Implementation and establishment of nationwide surgical site infections surveillance in Japan

Abstract

ISSUE: There used to be no database against which hospitals can compare their surveillance data for surgical site infections (SSIs). In 1998, the Japanese Nosocomial Infections Surveillance (JNIS) system was established by adopting the U.S. National Nosocomial Infection Surveillance System (NNIS). Its aim was to create the database of SSIs in Japan through surveillance. PROJECT: Hospitals that were willing to submit data were invited to participate in this project. Data collection began in 1999. Data were submitted via e-mail, using Excel™ spreadsheets. Feedback was given about once a year by paper reports to each hospital. RESULTS: The number of participating hospitals increased from eight in 1999 to about 80 by the end of 2004. A total 20,948 cases were surveyed from November 1998 through December 2003. The primary operative procedure was abdominal surgery because the persons responsible for surveillance are mainly gastrointestinal surgeons. Unlike in the United States, in Japan there are insufficient numbers of certified infection control practitioners, and surgeons are interested in performing surveillance on their own. Risk stratification is a necessary factor comparing hospital data against data in the JNIS system. We used the NISS SSI risk index system and it suited our data. On other hand, we modified system by dividing large bowel operation (COLO) into two different categories (COLN for colon, and REC for rectum) because many Japanese surgeons think that the risk of SSIs is lower for colon surgery than for rectal surgery due to differences in anatomy and the complexity of the operations. The overall SSI rates in COLN and REC procedures were 14.8% and 17.3%, respectively. Even though the difference was not statistically significant (p=0.12, chi square test), it seemed to support the surgeons' perceptions. Also, the SSI rate for esophageal operations, which we separated, was 21.6% (19/43) and much higher than that of OGIT category (5.2%), where it is included in the NNIS system. Feedback data included SSI rates by procedure and by other factors such as age, gender, and SSI rate-stratified NISS risk index. Questionnaires sent out in 2003 showed overall satisfaction among participating hospitals. LESSONS LEARNED: A nationwide SSI surveillance system was implemented in Japan, and we were able to create a national reference data in the past 5 years.