Two women died after caesarean sections at the General Hospital in Bundi, one admitted on 15 September and one on 16 September. The Rajasthan government has given a four-member committee two days to report. The Medical Department has already moved the doctor who performed both operations from the Women’s Hospital in Bundi to Kota, saying this keeps the investigation unaffected.

Both women bled after surgery, and both had balloon tamponade, drugs and blood before being referred to a higher centre. According to the hospital’s PMO, the second was sent out because the bleeding would not come under control and the hospital had no HDU or ICU.

The IMA and the All Rajasthan Serving Medical Association have submitted a memorandum to the District Collector and warned of agitation if the transfer is not withdrawn. Their objection is to the order of events. Nothing has been found yet, three separate inquiries are still running, and Dr Anil Jangid of the IMA points to nearly 15 years of service behind her.

The committee is examining the caesareans, the deterioration, the referrals and the treatment in Kota. The missing HDU is not on that list. If the findings land on one doctor and say nothing about the facility she was working in, the lesson every obstetrician draws tonight is that the safest thing to do with a bleeding patient is to get someone else’s name on the notes.

Read the full report →

🩸 Your systolic cutoff is 70 points too high: A pooled analysis of 15,954 STEMI patients across six tenecteplase trials found brain-bleed risk starts climbing from a systolic of 110, not the 180 every one of those trials used as its limit. Above 110, each additional point of systolic added roughly 2% to whatever risk that patient already carried — 2% of their risk, not two percentage points. Dose is the other lever you still hold: the crude rates run backwards because the lightest, oldest patients automatically get the smallest dose, but hold age, sex, weight and pressure steady and every extra 5 mg carries about 5% higher odds. At 30 mg an extra year of age barely counts. At 50 mg it adds close to 7%. The scoring chart, and the two worked examples.

⏱ 95 minutes, not 76 years: SENIOR-PCNL tracked 716 percutaneous nephrolithotomies in patients aged 70 and over, across 21 high-volume centres in 15 countries. One in six had a complication within 30 days. Cases running 95 minutes or longer carried roughly three times the odds of any complication and close to four times the odds of infection. Age, comorbidity score, hypertension and diabetes did not separate the two groups at all. The trustworthy half of the finding is the reassuring one: finish under 95 minutes and about 97 in 100 had no infection. Their clock starts at renal puncture, so a 95-minute case is a two-hour slot on your board. Why it is a checkpoint and not a stop sign.

💊 One box, three drugs, 14 days: Akums says it has marketing approval for a combikit pairing vonoprazan — a potassium-competitive acid blocker — with clarithromycin and amoxicillin for H. pylori in adults, dosed twice daily for a fortnight. The pitch is adherence: one pack instead of three strips. Five things a prescriber would actually want are missing from the announcement. What is in the pack, and what is not.

This slot goes to one brand per issue, in the middle of a brief that practising doctors actually finish. No display network, no retargeting pixel, no audience of “health enthusiasts”.

Verified Indian physicians. One sponsor a day. Rates and readership on request.

The story from the top of this email — and the number that decides whether you can use any of it.

Across 20 hospitals in five countries, staff photographed the inside of 350 mouths with whatever camera was lying around. No lighting rig, no cropping, no arrow pointing at the lesion. Each photo went to Gemini in a fresh chat with an identical prompt, and the model had to commit to a diagnosis before it knew the patient’s age, sex or smoking history. Then every lesion went for biopsy, and the same photographs went to four humans: a GP, a dentist, an oral pathologist and a head-and-neck surgeon.

On naming the exact lesion it was the same shape: 42 wrong out of 350, against the GP’s 182. Only the surgeon, at 8, beat the machine. And note who is usually first to look inside a mouth in this country — not a head-and-neck surgeon. A GP, a dentist, whoever is running the camp. That is the row it was really competing against.

Two things stop this being a victory lap. Those two cancers were misjudged, not missed — on the simpler question of whether anything was there at all it found all 320 lesions. But of 30 completely normal mouths, it invented a problem in 10.

All of that happened inside cancer hospitals, where 38 of every 100 patients actually had cancer. Your OPD is not like that. A camp is nothing like it. And that one difference changes the answer to the only question you care about.

Nothing about the model changed between those three rows. What changed is how many people in the queue had cancer: 38 in 100, then 5 in 100, then 1 in 100. When almost nobody in the queue has cancer, most of the alarms land on healthy people — simply because there are so many more healthy people available to land on.

The bottom line. Let it help you decide who walks to the specialist. Do not let it decide what they have.

Read the full piece → — the full scorecard, why its self-reported confidence turned out to be honest, where its urgency calls collapsed, the never-smoker trap every reader fell into, and the three uses that survive the fine print.

You can suspend the licence, but the implants stay on the shelf — and a proposed amendment to Rule 30(1) would finally let regulators order the batch itself destroyed. Still only a proposal.

Candidates who reportedly turned up for 10 Junior Resident posts at the Delhi State Cancer Institute, in a clip doing the rounds on medical Instagram this week. the interview was cancelled and the day became document verification.

🚨 Uttar Pradesh: Junior doctors were allegedly assaulted by nursing staff at GMC Urai, with a 27-second video circulating and resident bodies demanding an impartial inquiry. The post.

✊ Rajasthan: Foreign medical graduates have been sitting outside the DME for seven days over internship stipends. The All FMGs Association is now asking the state JDA and RDA for letters of support and a physical visit. The appeal.

🔑 And one habit worth 20 seconds: the AIIMS Patna story above is not a phishing story. No OTP was shared. Unauthorised debits ran for more than two months before anyone looked. Open your statement.

A test’s accuracy belongs to the test. Its usefulness belongs to the waiting room.

Know a colleague who should be reading this? Forward this email to them.

Forwarded this issue? Subscribe free →