A retrospective audit of referral letter quality from general practice to an inner-city emergency department

ObjectivesOur primary aim was to analyse the quality of letters from general practitioners (GPs) to the ED as defined by two checklists: the authors' own nine-item list and a template published in 2013 by the New South Wales Agency for Clinical Innovation. The secondary aim was to determine if referral quality was influenced by letter format (handwritten or computer-generated) or urgency of the patient's condition (defined by triage category). MethodsThe present study is a retrospective audit of records of patients presenting to Royal Prince Alfred Hospital ED from a GP during February and March 2014. ResultsOut of 12199 ED presentations, 575 (4.7%) were recommended by a GP and 414 (72.0%) had a letter. Greater than 60% completion was achieved in 80.9% of letters compared to our own checklist, and in 6.3% compared to the Agency for Clinical Innovation checklist. Computer-generated letters were more likely to be fully legible and include an accurate medical history and medication list. Handwritten letters were less frequent (11.9%) but of poorer quality and legibility. Overall, less than half of letters contained results, psychosocial history or vaccination history. Referrals for patients assigned an urgent triage category were more likely to contain examination findings, but there was otherwise no difference in quality. ConclusionsReferral quality was influenced by letter format but not by urgency of the patient's condition. Omission of information from referral letters potentially risks patient safety. Handwritten referrals should be abandoned. Comprehensive electronic letter templates and regular updating of medications, comorbidities and allergies are encouraged.