Data sources
The National Nosocomial Infections Surveillance (NNIS) system.
The NNIS system was a voluntary network of U.S. hospitals collaborating with CDC to monitor HAIs from 1970–2005. Hospitals participating in NNIS provided acute care, had 100 or more beds, and a minimum of one full-time equivalent infection control practitioner for the first 100 occupied beds. Detailed methods of the NNIS system are described …show more content…
For ICU patients, we calculated HAI rates by using NNIS ICU data for each major site of infection (number of patient-days × infection rate/patient-days = number of infections) (see Figure).
Figure
Calculation of estimates of health care-associated infections in U.S. hospitals among adults and children outside of intensive care units, 2002
Step 3. Estimate of HAIs among hospitalized adults and children outside of ICUs.
NNIS HAI rates outside of ICUs were likely underestimated from 1990 to 1995; therefore, we used a different method from those above to estimate HAIs among adults and children outside of ICUs. Specifically, we estimated the number of infections from a single major site and then used the distribution percentage for that site from NNIS hospital-wide surveillance to extrapolate to the total number of HAIs in adults and children outside of ICUs. We chose the surgical site for our calculations because the number of surgical procedures is available for the U.S. population in the NHDS. We multiplied the number of surgical procedures in the NHDS by the surgical site infection rate from NNIS 2002 surveillance. From this estimate, we subtracted surgical site infections among newborns and among adults and children in ICUs, which yielded the total number of such infections among hospitalized adults and children outside of ICUs, i.e., 244,385. Surgical site infections accounted for 20% of all HAIs in NNIS hospital-wide surveillance; thus, we used that percentage to estimate the number of infections for other body sites (see