Breast cancer staging turns on small operative details. The sentinel node episode clarifies which hot, coloured, clipped or suspicious axillary nodes count, and why documenting counts, channels and node yield matters for multidisciplinary decisions.

Haematemesis after open abdominal aortic aneurysm repair is not something a negative endoscopy should settle. The surgery scenarios are niche, but the bedside lesson is memorable: ask about prior graft surgery, recognise acute mesenteric ischaemia early, and do not let aortoenteric fistula bleeding wait.

Call burden is not just how often it happens but how hard the night was and what elective work follows. This is more workforce than bedside, but it gives surgery leaders a practical language for overnight intensity, OR access, administrative load and sustainable job design.

Training reform rather than bedside care is the focus here. It separates the Match algorithm from pay, mistreatment, mobility, and workforce problems, without treating abolishing the Match as an automatic fix.

Internalised shame matters even when nobody has openly humiliated the trainee. This follow-on episode is best opened after the earlier shame discussion, because it shows how self-blame after criticism or complications links with burnout and why mentorship and debriefing matter.

Humiliating feedback is separated clearly from firm correction of unsafe care. The useful point is not wellness language but teaching behaviour: name the safety issue, keep the trainee’s character out of it, and notice how shame, burnout, isolation, and suicidal thoughts can follow.

Shame is approached as a clinical and training problem, not just a personal feeling. The episode helps explain why silence, defensiveness, perfectionism, or overwork may follow errors or feedback, and why private behaviour-focused conversations protect learning better than public criticism.

Breast screening AI only matters clinically if it reduces interval cancers or catches aggressive disease, not just if it increases detections. The episode keeps attention on radiologist workload and automation bias, which makes it useful when someone claims software can safely replace readers.

The reason not to reassure is that VV ECMO only helps if it is still a bridge to transplant. This episode is most useful when a listed patient worsens, because it focuses on awake ECMO, ambulation, candidacy review and knowing when support has lost its destination.

The pitfall is turning an M&M case into a discharge summary and burying the complication. Worth opening for its practical rules on one-sentence headlines, clean timelines, decision points, and no-blame analysis of post-operative haemorrhage or anastomotic leak.