Is the present definition of health care-associated pneumonia the best way to define risk of infection with antibiotic-resistant pathogens?