Many clinicians and administrators in academic medical centers have ea
gerly embraced the idea of a comprehensive computer-based patient reco
rd (CPR), and either are contemplating implementing such a system or h
ave already begun implementing one. Most believe that CPR systems will
lead to greater productivity and clinical efficiency, and ultimately
to better patient outcomes at lower cost. But there is a gamble in all
of this. It is possible that in some settings a CPR system's potentia
l will not be realized because of poor implementation, poor organizati
on, or excessive and unanticipated costs. Given the high stakes associ
ated with CPR implementations, it is important that medical centers mo
ve more cautiously, always asking the question ''What if these systems
do not work?'' The author of this article considers worst-case scenar
ios of CPR deployment as well as evidence from industry and government
that undermines claims that CPR deployment will inevitably confer gre
ater productivity and efficiency. He challenges readers to think hard
about the cost - benefit ratios of both CRP systems and paper-based sy
stems, and to commit to an institutional ''bill of rights'' before for
ging ahead with CPR deployment.