Alex Nowbar’s weekly research reviews—26 February 2019

Alex Nowbar reviews the latest research from the top medical journals

Annals of Internal Medicine

Cardiovascular risk in China

It is hard for those of us who work in well developed healthcare systems to imagine a world without primary prevention for cardiovascular disease. Lu et al report on a national cardiovascular screening project in China, where 1.7 million people aged 40 to 75 with no known cardiovascular disease were screened. As one might expect, 9.5% had high cardiovascular risk. The more pertinent finding was the incredibly low rates of statin and aspirin prescriptions. Less than 3% of those at high risk were receiving statins or aspirin. There was also substantial underuse of antihypertensives, although this was not nearly as notable as for aspirin and statins. This study makes a clear case for infrastructure to educate and offer primary prevention on a large scale, despite financial and cultural barriers. The exciting prospect is that plans for this are being combined with research and patient centredness.


Testing in septic shock

Both capillary refill time and lactate levels have limitations as tests for tissue hypoperfusion. Capillary refill time has obvious appeal as it is cheaper and easier and fits into the popular paradigm that clinical assessment is more useful than biochemical assessment.

The ANDROMEDA-SHOCK trial compared two strategies to guide resuscitation in patients with septic shock requiring vasopressors. It took place in intensive care units in five South American countries. Over 400 patients were randomised to one of two resuscitation strategies. One strategy aimed to normalise capillary refill time which was performed every half an hour using a chronometer (presumably that just means formally timed). The other strategy was to normalise or decrease lactate levels at rates greater than 20% per 2 hours with lactate checks every 2 hours. The intervention only lasted 8 hours. At 28 days, 34.9% of patients in the capillary refill time group had died whereas 43.4% in the lactate group died. This difference was not statistically significant; the hazard ratio was 0.75 with a 95% confidence interval of 0.55 to 1.02.

Lack of blinding is likely to have affected these results as knowledge of the treatment group could have influenced the other treatments given during and after the intervention period. But I commend the authors for conducting a challenging trial. We are a very small step closer to intensive care being less of a dark art.


VOT is the new DOT

Video observed therapy for tuberculosis was compared with directly observed therapy in a randomised controlled trial set up at 22 UK sites. Video observed therapy was done by a daily recording on a smartphone app, whereas directly observed therapy was three to five times a week with a healthcare or lay worker. Video observed therapy was found to be more effective and cheaper over a two month treatment period. These data are promising for the target population of what they refer to as “socially complex” patients. There is a caveat to this though. Although smartphones were provided free of charge by the study team, patients who did not have access to a place to charge a phone were excluded from the study. It is disappointing that this was an exclusion criteria as it is likely to discriminate against the very people this intervention is supposed to be for. Various creative solutions could have been implemented to address this, but never mind.

The authors rightly state “The study was not powered to detect differences in culture conversion rate, treatment completion, loss to follow-up, relapse, or development of drug resistance.” Then they say “Nevertheless, it is reasonable to assume that improved adherence might improve all these outcomes.” I would say, it is also reasonable to assume that less of this sort of assumption might improve all medical research. Also of note, the declaration of interests states “RG reports involvement with SureAdhere Mobile Technology (but no funding was received from the company for work on this study).” I don’t see RG’s involvement as a problem per se, but perhaps the wording of the declaration could be more direct. RG is co-founder of SureAdhere.

Anal incontinence and caesarean delivery

Using a Swedish nationwide registry, Larsson et al compared rates of anal incontinence between four groups: women who had had a caesarean delivery, women who had had a vaginal delivery, age matched nulliparous women, and age matched men. They included women who had given birth between 1973 and 2015 and excluded women who had had a multiple birth delivery, those who had four or more deliveries, and those who had had both vaginal and caesarean deliveries. This enabled them to make the best assessment of comparative risks. They found a lower risk of anal incontinence after caesarean delivery compared with vaginal delivery, although not as low as that of nulliparous women, which suggests aspects of the pregnancy other than the mode of delivery contribute to anal incontinence. Of the four groups, men had by far the lowest risk of anal incontinence.


