Survival of death certificate initiated registrations: selection bias, incomplete trace-back or higher mortality?

Cases first notified to a Registry and successfully followed back have an apparently worse prognosis than cases registered in life. A simple approach can be used to assess whether this is due to selection bias, incomplete follow-back or intrinsically higher mortality. For the colorectal, breast and stomach cancers studied and for comparable registries, the main explanations are likely to be selection bias and higher mortality.