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

Persistent covid-19 symptoms

A group of Italian researchers examined what people experience after recovery from covid-19 in the months after discharge from hospital. Of the 143 men and women surveyed, now testing negative for SARS-CoV-2, only 12.6% had no persistent symptoms and 55.6% had three or more persistent symptoms. Ongoing fatigue, breathlessness, joint pain, and chest pain were commonly reported. The applicability of the results to the general population of people who have had symptomatic covid-19 is low because the sample size was small and included only people hospitalised for their illness. The survey is also limited because it is based on participant recall of symptoms—that is, symptoms could have been present before the covid-19 illness, and symptoms are very subjective. You’d better keep searching if you want to know how long these symptoms might persist, or even why they are happening, as this study is not able to explore this.

JAMA doi:10.1001/jama.2020.12603

Stillbirths during the pandemic

This retrospective study compared stillbirth rates at one London hospital in the three months before the first reported UK cases of covid-19 and the following three months. The rate before the pandemic was 2.38 per 1000 births and during was 9.31 per 1000 births. There are many possible explanations for this difference, including the play of chance. It seems reasonable to speculate that the pandemic had something to do with it, albeit indirectly as the affected women did not have covid-19. Possible explanations include this hospital being referred more high risk cases during the pandemic, changes in obstetric care due to covid-19 (such as reduced antenatal services), and perhaps reluctance to attend hospital when needed. While research in this area is vital, this particular study does not really further our understanding of the relation between covid-19 and pregnancy outcomes.

JAMA doi:10.1001/jama.2020.12746

How probiotics didn’t reduce infections

Butler and colleagues studied the effect of probiotics on days of antibiotic administration in one year in a double-blind randomised controlled trial of 310 UK care home residents aged over 65 years. The trial found no difference in antibiotic administration for any infection between the probiotic group and the placebo group. Days of antibiotics is a more objective, easy to measure, and continuous outcome than infections and is probably a relevant outcome because of the harms associated with antibiotic use. Death rates were similar in both groups. Weirdly, it is even possible that probiotics increased the rate of lower respiratory tract infections (one of the secondary outcomes). The hunt for interventions to reduce infections in care homes continues. I hope that this well designed trial will end the hopes of probiotic enthusiasts, at least with the intention of preventing infections.

JAMA doi:10.1001/jama.2020.8556

Baloxavir for influenza prophylaxis after exposure

Flu is still important even if covid-19 continues to dominate everything. This trial tests whether baloxavir is useful for post-exposure prophylaxis against influenza. Over 700 people in Japan who had been exposed to influenza through household contacts were randomised to receive a single dose of baloxavir or placebo in a double-blind fashion. The primary endpoint was clinical influenza with a positive PCR test in 10 days. Dramatically fewer people in the baloxavir group developed clinical influenza compared with the placebo group without an adverse safety signal for the drug. This benefit was present in the paediatric and the high risk subgroups, which is reassuring. And so it seems that this drug will be a useful intervention for preventing influenza. Hopefully it can be delivered with similar effects in real life. One way in which these results might not be generalisable to other populations is that the index patients, who could pass flu onto the participants in this study, tended to receive early antivirals (as per standard practice in Japan), which could have influenced the transmission rates. It is unclear whether this would change the outcome of the study though, especially since the difference between groups was so substantial.

N Engl J Med doi:10.1056/NEJMoa1915341

Breast cancer screening after childhood chest radiation

Modelling studies can range from being slightly off in their predictions to plain wrong. However, this study is a case where modelling seems like an appropriate approach. Women who had chest radiation for childhood cancer live with an increased cancer risk and thus undergo breast cancer screening. The frequency and type of screening is very challenging to test in randomised studies. Using existing published data (taken from survivors with cancer diagnosed between 1970 and 1986), Yeh and colleagues modelled initiating screening at ages 25, 30 or 35 years and compared annual MRI with annual MRI plus mammography. They concluded that early initiation (ages 25-30 years) may reduce breast cancer deaths and that the addition of mammography was beneficial without substantially changing rates of false positives and benign biospies. These modelling data may be some of the best data available to plan screening strategies, or at the least, inform risk versus benefit discussions with patients.

Ann Intern Med doi:10.7326/M19-3481

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

Competing interests: none declared.