Alex Nowbar reviews the latest research from the top medical journals.

Annals of Internal Medicine

Gadolinium based agents and risk of nephrogenic systemic fibrosis

Nephrogenic systemic fibrosis (NSF) is one of those rare conditions that knowing about made you feel proper clever at medical school but you’ve still never seen a case and nor are you likely to. It is known for its association with giving gadolinium-based contrast for MRI in people with renal failure and for being a potentially devastating condition. Lunyera et al’s systematic review found no cases of NSF with newer gadolinium-based contrast agents in one set of studies including over 80,000 patients. In another set of studies, again with tens of thousands of patients, there were 4 cases with the newer gadolinium-based agents and 37 cases with the older agents. So it is extremely rare. Giving it to people with renal failure remains a concern though and the studies were not able to address how to identify those at most risk. But this is always methodologically difficult when an adverse outcome is rare. This review is based on large observational studies which is a strength, but it could underestimate the risk because people at higher risk will not have been given gadolinium.

JAMA

Deaths in US nursing homes during the pandemic

Barnett et al’s analysis includes 189 “skilled nursing facilities” in Cleveland, Detroit and New York, making up a third of the facilities in these cities. Detroit and New York had a lot of covid cases in the study period (March to May 2020) while Cleveland had much fewer. Death rates (from any cause) in the skilled nursing facilities matched this i.e. there were spikes in the ones in Detroit and New York, but less in Cleveland ones. Data were collected from electronic health records aggregated by a software company called CarePort Health. The most interesting comparison in this study was between death rates in 2019 and 2020 in the skilled nursing facilities; death rates were higher in 2020 in all three regions’ facilities, but the rise was not statistically significant in Cleveland. Admissions to these facilities went down during the study period. I am not sure what purpose this study serves. It highlights the jump in deaths in these units. Perhaps this could have been prevented e.g. with better provision and use of personal protective equipment, but this study doesn’t shed any light on that.

Lancet

Valve versus valve

The PORTICO trial randomised 750 high risk aortic stenosis patients to the self-expanding, repositionable Portico valve or other commercially available TAVI valves. The Portico was found to be non-inferior for efficacy, defined as 1 year mortality or disabling stroke, and safety. For the safety endpoint, there were numerically more, in fact quite a lot more, events in the Portico group. Paravalvular leak rates and pacemaker rates were worse. Some of this could be accounted for by operator experience with the Portico. (Quite why patients would be prepared to try a valve that their operator isn’t experienced with is beyond me.) The authors conclude “The first-generation Portico valve and delivery system used in this study does not offer advantage over other commercially available valve”. I think this is an understatement. It actually looks worse.

NEJM

Triple therapy for COPD

Triple therapy means a lot of things nowadays. An anticoagulant with 2 anti-platelets. A proton-pump inhibitor and 2 antibiotics. Then there was triple therapy for asthma and now there’s triple therapy for COPD too. The ETHOS trial tested a combination of an inhaled steroid, long-acting beta agonist (formeterol) and long acting muscarinic antagonist (glycopyrrolate) against dual therapy (either formeterol and steroid or formeterol and glycopyrrolate) in a randomised fashion in people with at least moderate COPD and at least one exacerbation in the previous year. Triple therapy showed an improvement in the primary endpoint (rate of moderate or severe COPD exacerbations) compared to dual therapy at one year. I can tut about the fact that this pharma-sponsored trial was analysed by AstraZeneca and written by a medical writer paid for by the sponsor, but this evidence does seem robust.

Safety and doctor shift length

There are long running arguments (and indeed a variety of trials) about shift lengths for doctors. There is the longer shift equals more experience and hence better doctor in the long-term argument. And there is the longer shift equals tired doctor and hence worse doctor in the short-term argument. Testing hypotheses like these is not straightforward because any intervention takes place in a complex system where it is difficult to control all the other factors. However, the hypothesis that longer hours lead to more errors is important to test for planning work schedules going forward. Landrigan et al compared the patient safety of work schedules for paediatric residents in intensive care units in the US with or without extended shifts (24 hours or more) using a cluster-randomised, crossover design. Physicians randomised to extended shifts made fewer serious errors. This is counterintuitive because these physicians had had less rest. However the study found that their workload was lighter compared to those without extended shifts. The difference in error rates was eliminated when the results were adjusted for number of patients per physician as a potential confounder. I am not sure we are any the wiser after this study.

Alex Nowbar is a clinical research fellow at Imperial College London, UK.

Competing interests: none declared.