In 1995, Carl van Walraven and Anthony Weinberg reported in CMAJ on the assessment of quality in a discharge summary system.1 In a further report they noted that the quality of the reporting decreased as the length of the discharge summary increased.2 We evaluated the discharge summaries of 1712 sequential patients discharged from the respiratory division of Tsukuba University Hospital between April 1992 and December 2000.
Chief complaints, medical history, hospital course and discharge diagnosis were documented in all of the discharge summaries. However, physical examinations were not completely documented in 10.5% of the summaries, significant laboratory tests in 9.9% of the summaries and discharge medications in 3.4% of the summaries. The discharge summaries of the 171 patients who died in hospital were less likely to be complete than those of patients discharged alive in the categories of physical examination (83.0% v. 90.3%, p = 0.003) and significant laboratory tests (84.8% v. 90.7%, p = 0.014). However, the discharge summaries of the patients who died in hospital were not shorter than those of the patients discharged alive (1.48 v. 1.43 pages, p = 0.44).
For the records of patients who survived to discharge, summary length correlated significantly with completeness of reporting. The mean length of discharge summaries with complete reporting was 1.48 pages compared with 1.12 pages for summaries with incomplete reporting (p < 0.001).
We believe that discharge summaries should be routinely audited. This will ensure that problems with documentation are addressed and may improve completeness. It will also reinforce the importance of discharge summaries to physicians in training.