Psychiatrists ' letters to general practitioners : choosing the right format

Letters from psychiatrists to general practitioners (GPs) should provide an appropriate content in a format which is easy to write and assimilate. For content, GPs have requested "key items" (diagnosis, suicide risk, treatment, prognosis and follow-up), and an explanation which is educational (Williams & Wallace, 1974; Pullen & Yellowlees, 1985; Margo, 1982). For format, GPs preferred a one page letter with two or three sub-headings in a survey based on one fictitious case (Ycllowless & Pullen, 1984). Real letters from psychiatrists in one district averaged one and three quarter pages with four subheadings (Prasher et al, 1992). GPs' opinions about actual

Letters from psychiatrists to general practitioners (GPs) should provide an appropriate content in a format which is easy to write and assimilate.For content, GPs have requested "key items" (diagnosis, suicide risk, treatment, prognosis and follow-up), and an explanation which is educational (Williams & Wallace, 1974;Pullen & Yellowlees, 1985;Margo, 1982).For format, GPs preferred a one page letter with two or three sub-headings in a survey based on one fictitious case (Ycllowless & Pullen, 1984).Real letters from psychiatrists in one district averaged one and three quarter pages with four subheadings (Prasher et al, 1992).GPs' opinions about actual changes in the format and content of letters sent to them have not been reported.
Unnecessarily lengthy letters waste the time of medical and secretarial staff, a point emphasised for us while working in the busy psychiatric emergency clinic of the Maudsley Hospital.We therefore intro duced a semi-structured format which was typed on a single sheet of A4 paper (for convenience referred to as the summary).To encourage reasonable uniformity we chose seven "key items" as sub headings: reason for assessment, diagnosis, risk of self-harm, other problems, treatment, follow-up, and prognosis.We aimed to test the acceptability of the summary to GPs, and to assess potential savings in the time of staff if the summary was used routinely.We also hoped to gain an impression of how the type of format could affect GPs' satisfaction.

The study
For all new patients presenting to the clinic over 12weeks, a summary and a letter were sent to the GP.The time spent dictating each one was recorded, and the length measured in lines of A4 paper.With each summary and letter, GPs received a questionnaire which used four point scales to rate usefulness and information content, with individual ratings of seven areas of information (Table I).GPs were asked which format they would prefer to receive and which they had filed, and additional comments were invited.Non-respondents were sent a further questionnaire.

Findings
Sixty-nine questionnaires were sent to 62 GPs by 15 psychiatric trainees.Forty-eight GPs returned 52 (75%) questionnaires, corresponding to 17 GP referrals and 35 self-referrals.Thirty-three of these patients had no previous psychiatric history.The time taken for dictation averaged three minutes for summaries and 19 minutes for letters, and the average number of lines of A4 paper was 11 for summaries and 35 for letters.
Forty (77%) summaries were either "definitely" or "extremely useful", as compared with 48 (92%) letters.All the others were "of some use".Twenty-three (44%) summaries were less useful than the accompanying letter, 17 (33%) equally useful, and 12 (23%) more useful.The loss of usefulness with the summaries was statistically significant (Wilcoxon matched pairs signed rank test, P = 0.027).Usefulness was unrelated to length, and examination of "extremely useful" letters revealed no common characteristics which might account for their superiority.Summaries were judged to provide significantly less information than letters in certain areas (Table I).One summary and nine letters provided "too much" information in at least one area, whereas 12 summaries and six letters provided "too little" information.Twenty-six (50%) summaries and 30 (58%) letters received no criticism of information content.
While 23 (46%) respondents preferred to receive both formats, the others were divided equally between summaries and letters.. Forty-five (87%) filed both of them.Eighteen GPs made additional comments, and while eight of them appreciated both formats, others held differing views about readability and the amount of information provided.

Comment
We found that for assessments carried out in an emergency clinic, summaries which were sub stantially shorter and quicker to dictate than unstructured letters were not necessarily less useful  to GPs.Loss of usefulness occurred with some summaries (less than half of them) but in spite of this, three-quarters of the total were still "definitely" or "extremely useful".The obligatory sub-headings of the summaries ensured that items of particular importance (e.g.suicide risk and follow-up arrange ments) were always included; these were sometimes missing from letters.The main advantages of the summary, therefore, were reliability of content and speed of preparation.Against this must be considered a modest loss of usefulness in some.Use of the summary alone is attractive because of the potential savings in time, but translation of this into improved service to patients (e.g shorter waiting time) or speedier communication needs confirmation.

Ratings of information content (numbers refer to the number of letters (L) and summaries (S) given the rating) Area of information
Approval of the summary must be cautious because non-responders may have been the least impressed with the new format.Also, to enable evaluation on the same clinical material GPs received both a summary and a letter for each patient; nearly half of GPs preferred to receive both formats and they might have been less satisfied had they only received the summary.Greater familiarity with the summary might lead to improved ratings but GPs' satisfaction would need to be confirmed with use of the summary alone.
Ratings of information content of summaries were often similar to that of letters, even when letters actually contained much more information (e.g. about family and personal history).The additional information appeared to be redundant, which is in keeping with previous studies in which GPs already knew much of the background information (Williams & Wallace, 1974;Margo, 1982).We found no consistent relationship between content and usefulness, perhaps because we did not ask GPs to specify what was useful about the communication.Of course, the meaning of "useful" is not necessarily the same for the GP and the patient.and better patient care cannot be assumed to follow from a GP's evaluation of a letter as "useful".
Psychiatrists may also have different criteria for usefulness and emphasise points regarded as less important by GPs.
How important is the format of psychiatrists' letters?Both of the formats we used were appreciated, but some GPs' comments indicated that their preference for either format would depend on the clinical circumstances, and few evaluations uniformly favoured one format.If the importance of the format is in facilitating communication appropriate to the needs of patient and GP, then flexibility is required.Inevitably pressure of work will lead to compromise -the structured summary used in this study is an example.But good communication requires more than a check-list and psychiatrists could be more creative and flexible in the format as well as the content of their letters.GPs might like to encourage them by stating their requirements explicitly in their referrals.