricting access to the community, have a comprehensive range of psychosocial interventions, including appropriate psychological therapies and meaningful activities, which can enhance patients’ lives even within the confines of their setting.

The Care Quality Commission¹³ has reported the highly variable length of admissions of patients on rehabilitation wards, and expressed concern that some patients, particularly those in poorly defined ‘locked rehabilitation’ units,^{Reference Dye, Smyth and Pereira14} may be staying longer than necessary. The picture presented here, with widely differing durations of stay even within the same category of facility, reinforces this impression. Any analysis of the reasons for this must include a searching appraisal of the effectiveness of the interventions provided in these units. Definitions of rehabilitation services typically stress the importance of developing life skills, which help lead to greater independence and are critical to successful community living, with appropriate support.^{Reference Wolfson, Holloway and Killaspy9} Yet the evidence from the present study indicates that this may not be a realistic goal for many patients. Similar to the findings of a previous follow-up cohort,^{Reference Killaspy and Zis4} only around one in ten patients progressed to living independently. Occupational therapy, which seeks to improve everyday living skills and engage people in valued activities, has traditionally been at the core of rehabilitation practice. However, a recent major study of in-patient rehabilitation did not detect any clinical advantage of a newly designed activity-based intervention over standard care.^{Reference Killaspy, Marston, Green, Harrison, Lean and Cook15} As the authors acknowledge, the patients may have been too severely impaired to be able to sufficiently benefit; an evaluation of innovative new approaches is urgently required.^{Reference Meaden, Commander, Edwards and Cowan8,Reference Leff and Szmidla16}

One treatment domain clearly requiring improvement is the standard of physical healthcare in rehabilitation units; one in five patients had died by the time of follow-up, at a mean age of only 62 years. Comorbid physical health problems are common in this patient population and, as has been highlighted in a strategic review of mental health services,¹⁷ special attention is now crucial to tackle this premature mortality, involving more effective liaison with both primary and secondary care colleagues; the findings here give further weight to this assertion.

Recent commissioning guidance emphasises the importance of a system of rehabilitation services and cohesive rehabilitation care pathways.⁵ In this study, half of the patients who were discharged had a further hospital admission, with a quarter having had more than one episode of care on a rehabilitation unit. It is also notable that nearly two-thirds of the patients in hospital at follow-up were no longer in facilities run by the same NHS Trust, and just over one third were in units outside their original catchment area. The complex pathways of patients, even within in-patient rehabilitation services provided by just one NHS Trust, highlights the importance of effectively integrating a range of types of units while also being able to offer ready access to acute hospital admission. Certainly, the findings in this study contradict the notion of a simple linear flow of patients along a pathway from high-intensity settings through to the community setting and independence. Indeed, most of the patients living outside of hospital at follow-up were in some form of supported accommodation, pointing to the necessity for an approach to the commissioning and provision of services that embraces cooperation across the full gamut of residential options.^{Reference Edwards, Macpherson, Commander, Meaden and Kalidindi18}

Alongside the increasing plurality of in-patient rehabilitation and intensive community support services, the erosion of specialist community rehabilitation teams has arguably contributed to disruptions in patients’ rehabilitation pathways. There were no dedicated community rehabilitation teams in the NHS Trust involved in this study. Although a quarter of the patients were under an assertive outreach team, half were with a community mental health team, and a quarter had no community team involved in their care planning. This is consistent with the national picture, with only one in four NHS Trusts having specialty community rehabilitation services.¹⁹ It is recognised that the absence of dedicated teams compromises the ability of patients to progress from in-patient to community-based rehabilitation settings and, in turn, sustain these placements over the longer term. The development of these teams is highlighted as a priority in a recent report by Rethink Mental Illness and the Royal College of Psychiatrists.^{19,Reference Kalidindi, Killaspy and Edwards20}