Stephen J.LurieMD, PhD, Senior EditorIndividualAuthorJody W.ZylkeMD, Contributing EditorIndividualAuthor
In Reply: Dr Smith makes a thoughtful comment
about using clinical predictive indices to modulate excessive medical litigation.
However, the prognostic index for elderly patients may not be sufficient for
this worthwhile goal.1 First, only a minority
of patients will score in the highest bedside quintile at which the estimated
risk of death is about 65%. Furthermore, an estimated risk of death of 65%
may not excuse an adverse event that follows a perceived shortfall of care.
Second, most fatalities will occur among patients whose estimated risks of
death are less than 50%. Such fatalities in individuals who were not expected
to die are likely to arouse demands for more explanation. Finally, the use
of statistical reasoning is sometimes problematic in legal courts because
of ambiguities over the appropriate referent population.2
For example, how many therapeutic triumphs in other patients can justify a
single lapse in 1 identified patient?
Redelmeier DA, Lustig AJ. Use of Prognostic Indexes for Determining Malpractice Liability—Reply. JAMA. 2001;286(14):1712-1713. doi:10.1001/jama.286.14.1709