Background

The MBSAQIP maintains patient-level data for all bariatric procedures done in accredited centers, the Participant Use Data File (PUF). Conflicting results reported in recent publications using common PUF years have led to concerns about methodologies and the impact on clinical practice.Objective: To evaluate agreement in reported outcomes for studies using the PUF years 2015-2018 to investigate LSG and LRYGB.

Methods

An OVID search identified all studies using PUF data to investigate LSG and LRYGB outcomes. Studies evaluating only pediatric or geriatric patients were excluded. Outcomes assessed included cohort size and 30-day rates for readmission, reoperation, mortality, leak, and bleed. Individual outcome rates were compared to the corresponding median rate to determine the percent difference in values. Agreement categories were strong agreement (SA <=25% difference), moderate agreement (MA 25-50% difference), and disagreement (DA >50% difference).

Results

Seventeen of twenty-one studies met criteria for analysis. There was 82% SA for cohort size. Reoperation and transfusion had 88% and 89% SA, respectively. Over 1/3 of the studies could not demonstrate SA for mortality and leaks. Leak exhibited the greatest DA at 17%.

Conclusions

Overall, more than 80% of studies had SA for both cohort size and outcomes. However, this means that nearly 1/5th of studies demonstrated either MA or DA, and mortality reporting had only MA in 1/3rd of studies. Standards need to be developed for use of the MBSAQIP PUF to protect against either intentional or unintentional misuse.