The link between evidence and service development is poorly understood. Naylor (Reference Naylor1995) outlined the limitations of evidence-based medicine and pointed out that in many areas of medicine there is not, and in some cases is un-likely to be, clear evidence for interventions. This means that even with the current drive to provide evidence-based guidelines in all areas of medicine, the lack of evidence will often mean that other factors must be influencing decisions about service development (Reference Woolf, Grol and HutchinsonWoolf et al, 1999).

In this study we aim to determine the factors that influence service development using the example of liaison psychiatry services. We chose this area because liaison psychiatry is a specialist service that operates in both general and mental health hospital trusts and is not a government priority area, and therefore may be representative of other small specialist services. Also, despite two College reports (Royal College of Psychiatrists & British Association of Accident and Emergency Medicine, 1996; Royal College of Physicians & Royal College of Psychiatrists, 2003) recommending standard structures for liaison psychiatry services, and government directives aimed at reducing inequality in service provision (Secretary of State for Health, 1998), little change has been observed in the past decade and there remains a wide diversity in service provision (Reference Mayou, Anderson and FeinmannMayou et al, 1990; Reference Howe, Hendry and PotokarHowe et al, 2003; Reference Ruddy and HouseRuddy & House, 2003). A recent meta-review of interventions in liaison psychiatry shows that there is little high quality systematic review evidence for effective interventions (Ruddy & House, further details available from author). This prompts questions about what factors are influential in changing service provision in liaison psychiatry.

Our overall aim was to determine the role of evidence and other factors in specialist service development in liaison psychiatry.

Method

We held two 90-min focus groups, each with seven participants. One group was formed from consultant liaison psychiatrists from six European countries, who managed clinical services and were members of the European Association of Consultation Liaison Psychiatry and Psychosomatics (EACLPP). The other group was multidisciplinary, with clinical liaison psychiatry practitioners from the Trent, Yorkshire and Northeast Liaison Network (TYNEL) in the UK. We used focus groups with a topic schedule rather than in-depth interviews, because we wanted interaction between the research participants to lead to a larger pool of factors (Reference KitzingerKitzinger, 1994). Informed consent was obtained from all participants.

The groups were audiotaped and transcribed verbatim. In addition, we took notes during each group. R.R. listened to all the tapes and read the transcripts several times. Themes were identified and then checked to be sure that they had emerged from the data rather than being forced on the data. The transcripts were then re-read for illustrations of the themes. At the end of this process A.H. reviewed and commented on the data analysis and a subsequent version was mailed to the participants for validation.

Results

The themes from the content analysis of the focus groups are presented in Table 1.

Table 1. Barriers and facilitators to liaison psychiatry service development

Level	Broad themes	Specific themes
National	Guidelines	National service frameworks
	National reports
	National professional associations	Training requirements
		Recognition of the specialty as important
	Government policy
	User lobbying
	Political understanding of evidence	How politicians prioritise evidence
	Shortage of doctors
Service	Internal factors	Proximity of the service to the general hospital
		Stability of the service in terms of structure, function and management
		Staffing levels, inappropriate staffing, high staff turnover
		Whether the service is managed by the general hospital trust or the mental health trust
		The funding and resources that the service receives
	External factors	Links (both formal and informal) to psychiatry and psychology
		Other liaison psychiatry service providers working in the same general hospital
		The need of general hospital patients to receive mental health input
		Training of professionals in the general hospital to recognise mental health problems and to refer
	Reputation factors	Complaints and untoward incidents
		Data collection and evidence of effectiveness of interventions
		Feeling ashamed at the poor quality of psychiatric service provided to the general hospital
		Interpersonal relationships between members of the liaison psychiatry team and individuals in the general hospital trust
		Level of demand of mental health input from general hospital services
		Promotion and marketing of the service
		Quality of service including response times, clarity of advice, being available, good communication
		Significant individuals with specific skills, interests or links to general hospital departments
		Promotion and marketing of the service
		Understanding and interest of general hospital staff

First, we asked specifically about the impact evidence has on service development. One group member summarised the consensus about the impact of published evidence:

However, there was also recognition that a lack of adequate evidence may act as a barrier to service development at local and national level, with poor or absent evidence being used to prevent development. Evidence through local data collection was generally seen as important to prove demand:

Discussion

The participants were unanimous that many things were more influential in the development of their services than clinicians’ desire to implement evidence-based interventions. At a national level, evidence is incorporated in national service frameworks and government guidelines. These play a role in influencing the shape of services but there are problems; for example the UK National Service Framework for Mental Health supports 24-h access to accident and emergency based services but mentions little about wider general hospital or self-harm services – despite large variation in provision of these latter services (Department of Health, 2003). As trusts become more autonomous, service level internal factors become more important. The most influential factors, however, appear to be local external factors and the reputation of the service, suggesting that service development results from a process of negotiation and marketing that is swayed by the desires of the main stakeholders.

Nationally there is much variation in the provision of healthcare interventions and also variation in provision of whole services – especially specialist services. This inequality of provision has recently been highlighted by the National Institute for Clinical Excellence (NICE) guidance for in vitro fertilisation treatment: although it is recommended that women between 23 and 39 years should be allowed three cycles of treatment, it will take until 2005 to make even one cycle of treatment available for all (National Institute for Clinical Excellence, 2004a). In the NICE guidelines for eating disorder services, the evidence that is available is mainly of level C quality and there is little evidence about what to do with more severe cases (National Institute for Clinical Excellence, 2004b). In situations like this it is clear that service development will be influenced by factors other than evidence, such as local politics and individual relationships and attitudes.

Our study examines pressures perceived by clinicians developing specialist services in a particular area – psychiatric provision in general hospitals – but the principle of what we have found is widely applicable. That is, a realistic appraisal of all forces impacting on service development is needed if we are to develop more rational service planning and have a clearer idea of how to make evidence fit effectively into that process.