Abstract
BACKGROUND: Sharing information about hospital care with primary care in the form of a discharge summary is essential to patient safety. In the United Kingdom, although discharge summary targets on timeliness have been achieved, the quality of discharge summaries' content remains variable. METHODS: Mixed methods study in West Midlands, England with three parts: 1. General Practitioners (GPs) sampling discharge summaries they assessed to be "successful" or "unsuccessful" exemplars, 2. GPs commenting on the reasons for their letter assessment, and 3. surveying the hospital clinicians who wrote the sampled letters for their views. Letters were examined using content analysis; we coded 15 features (e.g. "diagnosis", "GP plan") based on relevant guidelines and standards. Free text comments were analysed using corpus linguistics, and survey data were analysed using descriptive statistics. RESULTS: Fifty-three GPs participated in selecting discharge letters; 46 clinicians responded to the hospital survey. There were statistically significant differences between "successful" and "unsuccessful" inpatient letters (n = 375) in relation to inclusion of the following elements: reason for admission (99.1% vs 86.5%); diagnosis (97.4% vs 74.5%), medication changes (61.5% vs 48.9%); reasons for medication changes (32.1% vs 18.4%); hospital plan/actions (70.5% vs 50.4%); GP plan (69.7% vs 53.2%); information to patient (38.5% vs 24.8%); tests/procedures performed (97.0% vs 74.5%), and test/examination results (96.2% vs 77.3%). Unexplained acronyms and jargon were identified in the majority of the sample (≥70% of letters). Analysis of GP comments highlighted that the overall clarity of discharge letters is important for effective and safe care transitions and that they should be relevant, concise, and comprehensible. Hospital clinicians identified several barriers to producing "successful" letters, including: juniors writing letters, time limitations, writing letters retrospectively from patient notes, and template restrictions. CONCLUSIONS: The failure to uniformly implement national discharge letter guidance into practice is continuing to contribute to unsuccessful communication between hospital and general practice. While the study highlighted barriers to producing high quality discharge summaries which may be addressed through training and organisational initiatives, it also indicates a need for ongoing audit to ensure the quality of letters and so reduce patient risk at the point of hospital discharge.