For referring physicians
Reconstruct, divert or stop: how I decide, and the diversions I do
The decision every other page hands off to. The radiated bladder, the devastated outlet, the reconstruction that has failed more than once, and the diversion that itself needs revising. The questions I ask, in order, with a video for each operation.
Does the reconstruction door close?
One finding closes it: refractory hematuria in a radiated bladder. That bladder is not salvageable and it comes out. Nothing else on the chart does the same work. A small bladder is an augmentation patient, not a diversion patient, including in a radiated pelvis, where the augment is close to building a neobladder. The count of failed repairs adds weight and never closes the door on its own.
Refractory means what has already failed elsewhere: a lot of interventions that do not seem to work. The man labeled refractory who has had none of the ladder and the man who has been fulgurated, put in a hyperbaric chamber and given formalin are not the same patient, and the label hides the difference. Ask what it means for this patient before it decides anything.
Which diversion: the sphincter, not the radiation
Continent diversion is not one thing. A neobladder depends on a sphincter that closes, so a neobladder through a radiated sphincter that does not close is contraindicated, and pelvic radiation that spared the sphincter leaves a neobladder reasonable. The question is never whether the pelvis was radiated; it is whether this sphincter works. The same variable decides augment versus neobladder for the bladder that stays.
An Indiana pouch is catheterizable and does not depend on the sphincter, so radiation does not exclude it. What excludes a pouch is the patient: one who cannot catheterize, which also excludes a neobladder; a creatinine above 2, as a default the goals conversation can move rather than a gate; and a short intestinal tract that leaves nothing to build from. Bowel is a limiting resource.
The bladder left behind
Three states. With refractory hematuria, the bladder comes out at the same sitting as the diversion. Otherwise it stays in with the mucosa stripped, which is the usual case and the move that prevents pyocystis. And when the bladder has a way to drain, or taking it out would risk the rectum and its neighbors, as it does when the bowel is stuck to the point that dissection means enterotomies, I leave the bladder alone and complete the diversion anyway. A retained bladder that keeps bleeding gets a cystoscopy and fulguration first, not a return to the abdomen.
What I ask before I operate
Of the frail patient on a chronic suprapubic tube: how is he tolerating the tube? What do the recurrent infections mean, is he being hospitalized? Will he accept a urinary diversion as the eventual outcome if the reconstruction does not work? How tolerant is he of incontinence, given that a reconstruction may leave him incontinent? The diversion question is asked before the reconstruction, which makes it a bailout agreed at consent rather than a salvage conversation after a failure.
Frailty changes the operation, not whether to operate. Age, the number of prior operations and the hostility of the field change the operation too. The ladder of interventions descends to a chronic tube, which is still something done, and never reaches zero. The only thing that stops me is a complete mismatch between what the patient expects and what the operation can deliver: no risk of incontinence, no risk of osteomyelitis, no catheter after a pouch. Every operation here hands the patient a job afterward, and a patient whose expectations exclude the job cannot do it.
Bailouts declared before the incision
There is no early tell. A case that is going wrong does not announce itself in the first ten minutes; it runs until it reaches something declared in advance. Two triggers are declared: too stuck to enter the pelvis means no cystectomy, and the diversion still happens; fecal incontinence the patient has not accepted stops the fistula takedown. I call it alone, because nobody else in the room can, which is why the trigger has to be said before the case rather than discovered during it.
A sphincter never goes in at the same sitting as a reconstruction. The repair proves itself on a cystoscopy first; the interval and the gate are on the incontinence page.
Revising a diversion
A neobladder that has failed can be converted to a conduit without resecting and reanastomosing bowel; a leak that will not heal, a fistula into bowel, and a stricture where the neobladder meets the urethra are each repaired robotically (videos below). A parastomal hernia after a conduit is repaired robotically with a dual-surface mesh through a 3 cm contralateral incision, most often with the SP robot, and often with a ureteroenteric stricture or an abdominal wall reconstruction in the same sitting. The ureteroenteric stricture itself has its own page.
Videos by operation
Most operative videos are age-restricted on YouTube, so you need to be signed in to watch them.
Cystectomy and ileal conduit
Continent diversion
Augmentation and the bladder kept
Revision of a diversion
For patients
Papers
- Robotic assisted repair of post-ileal conduit parastomal hernia: technique and outcomes. Urology. 2021. PubMed 34481825
- Robotic repair of ureteroenteric stricture following radical cystectomy: a multi-institutional experience. Urology. 2022. PubMed 35007620
- Completely intracorporeal robotic-assisted laparoscopic augmentation enterocystoplasty with continent catheterizable channel. Urology. 2014. PubMed 25432822
- Reconstruction of a devastated ureter: multi-institutional experience with robotic intracorporeal ileal ureter replacement. Eur Urol. 2026. PubMed 41577584
- Single-port robotic ileal ureter reconstruction: feasibility, technique, and early outcomes. World J Urol. 2026. PubMed 42068369
- Editorial comment on "Definition of benign ureteroenteric anastomotic strictures in ileal conduits after radical cystectomy". Urology. 2024. PubMed 38452942
- More on robot-assisted laparoscopic radical cystectomy. N Engl J Med. 2014. PubMed 25337765