Bounds for a joint distribution function with fixed sub-distribution functions: Application to competing risks
Arthur V. Peterson
DOI: https://doi.org/10.1073/pnas.73.1.11
IF: 11.1
1976-01-01
Proceedings of the National Academy of Sciences
Abstract:This paper gives sharp bounds for the joint survival function G ( t 1 , t 2 ,..., t r ) ≡ P ( X 1 > t 1 , X 2 > t 2 ,..., X r > t r ), and for the marginal survival functions S j ( t ) ≡ P ( X j > t ), j = 1,2,..., r , when the sub-survival functions S j * ( t ) ≡ P ( X j > t , X j = min k =1,2 ,..., r X k ) are fixed. Theorem 1 gives the bounds for r = 2, and Theorem 2 gives the bounds for general r . Theorem 3 applies the result to the competing risks problem, and presents empirical bounds based on the observations. Finally, an example illustrates the bounds.