Background

Staple-line leak is a well-known complication of sleeve gastrectomy. MBSAQIP has published contradictory guidance on leak counting strategies. Unfortunately, this has led to variability in reporting leaks and the use of organ-space surgical site infection (OSI) as a surrogate. This study evaluates the effectiveness of OSI as a surrogate for leak and the completeness of the leak variable introduced in 2020.

Methods

The MBSAQIP PUF was queried from 2015-2022. Leak was counted by methods found in the semiannual reports (SAR) for two distinct time periods. Before 2020, 'SAR def 15-19' was defined as readmission, reoperation, or reintervention with leak given as most likely reason. Since 2020, 'SAR def 20-22' is defined as 'SAR def 15-19' OR leak recorded under the new MBSAQIP variable: POSTOPANASTSLLEAK. These leak rates were compared to rates of OSI, and kappa statistics were calculated.

Results

Of 559,324 sleeves performed from 2015-2019, 0.21% (n=1,203) were positive for 'SAR def 15-19', 0.15% (n=854) were positive for OSI, and 0.09% (n=479) were positive for both (k= -0.46, p<0.01) (Figure). Of 373,671 sleeves performed from 2020-2022, 0.19% (n=698) were positive for 'SAR def 20-22', 0.21% (n=769) were positive for OSI, and 0.14% (n=509) were positive for both (k= -0.30, p<0.01) (Figure). When comparing 'SAR def 15-19' vs POSTOPANASTSLLEAK, there was no agreement between these variables (k= -0.28, p<0.01) (Figure).

Conclusions

Organ space SSI and leak do not have agreement in MBSAQIP and should be reported separately. We suggest counting leaks according to the SAR method during any given year.