A proof-of-concept study associating artificial intelligence surveillance of surgical site infections with antibiotic prophylaxis from 765,962 surgeries

A study has demonstrated the potential of artificial intelligence (AI) in detecting surgical site infections (SSSIs) at scale through surveillance of electronic health record (EHR) data. The proof-of-concept study analyzed data from 765,962 surgeries across 18 Danish hospitals between May 2016 and December 2021. The researchers hypothesized that an AI-based Natural Language Processing (NLP) model could identify SSSIs within 30 days postoperatively, associating the findings with breaks in relevant procedural events, such as antibiotic prophylaxis.

The primary exposure of interest was prophylactic antibiotic administration within 120 minutes before incision. The study identified 14,018 SSSIs (1.8%) among the surgeries, with periods exceeding 1 and 2 standard deviations (SD) of the mean infection rate. The exposure occurred in 37.3% of procedures overall, 94.4% of hip and knee arthroplasties, and 54.6% of laparotomies.

The results showed an overall reduction in infection risk associated with AI-based SSSI surveillance (OR 0.90; p < .001). This association was further noted in laparotomies (OR 0.65; p < .001) but not significantly in arthroplasties (OR 0.90; p = .58). The study suggests that AI-based SSSI surveillance can effectively identify infection incidences and potentially break procedural events, offering a valuable tool for infection prevention.

The research was funded by a grant to the corresponding author, and the study is available in Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.