Clinical correspondence between general practitioners and specialists remains fundamental to the process of referral from primary care and transmission of management advice from consultants. However, both older and more recent studies indicate that opportunities for good communication are commonly missed.1-3 Newton and colleagues explored the views of general practitioners and consultants on the desirable content of letters, and proposed standards against which the content of letters might be audited.4 After a decade of increasing emphasis on good communication, clear records, and patient involvement, we repeated that study, and also audited letters written by doctors who responded to the questionnaire.
Participants, methods, and results
Questionnaires were sent to 360 general practitioners, 157 in areas served by the Royal Devon and Exeter Hospital and 203 in areas served by the Freeman Hospital, Newcastle, and to the consultants doing outpatient clinics (107 in Exeter and 101 in Newcastle), asking for their views on the desirability (always/usually important or sometimes/never important) of defined items4 in the referral letter and replies. The response rate was 84% for both general practitioners (304/360) and consultants (174/208); the table shows their views. General practitioners now attached greater importance to documenting three items in their letters than in 1992: medical history, findings on investigation, and whether the referral is new. An increased proportion of consultants concurred with the need for medical history, but fewer consultants viewed what the patient expects from the referral as an important item. Fewer general practitioners and consultants thought that the general practitioner's expectation was an important item. A higher proportion of consultants now thought that including a summary of the case history in the consultant's letter was important.
Table 1.
Views on, and contents of, general practitioners' and consultants' referral letters. Values are numbers (percentages) of respondents who viewed each item as “always/usually important” and number (percentage) of letters containing the items
|
General practitioners' views of importance
|
Consultants' views of importance
|
Audit of letters (2002)
|
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Item of content | 2002 | 1992 | Odds ratio (95% CI) for difference | 2002 | 1992 | Odds ratio (95% CI) for difference | Exeter | Newcastle | Odds ratio (95% CI) for difference |
|
General practitioners' referral letters
|
|
|
|
|
|
|
|
|
|
| Initial sentence stating reason for referral | 263 (93) | 104 (90) | 1.52 (0.72 to 3.21) | 155 (96) | 146 (92) | 2.30 (0.85 to 6.21) | 132 (99) | 138 (80) | 33.48 (4.52 to 247.89)* |
| Outline of history | 278 (97) | 115 (100) | 0 | 160 (98) | 149 (94) | 3.58 (0.97 to 13.26) | 133 (99) | 155 (90) | 14.59 (1.92 to 111.08)* |
| Important medical history | 283 (98) | 104 (90) | 5.44 (1.99 to 14.87)* | 159 (95) | 138 (87) | 3.02 (1.30 to 7.05)* | 85 (63) | 117 (68) | 0.80 (0.50 to 1.28) |
| Findings on examination | 267 (92) | 99 (86) | 1.88 (0.95 to 3.70) | 123 (74) | 126 (79) | 0.75 (0.45 to 1.26) | 78 (58) | 66 (38) | 2.28 (1.44 to 3.61)* |
| Findings on investigation | 259 (90) | 91 (79) | 2.36 (1.30 to 4.25)* | 131 (80) | 116 (73) | 1.47 (0.88 to 2.47) | 65 (49) | 47 (27) | 2.58 (1.60 to 4.16)* |
| Current medication | 271 (95) | 110 (96) | 0.88 (0.31 to 2.50) | 141 (87) | 146 (92) | 0.60 (0.29 to 1.24) | 69 (52) | 111 (65) | 0.58 (0.37 to 0.93)* |
| Psychosocial matters | 126 (46) | 49 (43) | 1.13 (0.73 to 1.75) | 96 (59) | 83 (52) | 1.33 (0.86 to 2.06) | 32 (24) | 50 (29) | 0.77 (0.46 to 1.30) |
| Allergies | 187 (65) | 85 (74) | 0.65 (0.40 to 1.06) | 92 (58) | 97 (61) | 0.88 (0.56 to 1.37) | 17 (13) | 21 (12) | 1.09 (0.55 to 2.17) |
| Whether/how patient was involved in referral decision | 104 (36) | 33 (29) | 1.38 (0.86 to 2.21) | 51 (32) | 59 (37) | 0.80 (0.50 to 1.27) | 22 (16) | 39 (22) | 0.67 (0.38 to 1.20) |
| What patient or relative has been told | 105 (40) | 39 (34) | 1.30 (0.82 to 2.05) | 80 (49) | 86 (54) | 0.82 (0.53 to 1.27) | 10 (8) | 13 (7) | 1.16 (0.49 to 2.73) |
| What patient or relative expects from referral | 106 (38) | 49 (43) | 0.81 (0.52 to 1.27) | 60 (38) | 87 (55) | 0.50 (0.32 to 0.78)* | 16 (12) | 36 (21) | 0.51 (0.27 to 0.97)* |
| What general practitioner expects from referral | 216 (76) | 101 (88) | 0.44 (0.24 to 0.82)* | 98 (60) | 127 (80) | 0.38 (0.23 to 0.63)* | 131 (98) | 154 (89) | 5.39 (1.56 to 18.61)* |
| Whether new referral or re-referred | 245 (87) | 87 (76) | 2.05 (1.18 to 3.57)* | 129 (79) | 121 (76) | 1.19 (0.70 to 2.01) | 128 (96) | 81 (46) | 30.02 (11.71 to 76.97)* |
| Consultants' letters | |||||||||
| Summary of history | 219 (73) | 79 (69) | 1.20 (0.75 to 1.92) | 149 (89) | 126 (79) | 2.17 (1.16 to 4.04)* | 132 (99) | 172 (100) | 0 |
