Random error and undercounting in birth defects surveillance data: implications for inference.

Random error and undercounting in birth defects surveillance data: implications for inference.