Wound risk assessment in ventral hernia repair: generation and internal validation of a risk stratification system using the ACS-NSQIP.

BACKGROUND: There is a need for validated risk models to better stratify surgical site occurrences (SSO) following open ventral hernia repair (OVHR). The addition of more generalizable and validated risk models will serve to improve perioperative care in OVHR patients.
METHODS: We reviewed the 2005-2011 ACS-NSQIP databases identifying encounters for OVHR. The dependent outcome measure of interest was SSO, defined as superficial surgical site infection, deep infection, organ space infection, or wound dehiscence. Multivariate logistic regression of independently associated factors was performed and internally validated using a bootstrap technique. A composite risk score, the Hernia Wound Risk Assessment Tool (HW-RAT) was created using weighted beta coefficients. The HW-RAT was compared to existing models from the literature.
RESULTS: A total of 60,187 patients who met inclusion criteria were identified in the 2005-2011 ACS-NSQIP databases. The incidence of SSO in the study was 6.2% (N = 3,732). SSO risk factors were broken down based on rounded risk scores into the following groups: mild, intermediate, moderate, and severe risk. Severe risk factors related to operative time and degree of wound contamination. Moderate risk factors included class III obesity, component separation, dependent functional status, and inpatient hernia surgery. Patient stratification was performed based on total risk score into HW-RAT risk groups 1 through 5 which demonstrated significant discrimination between and across each group (P < 0.01, C-statistic = 0.71) with an incidence of SSO that ranged from 3.3 to 26.5%.
CONCLUSION: We present an internally validated risk model of SSO in OVHR (HW-RAT), which complements and builds upon current risk models. LEVEL OF EVIDENCE: Prognostic/risk category, level II.