i found a strange problem
when one of the host seems to be down and check_ping returns CRITICAL alert and when notification about this problem arrived it has an empty state field:
***** Nagios *****
Notification Type: PROBLEM
Date/Time: Wed Jan 16 10:56:27 MSK 2008
use generic-host ; Name of host template to use
use generic-service ; Name of service template to use
BTW, other critical services problems like fail to check disk usage via snmp or fail to check tcp port always return States correctly in notifications.