Surveillance system finds missed surgical wound infections
SSI rates drop under new procedures
A surgical site infection (SSI) surveillance system is picking up many infections that otherwise would be missed at Baystate Medical Center in Springfield, MA. The system, developed by nurses, also enables clinicians to tabulate meaningful overall SSI rates, as well as surgeon-specific rates, says Myrna Schulte, RN, MA, CIC, infection control manager at Baystate.
Under the system, the infection control department dispatches a relatively simple questionnaire and patient listing to surgeons, asking them to note any surgical wound infections that occur within 30 days after discharge, meaning they are most likely nosocomial infections.
The post-discharge component is but one component of a comprehensive surveillance system that includes several check points along the way to confirm infections and validate data against patient records and hospital lab reports. (See flowchart for procedure, p. 33.)
Somewhat skeptical initially, Schulte compiled one year of data and found that post-discharge reports were identifying many infections that would have been lost to follow-up. "About a third of all the SSIs that were picked up would not have been picked up otherwise," she says. "I was surprised."
Since the surveillance program went to effect in April 1991, nearly 30,600 procedures have been tracked, which accounts for about 25% of the total number of procedures performed in the hospital’s operating room, says Schulte. SSI rates have dropped from about 4% to less than 1%. Schulte notes that the so-called "sentinel effect" is very much in evidence. Namely, validating past findings and reporting individual rates to surgeons will result in lower infection rates. Baystate tracks 16 different procedures.
The response rate to the questionnaires was 97%, which Schulte attributes to strong support from the surgery department’s leadership. Because Baystate had no statistical guidelines by which to measure its results, they relied on self-made benchmarks determined during the first two years of the surveillance program, says Schulte.
Infections worse among readmitted patients
With the additional data generated by the questionnaires, Schulte is now able to determine that SSIs can be broken out into three segments:
• About a quarter were detected in the hospital after surgery but before the patient was discharged.
• Another quarter were detected when patients were readmitted to the hospital with post-surgical infection.
• About half occurred among discharged patients who did not return for readmission.
In general, readmitted patients had more serious infections than those who recovered at home and did not seek hospital readmission. While the latter were often minor infections, few were noted in hospital records before the advent of the post-discharge surveillance program.
"That makes it worth your while to do post- discharge surveillance," Schulte says. "We’re finding half of all the infections in the hospital through our regular surveillance, but only half of those occur during the initial hospitalization, and I expect that fraction to decrease because we are discharging people more quickly. The other half of those are readmissions. They went home and came back with an infection. The remainder we would not have heard about at all, other than through responses to the questionnaire."
System accounts for variables
As epidemiologists have noted, failure to track post-discharge SSIs undermines the validity of surveillance data and makes reporting surgeon-specific rates a questionable exercise. Likewise, variables such as patient risk factors for infection must be considered if meaningful rates are to be reported back to surgeons a practice that has been shown to reduce the number of nosocomial infections. Such concerns have been addressed in the Baystate program, where risk factors such as duration of surgery, type of procedure, wound class, and American Society of Anesthesiologists score are figured into rate computations.
In addition to control charts showing quarterly SSI rates, a standard risk chart of data from a rolling 12-month period is prepared. This gives an accurate look at trends over time and adjusts for risk as well as small numbers.
In general, the more risk factors, the greater the likelihood of post-surgical SSI. Surgeons who treat high-risk patients can expect the number of post-surgical infections to be higher than can surgeons who treat patients at lower risk. That approach, plus inclusion of the post-discharge data, allows the compilation of meaningful rate data by the Infection Control Department.
"The physicians who were saying you can’t compare apples to oranges’ now believe that the SSI rates are accurate," Schulte says. "That seemed to calm the waters. We have a lot of validity built into the system. Occasionally, someone has protested and said this doesn’t look right.’ We’re willing to take another look at the data. I think a rapport had to be built."
Adapting from past success
At Memorial Hospital in Worcester, MA, a system of post-discharge surveillance of wound infections is evolving from a tracking tool applied to discharged patients who receive peripherally inserted central catheters (PICCs). Using data collection sheets routed to the various post-discharge care personnel, Barbara Crocker, RN, MSN, CIC, infection control practitioner, set up a surveillance system to monitor the use of the PICCs. The data sheets are filled out initially by the health care professional who inserts the PICC. It is then forwarded to the hospital’s infection control department, which in turn sends it to the appropriate agency assigned to care for the patient: a nursing home, home health agency, or intravenous therapy provider.
Crocker says she intends to adapt the system to track surgical wounds infections. "We hope to initiate a surveillance tool similar to the one used for PICCs to let us know if post-surgical wounds develop, and if so, how they are doing. Did it heal well? If not, was there an identifiable cause for infection?"
Current methods of surveillance for post-discharge infections are inadequate, she adds. Discharged surgical patients who develop a wound infection may visit a physician who may start them on antibiotics without taking or culture. If a culture is taken, it may be done in the office and sent to a lab not associated with the hospital. "So we never find out about the infection. We need a better way of tracking these infections that may be related to the patient’s hospital stay."
For now, the infection control tracking reports are initiated only by the hospital-based Memorial Home Health Agency, but Crocker intends to expand the system to include other practice groups that care for discharged patients.