This Week in General Medicine — Aug 3, 2026

Welcome to This Week in General Medicine. This week we're covering nine notable papers spanning perioperative and cardiac surgical care, randomised trials that push drugs into unfamiliar territory, and a cluster of papers on chronic disease management and prevention that argue as much for stopping treatments as starting them. Let's dive in.

We start in the operating theatre, where two trials ask how we can make cardiac surgery less punishing. In The Lancet, the MIST trial randomised 170 patients with multivessel coronary disease across seven centres in Canada, India, China, Germany, the United States and Japan to coronary artery bypass grafting either through a small thoracotomy or through conventional median sternotomy [1]. The primary endpoint was patient-reported physical recovery at one month on the Short Form 36 physical component summary, and the minimally invasive group did better, with a mean difference of just under three points, which was statistically significant but modest in magnitude. What is arguably more reassuring is the safety picture: through twelve months there were no deaths and no strokes in either arm, and only a single major adverse cardiac or cerebrovascular event overall. The caveats matter, though. This was a small, open-label trial in selected patients operated on by experienced minimally invasive teams, ninety-one percent of participants were male, and recruitment stretched over six years. So this supports offering the thoracotomy approach to appropriately selected patients in high-volume hands, not a wholesale change in surgical practice. The second perioperative paper, in JAMA, is likely to generate more debate. Dutch investigators randomised 784 adults undergoing elective cardiac surgery at seven hospitals to four doses of dapagliflozin ten milligrams, starting the day before surgery and continuing through the second postoperative day, or to placebo [3]. Acute kidney injury within seven days occurred in twenty-eight percent of the dapagliflozin group versus fifty-two percent with placebo, a relative risk reduction of roughly forty-five percent. Atrial fibrillation and reoperation rates were essentially identical between groups, at forty-five percent and around ten percent respectively. That is a very large effect from a very short course of a cheap oral drug, and it sits awkwardly beside current perioperative advice, which generally tells us to hold SGLT2 inhibitors for several days before surgery because of euglycaemic ketoacidosis concerns. The trial population was predominantly male, ninety-seven percent White, and had largely preserved renal function, and the baseline acute kidney injury rate of fifty-two percent is higher than many centres would report, which will fuel discussion about the definition used. Still, this is the first drug shown in a properly powered randomised trial to prevent acute kidney injury in this setting.

Our second theme is trials taking established drugs, or entirely new ones, into unfamiliar disease territory. The BMJ published a Chinese trial of a thousand adults aged eighteen to sixty-four with migraine and an echocardiographically confirmed patent foramen ovale, randomised across thirty-nine centres to aspirin, clopidogrel, rivaroxaban, or metoprolol as an active control, each for twelve weeks after a twelve-week baseline screening period [4]. All three antithrombotic agents were non-inferior to metoprolol for the primary endpoint of a fifty percent or greater reduction in monthly migraine days or attacks. Rivaroxaban went further, with a responder rate of about seventy-eight percent versus roughly sixty-two percent for metoprolol, an absolute difference of sixteen percentage points, along with better secondary outcomes for migraine days, complete cessation, and migraine-specific quality of life. No major bleeding events occurred. Before anyone starts anticoagulating migraineurs, note that this was open-label with blinded outcome assessment, ran only twelve weeks, and enrolled a relatively young population with no history of stroke or transient ischaemic attack, so the long-term bleeding risk of indefinite rivaroxaban in this group is entirely unaddressed. But it does strengthen the thrombotic, rather than purely shunt-mechanical, hypothesis for patent foramen ovale associated migraine, and it is a striking finding in a field where device closure trials have been largely disappointing. Also in The Lancet, the HOPE-3 trial tested deramiocel, an allogeneic heart-derived cardiosphere cell therapy, in 106 people aged ten and older with advanced Duchenne muscular dystrophy, given as a quarterly outpatient intravenous infusion for twelve months [2]. The primary endpoint, percentage change in the Performance of the Upper Limb version 2.0 score, favoured deramiocel by four and a half percent, a statistically significant difference with a confidence interval that just cleared zero, and safety was comparable to placebo. For a non-ambulatory population where upper limb function determines independence, slowing decline matters, and the therapy is genotype-agnostic, which distinguishes it from exon-skipping and gene therapy approaches. Whether a four and a half percent difference on a functional scale over one year translates into meaningful preserved independence over years is the open question, and general physicians looking after adults with Duchenne should expect this to reach specialist multidisciplinary discussions soon.

