A major oncology center audited 15 trials and found that 51% of screened patients ultimately failed eligibility. Most were disqualified for not meeting inclusion criteria, discovered only after days of manual chart review and patient distress. The problem is structural: oncology protocols layer narrow biomarker requirements, performance thresholds, and molecular targets into eligibility gates that demand exhaustive chart mining. Manual screening takes up to an hour per patient, and coordinators spend over 30% of their time on this task alone. Disease can progress during the review window, and chart review accuracy studies show error rates exceeding 36% in clinical summarization tasks. Automated eligibility tools are now showing 85% to 90% reductions in manual review volume while maintaining over 90% sensitivity for truly eligible patients. These systems parse EHR data, match criteria in real time, and triage only uncertain cases to coordinators. The evidence spans pediatric and adult trials, breast cancer and solid tumors, and multiple academic centers. High screen failure rates cost trials weeks or months in delays. Automated pre-screening does not replace clinical judgment, but it narrows the field before coordinator hours are spent.