Mandatory Brief Operative Wound Classification Alone Does Not Increase Accuracy of Nursing Wound Classification

Author(s):
Samuel Zolin; Joseph Golob Jr M.D.; Brian Young; Vanessa Ho; Esther Tseng; Jeffrey Claridge

Background:

Accuracy of nurse-documented surgical wound classification (NDWC) is critical. These data are used to calculate hospital standardized infection ratios, which are utilized for a variety of quality programs, including national quality metrics, hospital quality comparisons, and compensation incentive programs. In February 2018, our hospital added a mandatory wound classification field to the surgeon’s brief operative note (BONWC) with the goal of improving documentation accuracy.

Hypothesis:

We hypothesized that mandatory BONWC would be associated with improved accuracy of NDWC.

Methods:

A single-center retrospective cohort study was performed on all adult patients undergoing colon operations from January 2017 to August 2018. Colon cases were selected given variation in wound classification (WC) and use in quality metrics. For each case, the final dictated operative note was reviewed to determine the correct WC by Centers for Disease Control criteria and was considered the gold standard. The BONWC and NDWC were compared to the wound class derived from review of the operative note to calculate overall accuracy of each WC. We used Fisher’s exact test to compare NDWC accuracy before and after mandatory documentation and Cohen’s kappa to determine inter-rater reliability of BONWC and NDWC.

Results:

A total of 386 operations were performed by 25 surgeons, with 125 (32.4%) occurring after BONWC became mandatory. Surgical indication was 52.3% elective, 28.2% emergency, and 19.4% trauma-related. NDWC was more accurate for elective cases than for emergency and trauma cases (69.8%, 49.5%, and 53.5% respectively, p < 0.001). Of 151 cases with misclassified NDWC, 91 cases (60.3%) were under-classified, with actual wound class higher than NDWC, while 60 (39.7%) were over-classified. BONWC accuracy was 72.5%, with 26 cases (78.8%) under-classified and 7 (21.2%) over-classified. Weighted kappa for NDWC and BONWC was 0.67, indicating moderate agreement. NDWC accuracy did not change significantly with mandatory BONWC (65.6% with BONWC vs 58.6% without, p = 0.22).

Conclusions:

NDWC and BONWC agreement was imperfect, and mandatory BONWC was not associated with a statistically significant increase in NDWC accuracy. Both surgeons and nurses tended to under-document wound class, potentially leading to underestimation of expected institutional surgical site infection volumes. Further work is needed to ensure that accurate wound classifications are available for use in hospital quality metrics.