| Findings on examination | 269 (90) | 102 (89) | 1.14 (0.57 to 2.28) | 146 (87) | 146 (92) | 0.59 (0.29 to 1.22) | 123 (92) | 145 (84) | 2.16 (1.03 to 4.52)* |
| Findings on investigation | 287 (95) | 105 (91) | 1.82 (0.79 to 4.18) | 150 (89) | 135 (85) | 1.40 (0.74 to 2.68) | 118 (88) | 75 (44) | 9.34 (5.11 to 17.08)* |
| Appraisal of the problem, including diagnosis where applicable | 291 (98) | 113 (98) | 0.86 (0.17 to 4.32) | 164 (99) | 157 (99) | 1.04 (0.15 to 7.51) | 132 (99) | 156 (91) | 12.69 (1.65 to 97.37)* |
| Management plan | 284 (96) | 114 (99) | 0.21 (0.03 to 1.62) | 163 (97) | 154 (97) | 1.06 (0.30 to 3.73) | 134 (100) | 172 (100) | 0 |
| What patient or relative has been told | 258 (86) | 105 (91) | 0.59 (0.28 to 1.21) | 142 (84) | 137 (86) | 0.84 (0.46 to 1.55) | 61 (46) | 91 (53) | 0.75 (0.48 to 1.19) |
| Time to follow up appointment | 262 (89) | 105 (91) | 0.78 (0.37 to 1.64) | 139 (86) | 135 (85) | 1.07 (0.58 to 2.00) | 129 (96) | 165 (96) | 1.09 (0.34 to 3.53) |
| Who saw the patient | 251 (88) | 98 (85) | 1.28 (0.68 to 2.40) | 147 (91) | 145 (91) | 0.95 (0.44 to 2.03) | 134 (100) | 172 (100) | 0 |
P>0.05.
Letters (including attachments) about two recent outpatient referrals from each consultant were audited, using uniform criteria for each item of content (table). The defined items were recorded more often in Exeter than in Newcastle by both general practitioners (six items) and consultants (three items). For two items, general practitioners in Newcastle recorded items more often.
Comment
In the past decade the views of doctors regarding the desirable content of letters written by consultants have changed little, but the desirable content of general practitioners' letters has changed somewhat. The audit showed that, despite the views they had expressed, general practitioners frequently did not include “important” items in their referral letters. Nearly all general practitioners considered documentation of medical history and findings both on examination and investigation as important, but these items were documented in only 27-68% of their letters. Consultants' letters more often contained the items they viewed as desirable, but only about half included what the patient had been told.
The study identified differences in the content of letters between Exeter and Newcastle, for both general practitioners and consultants. This indicates that there may be regional variation around the country in the thoroughness of communication which doctors expect.
As well as conveying information from one doctor to another, letters also form a valuable source of reference, evidence of the process of informed consent, and a medicolegal record. Some items may have important safety implications. Letters can also help to inform patients, and it will soon be normal practice in the NHS to send copies of letters to patients.5 For the guidance of healthcare practitioners and the wellbeing of patients, a more rational and consistent approach to defining the desirable content of letters is required.
We thank all the general practitioners and consultants who participated in this study, and the medical secretaries who helped by providing letters for the audit.
Contributors: BC conceived, designed, supervised, and coordinated the study and wrote the paper. KV, EW, CvdW, and RE collected and collated data; MW supervised and coordinated the Newcastle upon Tyne contribution to study. JC did the statistical analysis. BC and JC wrote the paper; all authors revised the manuscript and approved the paper. BC is guarantor.
Funding: None.
Competing interests: None declared.
Ethical approval: Chairmen of research ethics committees in Exeter and Newcastle upon Tyne.
This article was posted on bmj.com on 30 March 2004: http://bmj.com/cgi/doi/10.1136/bmj.38058.801968.47
References
- 1.Jacobs LGH, Pringle MA. Referral letters and replies from orthopaedic departments: opportunities missed. BMJ 1990;301: 470-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Rutherford R, Gabriel R. Audit of outpatient letters. BMJ 1991;303: 968. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Grol R, Rooijackers-Lemmers N, van Kaathoven L, Wollersheim H, Mokkink H. Communication at the interface: do better referral letters produce better consultant replies? Br J Gen Pract 2003;53: 217-9. [PMC free article] [PubMed] [Google Scholar]
- 4.Newton J, Eccles M, Hutchinson. Communication between general practitioners and consultants: what should their letters contain? BMJ 1992;304: 821-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Working Group on Copying Letters to Patients. Copying letters to patients: a report to the Department of Health and draft good practice guidelines for consultation. Leeds: Department of Health, 2002.