To the Editor. —
Escarce and Kelley's study on the limitations of Acute Physiology and Chronic Health Evaluation (APACHE) II scores at predicting hospital death lead time at time of medical intensive care unit (MICU) entry fails to state an important point that might modify the replication study they call for. Why not calculate an adjusted APACHE score, retrospectively using hospital admission physiology data? This should be readily available in patients' charts at time of transfer and should control for much of the lead-time therapy, allowing explanation of any selection effects that would muddle analysis of the utility of the APACHE system.