GPs' views on discharge summaries

have used the model of a specialised unit and patients are still managed on general medical wards or in accident and emergency depart ments in many cities. A single centre in a city facilitates the psychiatric or social evaluation that is essential (Royal College of Psychiatrists. 1983). It ensures that nurses in charge of the beds have psychiatric nursing experience and that medical, nursing and social work staff can receive regular in-service training in the manage ment of deliberate self-poisoning. Thus, liaison and training are improved and we hope that the increased opportunities for research will help in the prevention of poisoning and in developing more effective patterns of care for patients who have deliberately harmed themselves.


GPs' views on discharge summaries John Dunn and Stephen Burton
Aims and method After gaining the impressionthat our discharge summaries were too long for local general practitioners (GPs), we proposed to produce an abbreviated summary.We sent a questionnaire to a sample of GPs asking which aspects of the current summary were helpful.Results Although many GPsconsideredthe summary overly long, a majority considered all the items to be at least 'very helpful'.

Clinical implications Although previous work on
discharge summaries has indicated a demand for brief, focused reports it is important to establish local GP priorities before planning changes.

Effective communication between psychiatrists and general practitioners (GPs) is clearly essen tial in the care of people with mental health
Psychiatric Bulletin (1999), 23, 355-357  problems in the community.This is especially so following discharge from hospital, when the care plan and treatment are likely to have been changed.However, psychiatrists' letters and discharge summaries have been criticised for being too long and not meeting the needs of GPs (Pullen & Yellowlees, 1985;Craddock & Craddock, 1989).A number of studies have investigated what GPs want in letters (Williams & Wallace, 1974;Pullen & Yellowlees, 1985) and in discharge summaries (Walker & Eagles. 1994;Soloman et al, 1995).The tendency has been to favour short summaries under well-defined headings concentrating on diagnosis, drug treat ment and management and prognosis.We had gained the impression from informal feedback from GPs that our full discharge summaries, often more than three sides of A4 paper, were too long.We proposed introducing an abbreviated discharge summary including only those sections considered important in previous studies.We decided to assess the views of local GPs before piloting the new summary.

The study
We designed a questionnaire that broke the existing summary down into 18 sections.GPs were asked to indicate if they found the items 'essential1.Very helpful', 'helpful', 'helpful at times' or 'irrelevant'.We indicated those sections we intended to put into the abbreviated summary and asked if. in their opinion, any sections should be added to the summary or if any sections should be left out.We also asked if the current summary was 'too long', 'about right' or 'too short'.
From a list of in-patients over the 18 month period from July 1996 to December 1997 we identified a sample of GPs who had recently seen a discharge summary.The questionnaire was then sent to them with a covering letter and a copy of the discharge summary on their patient.A follow-up phone call was made two to three weeks later requesting the return of the questionnaire.

Findings
Twenty-one questionnaires were sent out, of which 15 (71%) were returned.The responses to the importance of the various parts are given in Table 1.Eight (53%) GPs indicated that the summary was too long, while seven (47%) felt it was about right.Only four GPs listed parts they would include in the abbreviated summary (one of whom indicated they would always want the full summary), most left this section blank.

Comment
This study was primarily intended to gather information from local GPs to guide us in imple menting an abbreviated discharge summary.By its nature therefore it is small in size, but the results were significant enough to justify a change to our plans.
GPs were nearly equally divided on considering the current summary too long or about right.However, every part in the existing summary had a clear majority describing it as essential or very helpful.It is possible we biased people's opinions by indicating those items we considered most important.These certainly form the bulk of items considered essential.From our study it seems that local GPs did want to know the full case histories of their patients.In addition there were a number of contradictions with previous reports (Walker & Eagles, 1994: Solomon et al 1995).Our GP sample were more likely to find history of presenting complaint essential than in previous studies.However, information given to patients or carers and prognosis was considered less helpful, although this had been emphasised as an important aspect before.It is possible to speculate that this could be related to the characteristics of our patient group and our type of service.Most in-patients will have extensive contact with the community team before and after admission.GPs are thus less involved in referral to hospital and thus interested in the reasons why.They are also less involved in immediate support afterwards, expecting this information to be given by the community team.GPs may also consider prognosis in the psychogeriatric population too obvious or too imprecise to be helpful.
The published literature on psychiatric dis charge summaries indicates they want short reports containing only certain items necessary for the continued care of the patient and family.However, our survey of local GPs indicates that although they wish the summary was shorter they do want all the information it contains.The immediate practical effect of this study was for us to abandon our plans to produce an abbre viated discharge summary.Instead by exercising tight editorial control and some change in presentation we hope to produce a complete summary on two sides of A4.Thus we will meet the information requirements of our GPs but also shorten the summary length.The wider implica tion is that while there is an extensive literature on GP requirements in psychiatric discharge summaries and letters it is important to estab lish local priorities, which may be different before planning changes in communication.

Table 1 .
Number of responses to each part of the summary (n=15) Indicates those parts we have intended to include in the abbreviated summary.