Alex Nowbar reviews the latest research from the top medical journals
Pain relief in premature newborns
In a population that much of the medical profession does not encounter, this high-quality randomised controlled trial looks at the safety and efficacy of morphine to relieve the procedural pain of heel lance or retinopathy of prematurity screening in premature neonates who weren’t being ventilated. The trial had to be stopped after randomising only 31 infants (of a planned sample size of 156) as it met its prespecified trial stopping boundary relating to respiratory adverse events in the group receiving morphine. There was no difference in the pain markers between the group receiving morphine and the group receiving placebo, but at this sample size, a small but significant effect would probably be missed.
Annals of Internal Medicine
Genetic testing for kidney and genitourinary disorders
Most of us don’t wake up in the morning for the delights of exome sequencing research studies. But this paper is cool, I promise. Firstly, I’d better explain the background (to the best of my understanding). Whole exome sequencing is similar to whole genome sequencing, but only focuses on the genes that code proteins, which is actually a minority of genes. Genetic variants are then used to make diagnoses. But like whole body imaging studies, incidental findings are common. This study looks at the prevalence of candidate pathogenic variants in 625 genes associated with Mendelian kidney and genitourinary disorders in almost 8000 supposedly healthy adults. The prevalence was many-fold higher than the number of people expected to have these rare disorders, so the authors suggest refining the test before using it routinely in clinical practice to avoid subjecting many patients to evaluation for conditions they do not have. And they suggest doing similar analyses for other conditions to avoid genetic misdiagnosis. It’s good to see careful thinking around genetic testing.
JAMA Internal Medicine
Team-based care in primary care for people with chronic illnesses
Because they are American, the authors of this paper call it “team-based primary care transformation collaborative initiative.” This observational study examines patient’s use of healthcare (visits and costs) in 18 practices who had this team-based intervention compared to 76 other practices. Findings were mixed. Patients in the intervention practices with 2 or more chronic illnesses had fewer hospitalisations and emergency department visits compared to this group of patients in the comparison practices. Patients with fewer comorbidities seemed to have more hospitalisations than patients from comparison practices. That being said, this paper describes an impressive piece of work; a learning collaborative directed at 18 practices and I expect this has had a positive impact on quality of care and patient satisfaction, but this wasn’t measured.
CART therapy for diffuse large B-cell lymphoma
Schuster et al conducted a single-arm Phase 2 study of tisagenlecleucel (an anti-CD19 chimeric antigen receptor T-cell therapy) for relapsed or refractory diffuse large B cell lymphoma. They measured efficacy by response rate in the group that received the therapy (which was actually 111 of the 165 patients who enrolled) i.e. not even the intention-to-treat group. They compared these data with previous observational data to declare the therapy a success. But all these data show is that it is worth proceeding to Phase 3 trials. A randomised controlled trial is the only trial that could be pivotal for patients and the haematology community.
Hydroxyurea for children with sickle cell anaemia in sub-saharan Africa
This is an uncontrolled study i.e. a single-arm study. But the design may be appropriate for three reasons. Firstly, the primary endpoint was safety. Secondly, hydroxyurea is already an established and effective treatment for sickle cell anaemia. Thirdly, the researchers were investigating feasibility in the under resourced setting of Africa where the children commonly have co-existing conditions such as malaria and malnutrition. Therefore it seems pragmatic to assess safety and feasibility by enrolling over 600 children to a single arm involving receipt of hydroxyurea. Apparently it was “reasonably” safe, feasible, and beneficial and therefore worth increasing this population’s access to this therapy. I don’t disagree with this conclusion and the study is very commendable. It’s just that I want to know how a control group would have fared in comparison.
ADHD and month of school enrolment
If you were wondering whether there was an association between attention deficit-hyperactivity disorder (ADHD) and month of school enrollment, you’ve come to the right paper. In this well-conducted US observational study, children born in August had higher rates of ADHD diagnosis and treatment than those born in September. I’m not sure that the many explanations for this observation are worthy of documentation as it doesn’t bring us any closer to preventing or treating the condition. You could attribute it to school entry cut offs, but how else to divide children into year groups? You could attribute it to physiology of pregnancies at different times of year, but advising couples when best to conceive borders on the ridiculous. The utility of the study is a little lost on me.
A test of time and a time of tests
Year 2000. It was the best of times, it was the worst of times. Who can say? All we know from this retrospective cohort study of UK primary care is that we were having fewer tests—1.5 tests per person per year, compared with 5 per year in 2015. The number of laboratory tests increased over time far more than imaging, or other tests, like endoscopy. Renal function tests were the most commonly ordered test. Second was full blood count, and third was liver function tests. No surprises there.
These data are fascinating and beautifully presented. ‘Dem big data can’t tell test-orderers what to do now, though. Ordering more tests might be good if “appropriate” but this is subjective. More tests might be bad if expensive, time-consuming, and leading to patients being exposed to unjustified risk. It is sad, and curious, given the reported patient and public involvement in this study, that the authors imply that test overusage is driven by patients rather than proposing shared decision-making to be part of the solution. If now is to be the best of times, patients should be involved in choice of tests irrespective of the downstream consequences. We only need the time for shared decision-making conversations and better evidence (and perhaps better training) to guide the discussions.
An analysis of respiratory mortality
Salciccioli et al’s analysis found that the UK is an outlier, with higher respiratory mortality than other countries. This sounds alarming, but on closer examination, the UK didn’t actually lie out of the set. And more importantly, the set were the EU15+ countries which are some of the richest in the world, including the US and all the Scandinavian countries, supposedly countries that are “similar” to the UK. I’m not sure being above the median for mortality for one set of conditions in a set of the richest countries in the world is really a cause for concern. By definition, half these countries will have above median mortality.
Moving past the alarmist nature of these kinds of reports, the breakdown by gender and subcategory of respiratory disease in this study does highlight a few interesting areas. Mortality decreased more in men than in women in the UK, and mortality from interstitial disease is rising. It is human to jump to conclusions about causation, but the authors wisely steer clear from this.
Justifying PPI in clinical trials
If you haven’t tried it, patient and public involvement (PPI) in research may sound like a dull box-ticking exercise. It’s not. However, it is hard to teach an old researcher new tricks. Fortunately researchers tend to respond to rigorous meta-analyses, like this paper by Crocker et al. This study assesses the impact of PPI interventions on enrolment and retention in clinical trials, one of the major challenges to the research universe which has finite resources. There was a gentle signal that PPI was good for enrolment in the 19 included trials. For retention there was no particular signal.
I think it’s a push to expect PPI to impact heavily on recruitment and retention. The proportion of people who say no to trials who could have had their minds changed if things were different is unknown. It isn’t a given that if the trial was designed “right” that all patients would say yes. Secondly, patients involved in trial design and management are often “trial-keen” so they aren’t best placed to tackle barriers to recruitment of diverse populations. However, they are well placed to make the research better suited to the experience of that medical condition and to focus on human priorities like satisfaction and ability to function instead of medicalised ones like “progression-free survival.” I look forward to seeing PPI interventions which attempt to engage disadvantaged or hard-to-reach populations. I also wonder if like lay ethics panels, the PPI could be a little independent from the trial, like a jury. Probably a bridge too far to ask of researchers at the moment though. Even I am only talking the talk; there was no PPI in the writing of these comments.
Alex Nowbar is a clinical research fellow at Imperial College London.
Competing interests: None declared