TP103A pilot patient preference RCT of admission to a women s crisis house compared with psychiatric hospital admission

A small number of women?s crisis houses have opened around the UK. They aim to offer a service which is more acceptable, less coercive and less stigmatising than psychiatric hospital admission for women. However, there has been little research examining whether these services really do lead to better outcomes for women in terms of their symptoms, quality of life, needs or satisfaction. We therefore propose to evaluate women s crisis houses, comparing them with psychiatric inpatient wards. There are two phases to this proposal. In phase one we will explore with women and staff from women s crisis houses and psychiatric wards in the same catchment areas, suitable points in their admissions for recruitment and randomisation. We will discuss what they consider to be the most important outcome measures e.g. quality of life or symptoms. We will also use published literature about research in mental health crises to help us to identify the best and most acceptable ways of inviting people to take part in research at such times.

In phase two we will use a study design known as a patient preference randomised controlled trial (RCT). With this type of design, women are first asked whether they have a strong preference for a particular type of service: if they do, they will receive the service they prefer. Women who do not have a preference will have a 50:50 chance of being admitted to either a crisis house or a psychiatric inpatient ward ? this process is called randomisation. This design has been chosen because we expect many women to prefer women?s crisis houses. A more traditional study using randomisation would not allow for such preferences: many women might therefore be unwilling to take part.

In this pilot study, we will test the outcome and cost measures, which will include questionnaires asking about quality of life, symptoms, satisfaction with the services, needs and contact with services. We will also explore the number of women s services nationally that could participate in a future larger study.

If this initial trial suggests that a patient preference RCT is a good method for evaluating crisis houses, we plan to conduct a larger national study including several more women?s crisis houses: this should give a clear answer as to whether this type of service is effective.

Women s crisis houses offer women an innovative alternative to psychiatric admission which may be less stigmatising and less coercive. Policy makers and service users favour their introduction throughout the country, but this is not yet supported by evidence from a rigorous evaluation. We propose to develop and test a framework for such an evaluation. Recruiting trial participants at the time of a psychiatric crisis poses various challenges, so that we need to carry out careful developmental work to establish the best way of doing this: this is phase 1 of our study, to be followed by a pilot patient preference trial in phase II.
Phase 1 ? the developmental work ? includes the following objectives:
1. To explore best methods of recruiting study participants at the point of a psychiatric crisis, including ways of managing both the practical and ethical difficulties this poses ? by systematic review
2. To examine pathways to care for women referred to women?s crisis houses or psychiatric wards, and find suitable points for recruitment and randomisation ? by quantitative data collection of a large series of admissions and more detailed qualitative interviews with a sub-group of service users and referring professionals.
3. To establish candidate primary outcome measures for appropriateness, acceptability, sensitivity to change and completeness of data collection ? by qualitative interview and quantitative data collection.

Phase 2 ? the pilot study ? includes the following objectives:
1. To establish the feasibility of a patient preference RCT study design for investigating the effectiveness and cost effectiveness of women?s crisis houses
2. To pilot the outcome and cost measures
3. To estimate recruitment and drop out rates, and the levels of outcome variability to calculate sample sizes for the main study.
4. To develop a user focused instrument designed by users to evaluate women?s crisis houses (user focused monitoring).

Method of pilot study: Patient preference RCT, 6 month recruitment phase with assessments at baseline and 3 month follow up. Depending on the outcome of Phase 1, measures will be allocated primary or secondary outcome status for the follow up.

A proposal for a larger multicentre patient preference RCT evaluating women?s crisis houses will be developed if this study indicates that it is feasible.