Many organizations use knowledge hierarchies (KHs): frontline workers attempt tasks and pass unresolved ones to experts. But the implications of using KHs are poorly understood, especially when firm revenues depend on who resolves tasks or the firm generates downstream externalities. We study a KH using 500,000 patient cases from an online primary care firm, linked to outcomes and costs across Sweden's healthcare system, exploiting assignment variation from temporary doctor congestion. Nurses resolve 70% of KH cases, passing the rest to doctors. Relative to doctor-only care, the KH modestly increases external primary care and emergency department visits but has no adverse effects on hospitalizations, earnings losses, or mortality. Its mean effect on systemwide (including downstream) costs is small but masks heterogeneity. Compared with doctor-only care, the firm's selective KH use saves 1.6% of online primary care costs; however, crucially, reassigning cases within 15-minute windows to minimize systemwide costs would save 6.8%.