Risk Factors for Retained Surgical Foreign Bodies: A Meta-Analysis

Document Type


Publication Date



Introduction. Retained surgical foreign bodies (RSFB) constitute a theoretically preventable surgical morbidity. Despite numerous case reports and clinical series, very few studies exist that adequately address the issue of RSFB. Moreover, published reports lack the ability to detect clinically important differences among the proposed risk factors for RSFB. This meta-analysis examines existing data for previously unrecognized risk factors for RSFB.

Methods. Eighteen potentially relevant studies were considered for this meta-analysis. Two retrospective, case-control studies of RSFBrelated risk factors contained suitable group comparisons between patients with and without RSFB, thus qualifying for further statistical analysis. Comprehensive Meta Analysis 2.0 (BioStat, Inc., Englewood, NJ) software was used to analyze the following proposed risk factors for RSFB: (a) patient age and gender; (b) body-mass index; (c) surgical count not performed; (d) incorrect surgical count; (e) duration of operation; (f) estimated blood loss >500 mL or transfusion requirement in the operating room; (g) emergent nature of surgical operation; (h) operation after hours (5 pm); (i) unplanned change in surgical procedure; (j) more than one major procedure performed; (k) change in operating room nursing staff; and (l) involvement of multiple surgical teams.

Results. Frequent changes in nursing staff and operation after hours were not significantly associated with RSFB in either study but reached near-significance in our meta-analysis. Of note, six variables found to be independently associated with RSFB in only one of the two studies were determined to be significantly associated with RSFB in this meta-analysis. In addition, duration of operation, while not statistically significant in either of the two studies, was found to be significantly associated with the occurrence of RSFB by meta-analysis. Six remaining variables were confirmed not to be significantly associated with RSFB.

Conclusions. This analysis demonstrated that seven of 13 previously studied variables are significantly associated with the incidence of RSFB. These findings provide a meaningful foundation for future patient safety initiatives and clinical studies of RSFB occurrence and prevention. Further, large prospective studies evaluating effects of specific changes at the institutional level (i.e., universal surgical counts, radiographic verification of the absence of RSFB, radiofrequency labeling of surgical instruments and sponges) on the incidence of RSFB should be undertaken using variables derived from the current analysis.