Three Things Worth Knowing This Week

Sent to subscribers on 7 September 2026. Details may have changed since.

Starting this issue, a new standing format: training, research, and the specialty itself, in one letter.

Residency and Fellowship · The ERAS date that actually matters
Research · The outcomes that quietly disappear
Cardiothoracic Surgery News · Three trials from ESC Congress 2026

Part One · Residency and Fellowship

The Date That Actually Matters Is September 23

ERAS 2027 opened for submission on September 2. Applications are going in right now. And a large number of applicants are about to make the same avoidable mistake, because they are optimizing for the wrong date.

Here is the mechanic almost nobody explains clearly.

Programs cannot review anything until September 23 at 9 a.m. ET. Everything submitted between September 2 and September 23 lands in front of programs at the same moment. An application submitted on day one and one submitted on September 22 arrive together, indistinguishable. There is no first mover advantage inside that window.

What does exist is a real penalty on the other side of it. Submit after September 23 and many programs have already downloaded, sorted, and begun extending interview invitations on a rolling basis. You are not evaluated late. You are evaluated after decisions have already started being made.

Submitting on September 2 buys you nothing. Submitting on September 24 costs you something real. The entire game is being complete before the gate opens, not first through it.

Three structural changes this cycle are worth knowing if you have not looked recently. The Publications section has been redesigned and renamed Scholarly Work, and you can now flag up to three entries as Most Meaningful, which is a genuine opportunity to direct attention rather than let a reviewer skim a list. All letters of recommendation now route through the centralized AAMC Letter Writer Portal, so your writers need lead time with a new system, not just a deadline. And Urology, Dermatology, and Integrated Plastic Surgery have moved to specialty specific standardized letters.

Program signaling is now close to universal: in the 2026 season, over 90 percent of ERAS participating specialties opted in. Signals to programs you have applied to lock when programs open applications on September 23 at 9 a.m. ET, so treat them as final once you submit.

What to do with this

Work backward from September 23, not forward from September 2. Your MSPE, transcript, and letters need to be uploaded and verified before programs open applications, and for IMGs, ECFMG verification adds its own timeline on top of that. Aim to be fully complete several days early, then spend the remaining time on signal allocation rather than on the submit button.

Part Two · Research

The Outcomes That Quietly Disappear

A foundational audit by the COMPare Project at the University of Oxford, examining all 67 trials published across NEJM, JAMA, Lancet, BMJ, and Annals of Internal Medicine between October 19 and November 30, 2015, found that 58 of them, 87 percent, misreported their pre-specified outcomes, and that 365 new outcomes were added without declaration. Only 9 of the 67 trials reported their outcomes with no discrepancy at all. This is the actual mechanism of outcome switching, documented directly in some of the most visible journals in medicine.

This is not only a historical concern. In March 2026, the FDA sent reminder messages about results reporting requirements to more than 2,200 sponsors and researchers tied to more than 3,000 registered trials, and the agency's own analysis found that 29.6 percent of studies highly likely to fall under mandatory reporting had no results posted.

A pre-registered outcome that quietly disappears, and an unregistered outcome that quietly appears, are the same problem viewed from two directions.

For anyone designing or currently running a trial, this is directly actionable, not just a cautionary statistic about other people's work.

What to do with this

Before you analyze your data, pull up your own original registration and check it against your current analysis plan. If a primary outcome has shifted, document why, explicitly, in the methods section, rather than letting the discrepancy go unexplained. Reviewers and readers increasingly check this directly.

Part Three · Cardiothoracic Surgery News

Three Trials From ESC Congress 2026

ESC Congress 2026, Munich · August 31

LAACS-2: routine left atrial appendage closure during cardiac surgery is not supported

This multicenter randomized trial across Denmark, Spain, and Sweden enrolled patients undergoing first time planned open heart surgery, with and without prior atrial fibrillation, regardless of CHA2DS2-VASc score. The conclusion: results do not support routine concomitant LAA closure in all patients, though those at high stroke risk may still benefit. For a procedure many surgeons have been adding reflexively, this reframes the decision as an individualized one based on stroke risk rather than a default add on.

ESC Congress 2026 · Simultaneously published in NEJM

TAVI-PCI: valve first is noninferior to coronary first, and the order changes the decision

986 patients with severe aortic stenosis and concomitant coronary disease across 48 centers in six European countries, mean age 82, randomized to TAVI followed by PCI or the reverse. TAVI first was noninferior at one year. The genuinely interesting finding, per presenter Barbara Stähli of University Hospital Zurich: fewer patients in the TAVI first group ultimately underwent PCI at all, not because coronary access was impaired, but because once the valve was treated the clinical judgment about whether PCI was still needed sometimes changed. Life threatening, disabling, or major bleeding was numerically lower with TAVI first, 6.6 versus 9.7 percent, a secondary finding Stähli said should be interpreted cautiously.

ESC Congress 2026 · Simultaneously published in JAMA

ACASA-TAVI: NOAC monotherapy reduces leaflet thickening after TAVI

360 patients undergoing TAVI for severe aortic stenosis were randomized to 12 months of non vitamin K antagonist oral anticoagulant monotherapy or aspirin monotherapy. NOAC monotherapy reduced the incidence of hypoattenuated leaflet thickening and was noninferior for bleeding, thromboembolic events, or death.

Whether you are navigating a training requirement in flux, tightening a trial's outcome reporting, or simply keeping current with the field, this is exactly the kind of thinking AskDimitri is built to work through with you.

Work through it with AskDimitri →

Editor's note: in October 2026 the figures in this letter were checked against the sources listed below, and the text was corrected where the letter as sent differed from them.

Until next week,

Sources

A new letter goes out every week, to subscribers first. Join to get the next one.