Feachem's Kaiser study not credible