To the Editor: Researchers have used various
methods to identify cause of death for case-mix classification at the end
of life, but there has been no systematic evaluation of the consistency between
these methods. The most common approach is to use the death certificate underlying
cause of death, but the limitations of this method include problematic data
quality,1 prohibitive cost of obtaining National
Death Index data, and identification of the underlying cause of death (the
condition that started the train of fatal events) may not be consistent with
a particular research objective. Some researchers have used health care utilization
data from their study data sets to identify cause of death for case-mix classification.
The methods have included identifying the last major diagnosis (last-diagnosis
method)2,3 and using the diagnosis
responsible for the most resource use near death (cost method).4,5 We
assessed the consistency among death certificate data and these 2 other methods.
Controlling for Patient Case Mix at the End of Life: Issues in Identifying
Cause of Death. JAMA. 2005;294(7):791-794. doi:10.1001/jama.294.7.793-c