Invasive Group B Streptococcus

It is easy to become disillusioned about epidemiological studies because the findings can seem far removed from day to day clinical practice. This study of invasive Group B Streptococcus in non-pregnant adults in the US from 2008 to 2016 bucks the trend by being interesting and also concerning. Invasive Group B Streptococcus was defined as Group B Streptococcus isolated from a normally sterile site (blood, joints, and bones most commonly). Francois Watkins et al found that the incidence has been increasing. Invasive Group B Streptococcus was more common in men than women, in black people than white people, and in older age. In 2016, the incidence in people over 80 was more than 40 cases per 100000 people. Antibiotic resistance rates are also increasing. Ninety five percent of affected people had at least one underlying condition, usually obesity or diabetes. Perhaps understanding the characteristics of who is affected by this condition will provide insight into how Group B Streptococcus transfers. All in all, this important study provides an excellent basis for planning strategies for prevention and treatment of this condition.


Prevention of venous thromboembolism in cancer

More than 10 years ago, the Khorana risk score was created to identify risk of venous thromboembolism in patients due to start chemotherapy for a solid tumour. Higher scores indicate higher risk of venous thromboembolism, with the main driver being a high risk cancer site and high platelet levels. The hope was that these patients could be given thromboprophylaxis and that this would be good in all manner of ways. The CASSINI trial (on which Khorana is an author) and AVERT trial tested rivaroxaban and apixaban respectively in patients with cancer in a double blinded fashion. Rivaroxaban (10 mg daily) did not reduce the primary endpoint of deep vein thrombosis, pulmonary embolism, or death from venous thromboembolism, even with screening ultrasound of both legs every eight weeks to pick up asymptomatic venous thromboembolism. There was no difference in major bleeding between groups. Apixaban (2.5 mg twice a day) showed a reduction in the primary endpoint of deep vein thrombosis or pulmonary embolism (half of the pulmonary embolisms in the study being incidental findings), but also a substantial increase in major bleeding events. So, there you have it.


Induction of labour

The INDEX trial randomised 1801 women in the Netherlands to induction at 41 weeks or expectant management with induction at 42 weeks if necessary. The primary outcome was a composite of perinatal mortality and neonatal morbidity. It was designed as a non-inferiority trial. Expectant management was not found to be non-inferior to induction. There were no differences in maternal outcomes or rates of caesarean section. This is a key piece of evidence for planning deliveries going forward.

Need new knee?

Wilson et al conducted a thorough meta-analysis of trials comparing unicompartmental and total knee replacement. Unlike a traditional meta-analysis this paper focuses on patient relevant outcomes. The authors divided studies into three groups: randomised controlled trials, large registry studies, and large cohort studies and for each outcome reported the findings of each group of trials. Interestingly unicompartmental knee replacement had better patient-reported outcome measure scores in the 6 meta-analysed randomised controlled trials. Unicompartmental knee replacement did better in many areas actually, including shorter hospital stay and quicker recovery. And although observational data suggest increased revision rates with unicompartmental versus total knee replacement, this was not reflected in the one randomised controlled trial that assessed revision rates at 15 years.

For a high-end scientific paper, this is an easy read and serves as a good summary of published data that should be useful for shared decision-making.

Shouldering the burden

Craig et al’s cohort study of over 50,000 elective shoulder replacements using hospital episode statistics in the UK found increasing rates of shoulder replacement from 1998 to 2017. The authors report useful statistics on lifetime risks of revision surgery. Risk of revision surgery was higher in younger patients with a striking risk of one in four for men aged 59 years and younger.

The reported risks of serious adverse events have to be interpreted with more care though. 1 in 22 had a serious adverse event at 90 days. This could have been pulmonary embolism, myocardial infarction, lower respiratory tract infection, acute kidney injury, urinary tract infection, cerebrovascular event, or death. It’s easy to assume that the rate of these events would have been zero if the patient had not undergone the operation. But outcomes without the operation can only be assessed properly in a randomised controlled trial.

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

Competing interests: None declared