The third theme is one of the most useful for everyday practice: evidence that tells us to do less, or to choose differently for long-term risk. The BMJ published a systematic review and network meta-analysis of forty-seven randomised trials and 6,230 children and adults examining add-on oral antihistamines for atopic dermatitis [7]. With moderate certainty, H1 antihistamines did produce statistically detectable reductions in eczema severity and itch, but both were far below the minimal important difference, meaning patients would not notice them. With low certainty, none of the interventions improved sleep disturbance or reduced flares. Meanwhile first generation agents probably increase cognitive impairment and may increase discontinuation from adverse events, with an estimated sixty-six more discontinuations per thousand patients treated, and even among second generation agents cognitive effects differ, with loratadine appearing neutral and cetirizine and levocetirizine showing small increases. The authors state plainly that this is evidence against routine antihistamine use in eczema. That is a deprescribing message worth acting on this week, particularly for the sedating antihistamines still handed out for night-time itch. Pair that with The Lancet's seminar on psoriasis, which reframes the disease as one point of a psoriatic spectrum affecting more than forty million people, driven by the interleukin-23 and interleukin-17 axis and carrying a substantial burden of mental health and cardiometabolic comorbidity [6]. The practical message for general physicians is that the skin is the visible part of a systemic inflammatory disease, and that screening for psoriatic arthritis, depression, obesity and cardiometabolic risk belongs in routine review, not only in the dermatology clinic.

Staying with long-term risk, JAMA published the Opti-DOR trial, which randomised six hundred antiretroviral-naive adults at two South African sites, ninety-nine percent Black African and sixty-nine percent assigned female at birth, to doravirine with lamivudine and tenofovir disoproxil fumarate, or to dolutegravir with emtricitabine and tenofovir alafenamide [5]. At forty-eight weeks, viral suppression was eighty-nine percent versus about ninety-one percent, meeting the non-inferiority margin, and median weight gain was three kilograms versus five kilograms, a two kilogram difference favouring the doravirine regimen. But there is a trade-off: hip and spine bone mineral density fell significantly more with tenofovir disoproxil fumarate, and seven of the nine virologic failures on doravirine developed doravirine resistance, though most resuppressed after switching to dolutegravir. So this is a genuine option for patients in whom weight gain is a dominant concern, with bone density and resistance as the price. On the prevention side, JAMA also published a position statement from a working group of the Society for Research on Nicotine and Tobacco making a single overarching recommendation to clinicians in the United States: integrate nicotine e-cigarettes into shared decision-making conversations about all evidence-based pharmacological smoking cessation treatments [8]. Their argument is that e-cigarettes are less harmful than combustible cigarettes and have outperformed approved nicotine replacement therapy for cessation, while misperceptions about their harms remain common among clinicians themselves. This is a position paper, not new trial data, and it will remain contested, but it does mean being ready to have that conversation with an evidence base rather than a reflex. Finally, a PLOS Medicine commentary argues that the Developmental Origins of Health and Disease framework has leaned too heavily on maternal pregnancy behaviours, and that the largest gains in child health may come from tackling socioeconomic disadvantage rather than from further exhortation of individual mothers [9] — a useful corrective to how we frame risk in antenatal counselling.

If you only have time for one paper this week, make it the dapagliflozin cardiac surgery trial in JAMA [3]. A four-dose perioperative course of a widely available, inexpensive drug nearly halving postoperative acute kidney injury is the kind of result that will force a rethink of the current advice to withhold SGLT2 inhibitors before surgery.

Here are the key takeaways from this week in General Medicine. First, dapagliflozin started the day before elective cardiac surgery cut acute kidney injury from fifty-two percent to twenty-eight percent, with no excess atrial fibrillation or reoperation, and this directly challenges routine preoperative withholding of SGLT2 inhibitors. Second, minimally invasive multivessel bypass through a small thoracotomy gave modestly better patient-reported physical recovery at one month with no safety penalty to a year, but only in selected patients and experienced hands. Third, in patients with migraine and a patent foramen ovale, aspirin, clopidogrel and rivaroxaban were all non-inferior to metoprolol over twelve weeks, and rivaroxaban was superior, without major bleeding, though long-term safety is unknown. Fourth, oral antihistamines for eczema produce reductions in itch and severity too small for patients to notice, do not improve sleep or flares, and first generation agents impair cognition, so routine use should stop. And fifth, a doravirine-based regimen matched dolutegravir for viral suppression with two kilograms less weight gain, at the cost of greater bone density loss and emergent resistance in most of those who failed.

That's your roundup for This Week in General Medicine. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And if you have a paper you have been meaning to read, upload the PDF, or paste any link, at audioscholar dot C C. We will turn it into audio like this one, in any of thirty-one languages.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.