These interactive databases can be searched by hospital, by infection, and by peer group. The Readerâ€™s Guides below provide background information important to the interpretation of the data.

**Understanding the peer groups.** Because there are differences among hospitals in the services provided and the number of patients treated, hospitals are grouped into "peer groups" to assist the reader in recognizing similar hospitals. Each peer group is comprised of hospitals that are similar in the complexity of services offered (including percent of surgical procedures performed) and the number of patients treated. Click here to access the

2005 Peer Group Directory,

2006 Peer Group Directory or the

2007 Peer Group Directory.

**Understanding device-associated and non device-associated infections.** There are several types of infections that can occur with or without the presence of a medical device. For example, patients with indwelling urinary tract catheters may contract urinary tract infections, those on ventilators may contract pneumonia, and patients with central lines may contract bloodstream infections. These infections are referred to as device-associated. Infections that occur when a device is not present are referred to as non device-associated. Beginning with Quarter 4, 2005, urinary tract infection, pneumonia, and bloodstream infection data is broken out for device- and non device-associated infections. Data associated with the total number of infections is also reported. For example, data for urinary tract infections includes data for the total number of urinary tract infections, the number of device-associated urinary tract infections, and the number of non device-associated urinary tract infections.

**Understanding mean and median length of stay and charges.** Both the mean and the median are averages. The mean length of stay represents the number of days a patient would have been in the hospital, if all patients had an equal length of stay. The mean charge represents the amount that a patient would have been charged for their hospital care, if all the patients had equal charges. The mean is the measure most often referred to as the average. The median length of stay represents the midpoint of all the lengths of stay for all patients in a particular hospital. In other words, half of the stays are longer in length than the median and half are shorter in length than the median. The median charge represents the midpoint of all charges for all the patients in a particular hospital. In other words, half of the charges are more than the median and half are less than the median. Both the mean and the median include extreme values, also known as outliers. Because outliers have more of an effect on the mean than the median, the mean offers greater insight regarding the presence of extreme lengths of stay or charges. On the other hand, the median offers greater insight into mid-range lengths of stay or charges.

**Understanding the relationship between hospital charges and reimbursements.** In almost all cases, hospitals do not receive full charges from private insurance carriers or government payors, because insurance companies and other large purchasers of health care usually negotiate large discounts. During 2006 and 2007 on an average basis, across all inpatient hospital cases statewide (not just infection cases), hospitals were reimbursed or paid for approximately 27% of established charges. (The statewide/regional revenue-to-charge ratios were derived from the annual net inpatient revenue and inpatient charges provided by each hospital as part of their annual financial filings to PHC4.)