NephSpace

For years, most of us thought it didn't matter whether critically ill patients started on intermittent hemodialysis or continuous therapy. In the ICRAKI trial, starting on intermittent hemodialysis was linked to more deaths, and the trial stopped early. We explain why the size of that gap is uncertain, but its direction is harm, and why the trial couldn't show that intermittent was no worse. A JAMA study follows last week's fluids-or-pressors thread: resuscitation was only about one liter in six of ICU fluid, with maintenance and the fluid that comes in with drugs making up most of the rest...

Three reflexes worth rethinking. In septic shock in the emergency department, stopping the fluid and starting the vasopressor early showed no difference in days alive and out of the hospital compared with the classic fluids-first approach, with about half the extra fluid. In a blinded ICU trial, sodium bicarbonate for metabolic acidosis in shock fixed the blood gas but did not significantly lower death, dialysis or lasting kidney damage. And a review of 13 interview studies asks why so few patients start peritoneal dialysis. Each with what's new, the caveats and what it means on Monday.

A 60-page referral, nine years of creatinine, and an AI summary waiting on two desks: the attending's and the first-year fellow's. A Kidney360 perspective says to hand drafting, summarizing and paperwork to AI first, because a clinician checks the draft. A CJASN perspective says that same drafting is how fellows learn judgment, and warns of "never-skilling". We look at what the evidence actually shows: an ambient-scribe trial in which the better of two tools saved about 20 seconds per note, trainees who accepted roughly three of every four wrong AI answers, and endoscopists who detected fewer adenomas on their own...

A new lipid guideline makes reduced kidney function a reason for a statin; in 2024, before it arrived, most patients for whom kidney function is now the reason had no statin prescription. An editorial asks whether GLP-1 drugs belong in dialysis units before the trials are done. A review says dialysis adequacy should be measured by what it does for the heart, troponin turns out to be a moving target during a session, and an AI reads photos of fistulas. Five picks from the week of September 25, 2026, each with the background, what's new, the caveats and what it means on...

I want to keep up with nephrology. But the new papers sit behind logins and scattered across sites, so I end up doomscrolling instead. NephSpace brings the week's kidney research and news together, out loud: one weekly episode on what's new, plus deep dives in between, made for the couch, the run and the drive to work.

Let's transform scroll time into learning time. Knowledge, transformed: easy to reach, hard to stop.

Follow on Apple Podcasts: https://podcasts.apple.com/us/podcast/nephspace/id6818364609
Spotify: https://open.spotify...

Two dialysis patients, same city, same week. One gets a new machine that filters and replaces more than 20 liters every session. The other gets the same old machine with a new filter. Both are told it's the future. Hemodiafiltration has a survival trial but needs new machines. Medium cutoff dialyzers drop into today's machines but have no survival evidence against standard dialysis. A detective story with a Fight Night verdict, including who makes, sells and funds each road.

The takeaway: Both roads clear more middle molecules, but only hemodiafiltration has shown that people live...

Two ways to clear more middle molecules arrive in US dialysis units at once. A urine dipstick can show whether patients are taking their SGLT2 inhibitor. Two papers disagree about who should let AI write the notes. Nine picks from the week of September 21, 2026, each with the story, the numbers and a take.

Disclosure: Dr. Hussein serves on the Fresenius Medical Care US Home Medical Advisory Board. Fresenius sells hemodiafiltration machines, which are discussed in this episode.

Chapters