Sampling and non-response bias on health-outcomes in surveys of the oldest old.

Sampling and non-response bias on health-outcomes in surveys of the oldest old.