Ann Robinson’s research reviews—1 October 2021

Ethics of paying people to enrol in trials

Is it ethical to incentivise people to take part in research studies? Do inducements blunt their assessment of the risks involved and mean that poorer people are more likely to sign up, or do they widen and improve participation? Two randomised clinical trials were embedded in a smoking cessation trial and an ambulation trial in which potential participants were offered between $0 and $500 if they consented to take part. Incentives substantially increased consent rates in the smoking trial (21.8%, 35.9%, and 47.1% in the $0, $200, and $500 groups respectively), but not in the ambulation trial (around 45% in groups offered $0, $100, or $300). Consent rates were probably relatively high in the ambulation trial, regardless of inducements, because people thought it was likely to be low risk and potentially beneficial. Non-inferiority tests in both trials suggested that the payments didn’t act as “undue or unjust inducements” as there was no evidence that incentivised groups were more likely to underestimate risks or that people on low incomes were more likely to sign up. The authors call for research regulators to relax restrictions on the use of incentives in low risk trials; but their own study shows that incentives sometimes work and sometimes don’t, and it remains a worry that people who need the money most may be more likely to put themselves in potential danger in high risk trials.

JAMA Intern Med doi:10.1001/jamainternmed.2021.5450

Probiotics don’t prevent pneumonia in ventilated patients

Does boosting the gut biome with probiotics improve outcomes in critically ill people? This randomised trial of 2650 patients in 44 intensive care units (ICUs) found no significant difference in ventilator-associated pneumonia among critically ill patients on ventilators who were treated with probiotics (Lactobacillus rhamnosus GG) compared with placebo (21.9% v 21.3%, respectively). There was also no difference in secondary outcomes such as mortality, length of hospital stay, other ICU-acquired infections, or diarrhoea. Other doses, strains, or types of probiotics and different patient groups may yield more positive results. But this useful study suggests that, although the gut biome shows reduced microbial diversity in critical illness, boosting it will not necessarily help to prevent infection or improve outcomes.

JAMA doi:10.1001/jama.2021.13355

Surgical options for bicuspid aortic valve disease

Congenital bicuspid aortic valve disease (which involves gradual fibrosis, calcification, and stenosis) affects 1% of the world’s population, almost all of whom will need valve replacement to improve symptoms and enjoy normal life expectancy. The question is whether transcatheter aortic valve replacement (TAVR) is as safe and effective for bicuspid aortic stenosis as it is for tricuspid aortic stenosis. This registry-based cohort study of matched patients at low surgical risk undergoing TAVR for bicuspid versus tricuspid aortic stenosis found no significant difference in death or stroke at 30 days and at 1 year. There is an inherent potential for selection bias in this type of study, so randomised trials are needed to test the efficacy and safety of TAVR for bicuspid aortic stenosis in more detail.

JAMA doi:10.1001/jama.2021.13346

Eyeing up a new melanoma treatment

Melanomas in the eye (uveal) have low tumour mutational burden (fewer genetic mutations in their cells so less susceptibility to immunotherapy) and poor prognosis (up to 50% of patients develop metastases, mainly in the liver, and median survival in those cases is only a year). This open label, phase 3 trial of 378 previously untreated patients with metastatic uveal melanoma and a particular HLA serotype randomised patients to receive tebentafusp (a drug that can redirect T cells to target glycoprotein 100-positive cells expressed in uveal melanoma) or a control drug (a single immunotherapy agent, such as pembrolizumab, chosen by the investigator as there is no proven standard treatment). One year survival was 73% in the tebentafusp group and 59% in the control group. Progression-free survival at six months was 31% versus 19% in the two groups. Side effects such as fevers and rashes were common, but they improved over time and there were no treatment related deaths.

N Engl J Med doi:10.1056/NEJMoa2103485

Shining a light on a rare complication of the AZ vaccine

Vaccine-induced immune thrombotic thrombocytopenia (VITT) is a rare syndrome, similar to heparin-induced thrombocytopenia, and has been associated with the Oxford/AstraZeneca and Johnson & Johnson/Janssen covid-19 vaccines. The most common sign of VITT is cerebral venous thrombosis (CVT). The proposed diagnosis criteria are post-vaccine CVT, a low platelet count, and antibodies to platelet factor 4. A multicentre UK cohort study of 95 patients who had CVT within 28 days after any covid-19 vaccination found that 70 had VITT and 25 did not. Those with VITT and CVT fared worse, with more intracranial veins thrombosed (3 v 2), a greater risk of extracranial thrombosis (44% v 4%), and higher rates of death or dependency (47% v 16%.) Patients with VITT and CVT do worse than patients with other types of CVT, although treatment with non-heparin anticoagulants and immunoglobulins may improve the outcome. Numbers were small as VITT and CVT are mercifully rare, and the study was underpowered to make robust comparisons between VITT and non-VITT groups or definitive conclusions about treatment options.

Lancet doi:10.1016/S0140-6736(21)01608-1

Ann Robinson is an NHS GP and health writer and broadcaster