Mortality rates as a measure of quality and safety, "caveat emptor".

The objective of this study was to demonstrate the impact of a single ICD-9 (International Statistical Classification of Diseases and Related Health Problems, Version 9) code on the observed-to-expected mortality ratios for acute care hospitals, calculated using administrative data. The study was a retrospective analysis of mortality data and prospective measurement of the impact of a change in coding on expected mortality rates. Measurement included overall mortality observed-to-expected mortality index for 2 hospitals and rate of use of the palliative care ICD-9 code. The main result was that both retrospective and prospective applications of this single ICD-9 code significantly reduced observed-to-expected mortality ratios. Both regulators and hospitals need to be aware of the impact of the quality of coding on publicly reported quality and patient safety data.