This dataset shows the Centers for Disease Control and Prevention National Healthcare Safety Network (NHSN) risk adjusted standardized infection ratios (SIR) for surgical site infections (SSIs) reported by hospitals for 24 operative procedures that are listed in the Data Dictionary. The SIR is calculated by dividing the number of observed infections by the number of predicted infections. The number of predicted infections is calculated using SSI probabilities estimated from multivariate logistic regression models constructed from NHSN data during a baseline time period 2006 through 2008, which represents a reference population’s SSI experience. To enforce a minimum precision criterion, an SIR is calculated only if the number of observed infections is at least one.
SIRs are based on the date of procedure, not the date of the SSI event. All procedure and infection data entered before May 1, 2015 for surgical procedures performed between January 1 and December 31, 2014 were downloaded from NHSN for this dataset. The SIR used for this dataset is a risk adjusted SIR that considers only complex SSI identified during hospital admission or readmission to the same hospital following inpatient surgeries. The procedure counts and infection counts submitted by each hospital are displayed along with an SIR, the 95 percent confidence interval for the SIR, and the statistical comparison as follows: No difference – no difference in number of observed and predicted infections, High – more infections than predicted, or Low – fewer infections than predicted.
Reporting of SSI data by California general acute care hospitals to CDPH and public reporting of hospitals’ risk-adjusted SSI rates by CDPH‘s HAI Program were mandated by Health and Safety Code 1288.55 (a)(3), (b)(2), and (c)(1). The frequency of SSIs varies depending on modifiable risk factors, such as surgical technique and infection prevention measures, and non-modifiable risk factors, such as underlying patient illnesses, whether the surgery was an emergency, and whether the wound was contaminated prior to surgery. The distribution of non-modifiable risk factors among patients, referred to as patient case mix, can vary widely among different hospitals. To report SSI rates that allow for meaningful comparisons between hospitals, it is critical to adjust for the differences in patient case mix. Instead of a risk-adjusted SSI rate, NHSN produces a standardized infection ratio (SIR) that compares the number of SSIs reported by a hospital to the number that is predicted based on a surgical procedure category-specific algorithm derived from national data reported to NHSN from 2006 through 2008.
NHSN produces two SIRs for each surgical category, one that includes all SSIs (including superficial incisional) and the second, the Complex A/R SIR, that includes only deep incisional and organ/space SSI identified during the index hospital admission or on readmission to the same hospital that performed the inpatient surgery. For this dataset, data was extracted from CDPH that used only the Complex A/R SIR, hereafter referred to as SIR. Ten surgical procedure categories were selected for presentation based on the number of SIRs generated. SIRs were generated more frequently for high volume surgical procedure categories, surgeries that have a higher predicted number of infections, or both. As such, these were the surgical procedure categories where infection prevention strategies would have the largest impact. These ten surgical procedure categories, abdominal hysterectomy, cesarean section, coronary artery bypass graft, colon surgery, hip prostheses, knee prostheses, open reduction of fracture, spinal fusion, small bowel surgery and bile duct, liver and pancreatic surgery, account for 66% of SIRs for all the risk-adjusted SP categories. The data was collected through reports from hospitals in California for 24 common procedures plus 5 procedures without risk adjusted comparisons: heart transplant, kidney surgery, ovarian surgery, pacemaker surgery, and spleen surgery.
The datasets used from CHHS Open Data Portal has been modified.