The Feeding Tube Is Usually a Detour, Not a Destination

Every consent conversation about metabolic and bariatric surgery mentions nutritional deficiency. Very few mention the patients who stop being able to nourish themselves by mouth at all — the ones who leave hospital with a feeding tube or a central line, and whose recovery is measured in months.

A retrospective cohort published on 13 October in the Journal of the American College of Surgeons puts a number on that group. The important figure is not how many needed the support. It is how many came off it.

168 patients, and a strict definition of recovery

The authors, from the Cleveland Clinic’s digestive disease institute, searched records from 2004 to 2023 for adults who required enteral or parenteral nutrition after a bariatric operation. They found 168. Median age 41.2 years, median body mass index 47.3 kg/m², 85.1% women.

Independence was defined tightly: off all supplemental nutrition for at least three months. Over a median 11.6 years of follow-up, 148 of the 168 — 88.1% — met it. Median time from starting nutritional support to independence was 12.1 months.

This is a single-centre retrospective review, and 168 people is a small number for an outcome this uncommon. But the direction of the finding is hard to argue with, and it runs against the way these patients are often pictured: as permanent casualties of an operation that took away too much absorptive surface.

The repair was mostly plumbing

The detail that carries the paper is what the 148 went through. Of those who became independent, 126 — 85.1% — required an endoscopic or surgical intervention, most commonly to address a complication at the gastrojejunal anastomosis, the join between the stomach pouch and the small intestine in a bypass.

Set that beside the first number. Nutritional support kept these patients alive and out of trouble while the problem was identified; it was not what solved it. The solution was mechanical — dilating a narrowing, revising a connection, correcting anatomy that had kinked or scarred.

The timing carries its own message. Among 93 patients still dependent at twelve months, 74 — 79.6% — later became independent. On this cohort’s evidence, a year of tube feeding is not a verdict.

A mortality signal that stayed short of significance

The survival analysis is where the paper is most careful and least conclusive. Among 165 patients alive and in follow-up at twelve months, mortality was 20.4% in the 93 who were not yet independent, against 6.9% in the 72 who were. The hazard ratio was 2.56, with a confidence interval running from 0.95 to 6.90 and a p value of 0.064.

The authors used a twelve-month landmark analysis to keep patients who died early from being counted as failures of recovery — the standard defence against immortal time bias, and the reason the comparison is worth reporting at all. It still does not reach significance, and with 165 people and 19 deaths it could hardly be expected to. What it does is mark a group — patients whose nutritional autonomy is delayed past a year — that deserves closer follow-up, without claiming that dependence causes death rather than accompanying it.

Anatomy you can build two ways

If the gastrojejunostomy is where these problems concentrate, the question of how the join is constructed is not academic. A nationwide Swiss administrative study, published in Obesity Surgery on 16 October, took it on with 21,320 minimally invasive RYGB operations recorded between 2015 and 2022. Linear stapling accounted for 18,161, circular for 3,159, and circular use declined across the period.

The results are a lesson in method. In the multivariable model, circular stapling carried an odds ratio of 1.221 for a major in-hospital complication, with a confidence interval from 0.824 to 1.809 — no association. After propensity-score matching, the same comparison produced 2.464, interval 1.225 to 4.956. Same file, same exposure, opposite verdicts, and the authors say so plainly rather than picking the more flattering number.

What held in both analyses was narrower: surgical site infection (odds ratios 2.251 and 7.647) and a length of stay above the median (relative risks 1.136 and 1.318). Even the distribution’s tail behaved inconsistently — length of stay above the 90th percentile was associated in the multivariable model and not after matching. The in-hospital window also closes before the strictures and narrowing that fill the long-term picture, which makes the study better evidence that circular stapling costs infections and bed-days than that it explains nutritional failure years later.

What duodenal continuity does not buy

The counter-argument in this field has always been that malabsorption is a price paid for a particular anatomy, and that operations preserving duodenal continuity should avoid it. A prospective single-centre series from Turkey, published in Obesity Surgery on 12 October, tests that intuition at five years.

Thirty-five women with severe obesity and insulin-treated type 2 diabetes underwent sleeve gastrectomy with transit bipartition in 2016 and 2017. Thirty retained the anatomy with complete efficacy data at five years. HbA1c fell from 9.2% to 5.5%, weight from 118.0 to 75.6 kg, total weight loss averaged 35.7%, and medication-free diabetes remission was 23 of 30 — 76.7%, or 65.7% counting all 35 enrolled.

Then the nutritional ledger. At year five, hypoalbuminaemia affected 20%, low serum iron 30%, vitamin B12 deficiency 13.3% and vitamin D deficiency 73.3%. Twenty-seven of the 30 patients — 90% — had at least one predefined biochemical abnormality, and two needed an invasive intervention related to the operation, including one revision.

One surgeon, one centre, thirty women, no control arm: this is a small study and it should be read as one. The point it makes is narrow and it lands. Preserving the duodenum changes the shape of the nutritional risk; it does not remove the need to look for it.

The arithmetic the field already publishes

None of this is news to the agencies. NIDDK’s patient information on side effects states that follow-up interventions, surgery and hospitalisations are “relatively common within 5 years of weight-loss surgery, affecting about one-third of patients,” and that follow-up procedures are needed more often after gastric bypass than after sleeve. It also notes that inadequate nutrient absorption can produce anaemia and osteoporosis — the slower, quieter version of the same problem, documented in this archive’s earlier reading of bone density, trace minerals and anaemia after metabolic surgery.

What the new cohort adds is a denominator and an outcome for the severe end of that distribution. About one patient in three needs another procedure within five years; a small fraction of those end up dependent on tube or intravenous feeding; and of those, nearly nine in ten eventually eat again, usually because someone fixed the anatomy.

The LABS study, funded by NIDDK, remains the design that would answer the parts a single centre cannot — a prospective multi-centre cohort with structured nutritional follow-up, described in this site’s LABS archive. This archive is independent of NIH and NIDDK, and reports what the published record shows rather than what any institution would prefer it to show. On that record, the nutritional tail of bariatric surgery looks less like an irreversible consequence of malabsorption and more like a plumbing problem with a clock on it: the sooner it is diagnosed, the less of a patient’s life it takes.