Is disturbed sleep a clinically useful marker to determine the suicide risk in patients with posttraumatic stress disorder?