Robotic duodenal switch had a higher 30-day readmission rate than laparoscopic in a 17,615-case registry

An analysis of ten years of United States registry data has found that duodenal switch operations done with a surgical robot were followed by more 30-day hospital readmissions than the same operation done laparoscopically. The paper was published online in Obesity Surgery on September 1, 2026, by a team at the University of Minnesota with a co-author at Dartmouth College.

The duodenal switch is one of the less common bariatric operations. It combines a sleeve gastrectomy with a long rerouting of the small intestine, and it typically produces the largest weight loss of any standard procedure. It also carries the highest ongoing nutritional demands, because a long stretch of intestine is bypassed and fewer nutrients get absorbed.

The data came from MBSAQIP, the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program, a national registry run jointly by the American College of Surgeons and the American Society for Metabolic and Bariatric Surgery. Accredited centers submit standardized 30-day outcome data on their cases.

Readmission was 6.2% after robotic surgery and 5.0% after laparoscopic

The study included 17,615 minimally invasive duodenal switch cases performed between 2015 and 2024. Of those, 5,496, or 31.2%, were robotic-assisted. Open, endoscopic, hand-assisted and single-incision cases were excluded, as were revisions and conversions from an earlier operation.

Within 30 days of surgery, 6.2% of robotic patients were readmitted to hospital, against 5.0% of laparoscopic patients. That difference was significant at p below 0.01.

The absolute gap is 1.2 percentage points. Put another way, about 1 extra readmission for every 83 robotic operations.

After adjusting for other factors, robotic assistance was associated with an odds ratio of 1.78 for readmission, with a 95% confidence interval of 1.25 to 2.55. An odds ratio is not the same as a risk. It compares the odds of an event in two groups, and when the event is uncommon it tends to look more dramatic than the underlying difference in plain percentages. Both figures are describing the same 6.2% against 5.0%.

Longer operations were also independently associated with readmission, at an odds ratio of 1.27 for each standard deviation of extra operating time, with a confidence interval of 1.18 to 1.38.

Operating faster did not close the gap

The researchers built a pre-specified interaction term to test whether operating time explained the robotic disadvantage. It did not, and the direction was the opposite of what a simple time explanation would predict.

Adding the interaction significantly improved the statistical model, with a likelihood ratio p of 0.007. The interaction odds ratio was 0.79, with a confidence interval of 0.67 to 0.94. In practical terms, each standard deviation reduction in operating time made the robotic disadvantage relatively larger, not smaller.

The authors draw the obvious conclusion: improved operative efficiency will not resolve the robotic readmission disadvantage. Whatever is driving the extra readmissions, it is not simply that robotic cases take longer.

The disadvantage sat in the cases from 2015 through 2019

A subgroup analysis by time period found the effect concentrated in the earlier years of the decade, 2015 through 2019. That pattern is what a learning curve looks like. Robotic bariatric surgery was far less established in 2015 than it is now, and centers adopting it were early in their experience with it.

The registry does not record how many robotic cases an individual surgeon had done, which the authors identify as a critical missing confounder. A surgeon on case number five and a surgeon on case number five hundred appear identically in this data.

What a registry of this kind can and cannot settle

This is an observational analysis, and the authors say plainly that it does not establish cause. Patients were not randomly assigned to robotic or laparoscopic surgery. Surgeons choose the approach, and that choice can be influenced by patient factors, including ones the registry does not capture well.

MBSAQIP only covers accredited centers, so results from non-accredited centers are absent. The registry window closes at 30 days, which means nothing here speaks to weight loss, nutritional problems, revision rates or anything else that unfolds over months and years. Cost and resource use were not available.

Readmission is also a blunt outcome. It counts a return to hospital without distinguishing a short stay for dehydration from something serious.

For a patient weighing a duodenal switch, the practical takeaway is narrow. This is a reason to ask a surgeon how many robotic duodenal switches they personally have done and what their own readmission rate is. It is not a reason to rule out a robotic operation at an experienced center.