527. The Predictive Validity of Risk Factors for Colon SSI
Session: Poster Abstract Session: Surgical Site Infections
Thursday, October 3, 2013
Room: The Moscone Center: Poster Hall C
  • Colon SSI Risk Poster.pdf (401.3 kB)
  • Background: CMS mandates reporting of Colon SSI rates to NHSN. Young, et al. (2012) concluded that the National Healthcare Safety Network (NHSN) risk-adjustment model for colon surgery does not accurately adjust for colon SSI risk. We sought to develop a simple colon SSI risk-adjustment model using electronically-available variables from the NHSN model and published literature.

    Methods: A retrospective cohort of colon surgeries reported to NHSN from October 2010 through December 2012 was identified at an public academic level I trauma hospital. Variables required for the NHSN model, including age, duration, Anesthesiologists Society of America (ASA) score, wound class, anesthesia, endoscope, bed size, and medical school affiliation, were collected. Other potential risk factors described in the literature – including emergent and trauma designation, gender, and body-mass index (BMI) – were also collected. Univariate analysis to assess association with SSI used Chi-square tests of association, Fisher’s exact test, or the Wilcoxon rank-sum test where appropriate. Factors with a p-value of less than 0.25 were considered for inclusion in the model. Risk factors with over 10% missing data were excluded. A logistic model of SSI risk was constructed and assessed for goodness-of-fit and area under the curve (AUC). The AUC was then compared to NHSN models for colon, hip and knee arthroplasty, and vaginal and abdominal hysterectomy.

    Results: Two hundred eighty-four procedures and 17 SSI (6.0%) were identified. Medical school affiliation and hospital size were excluded because they are institution-specific. General anesthesia, endoscope, ASA, wound class, trauma, emergent, and gender were excluded for lack of association with SSI. BMI was excluded due to missing data. Ultimately, only duration and age were included in the model. The Hosmer-Lemeshow goodness-of-fit test indicated adequate fit (p=0.86). AUC (0.664) was greater than the NHSN model for colon surgery, and comparable to hip and knee arthroplasty and vaginal and abdominal hysterectomy.

    Conclusion: We demonstrated that most variables included in the NHSN model are poor predictors of colon SSI in a single center academic level I trauma public health hospital. Given that colon SSI reporting is mandated, an NHSN SSI risk-adjustment model with greater predictive validity is needed.

    Bryan Knepper, MPH, MSc, Heather Young, MD and Connie Price, MD, Denver Health Medical Center, Denver, CO


    B. Knepper, None

    H. Young, None

    C. Price, None

    Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 2nd with the exception of research findings presented at the IDWeek press conferences.