Estimation of confidence limits for the cumulative probability of survival in life table analysis

Rothman, K.J.

Journal of Chronic Diseases 31(8): 557-560

1978


ISSN/ISBN: 0021-9681
PMID: 711848
DOI: 10.1016/0021-9681(78)90043-7
Document Number: 453768
Interval estimation for the cumulative probability of human survival in life table analysis is usually based on Greenwood's formula and does not reflect the asymmetry of the sampling distribution. Confidence limits are set by assuming the life table estimate of cumulative survival probability follows a binomial distribution with the number (N) equal to the effective sample size in the life table analysis. Wilson's quadratic limits for the binomial proportion offer an accurate non-iterative approximation to the binomial limits which would be convenient to apply routinely in life table analyses.

Document emailed within 0-6 h
Secure & encrypted payments

Estimation of confidence limits for the cumulative probability of survival in life table analysis