Can general practitioner referral letters for acute medical admissions be improved?

General practitioner (GP) referral letters for emergency medical admissions should contain enough information to ensure that patients are managed safely and effectively.

Our aim was to assess the quality of referral letters for acute medical admissions.

GP letters from 300 consecutive acute medical admissions were prospectively assessed by the admitting doctor and then independently assessed by a senior house officer (SHO), and a senior registrar. A random sample of 25 were assessed by a general practitioner. Content and legibility were evaluated for demographic details, current history, past history, social history, drugs and allergies, in order to grade the overall quality as "excellent", "good", "adequate" or "inadequate".

Two hundred and ninety one letters were received, of which 208 (71%) were from the patients' own GP practice and 83 (28%) from a co-operative or deputising service. The admitting doctor rated 38 (13%) as excellent, 108 (37%) as good, 96 (33%) as adequate and 49 (17%) as inadequate. Kappa values for overall quality between the admitting doctors and the other doctors (senior house officer, senior registrar and our general practitioner) were 0.32, 0.26 and 0.44 respectively, representing fair to moderate interobserver agreement. Co-operatives or deputising services used proformas more often than GP practices (65/83, 78% and 17/208, 8% respectively, p < 0.01). There was no significant difference in overall quality between GP practices and co-operatives or deputising services, or between letters received on headed note paper or proformas.

Most letters are satisfactory for the individual management of acutely ill patients, but collaborative work could usefully be undertaken to address the factors which result in the production of sub-optimal referral letters.