Case management programs, a specialized form of utilization review (UR), focus their interventions on catastrophic cases, which comprise the minority of patients while accounting for the majority of costs. Many case management programs base their case identification and review criteria on diagnosis alone, although research has shown that it is a poor predictor of length of stay. Using claims data from a large nationwide insurer, the authors developed an empirical approach to identifying potentially catastrophic cases. The findings suggest that, in addition to diagnosis, other factors such as age and treatment setting contribute to long stays and high costs and thus should be used to identify catastrophic cases for case management interventions. Strategies to target case management programs must be considered not only in light of their impact on cost but on the quality of care for individual patients.