Referral letters from general practitioners to hospitals in Sri Lanka; lack information and clarity

Abstract

BACKGROUND: Referral of patients to hospitals, specialists and other institutions is an essential part of primary health care. In many instances the referral letter is the sole means of communication between general practitioners (GPs) and specialists/hospital doctors. This study was planned to assess the quality of referral letters sent by general practitioners to out patient departments (OPD) of hospitals. METHODOLOGY: This descriptive cross sectional study was conducted in four hospitals of different levels of care provision in Sri Lanka. Referral letters received by the OPDs during a period of 2 weeks were analyzed. A check list to extract data was developed based on the items of information expected in a referral letter and legibility. Each item was assigned a score. This scoring system was validated using a panel of experts by means of Delphi method. Maximum score possible for a letter was 30. RESULTS: A total of 461 letters were analyzed. Items of information most often present were; to whom referred (96.7%), symptoms (91.5%), reason for referral (90.2%) and date (88.9%). The least often present items were; family history (0.2%), history of allergy (1.1%) and social history (1.7%). Most of the words were not legible in 42.3% of the letters. Median score of the sample was 16 (mean=15.69) Mean score of structured form letters was 18.61 (n=33) and in conventional letters it was 15.53 (n=428). The observed difference was statistically significant (z=-3.544, p<0.01). DISCUSSION: Most of the letters did not have the required information and legibility was also poor. Expected benefits of a referral letter to the patient, recipient and the referring doctor will not be achieved due to these short comings. Form letters were comparatively better. Measures should be taken to improve the content and clarity of referral letters.