Ann Robinson’s journal review—5 September 2019

Ann Robinson reviews the latest research from the top medical journals


PCI after STEMI 

In ST-segment elevation myocardial infarction (STEMI), opening up the offending or “culprit” coronary artery with percutaneous coronary intervention (PCI), saves lives and prevents further myocardial infarctions (MI). But is it a good idea to open up other occluded coronary arteries seen on angiography before they too cause problems? This study found that achieving complete revascularisation was beneficial; a composite measure of cardiovascular deaths or new MIs was 7.8% in the complete group and 10.5% in the culprit-lesion-only PCI group. There was no increase in major bleeding or stroke rates between the two groups. Timing of the non-culprit-lesion PCI wasn’t critical in this study; complete revascularisation showed consistent benefit regardless of whether it was performed during the initial hospital stay or up to several weeks later. It would be handy to know whether cleaning up all the significantly blocked coronary arteries at the same time as the PCI for STEMI is more effective and as safe as doing it at a later time but this study couldn’t evaluate that. 


HRT and breast cancer—a landmark study

A major review of worldwide evidence on hormone replacement therapy (HRT) and breast cancer is required reading for all prescribers. In a nutshell; the risk of breast cancer in women aged 50-69 who haven’t taken HRT is 6.3% and the risk increases to 8.3% in those who have taken oestrogen and daily progestagen (over half HRT users) for 5 years or more. Some excess risk persists for longer than 10 years after stopping. 

The authors estimate that around 1 in 20 breast cancer cases in the UK are attributable to HRT and that  around 1 in 50 HRT users will develop breast cancer that they would have otherwise avoided. And yet and yet. The increase was in breast cancer incidence, not mortality. Women will want to take into account other health benefits, impact on quality of life, the fact that using HRT for less than a year doesn’t increase breast cancer risk, the impact of duration of progestagens and their own baseline level of risk. 

Rather counterintuitively, one group that doesn’t need to worry about this study’s findings are women who are obese; their level of risk is already raised by being obese but not further raised by taking HRT. The study tells us what we know, but don’t always practice; every potential HRT-taker needs to be helped to conduct her own cost-benefit analysis before making an informed decision about whether, which, and for how long to take HRT. This study should help inform practitioners about what to advise their patients.

Safe delivery

This multicentre, randomised trial across England and Wales compared planned delivery versus expectant management (usual care) in women with late preterm pre-eclampsia (34 to 37 weeks gestation). They found strong evidence that planned delivery is preferable, with less maternal morbidity and severe hypertension. There was no increase in neonatal morbidity despite more neonatal unit admissions than with usual care. The authors urge shared decision making to decide on the timing of delivery with the pregnant women in these cases.

Annals of Internal Medicine

Suicide prevention—which interventions work?

Which interventions are effective in preventing suicide? US researchers analysed eight systematic reviews and 15 randomised controlled trials as part of an updated clinical practice guideline. They found that cognitive behavioural therapy, and its offshoot, dialectical behaviour therapy, reduced suicidal ideation to a modest extent compared to the ill-defined category of “treatment as usual” or keeping the person on a waiting list. Cognitive behavioural therapy appeared to reduce the risk of suicide attempts compared to treatment as usual. Lithium reduced the suicide rate in uni or bipolar mood disorders and ketamine, recently licensed as a nasal spray for use in major depression, reduced suicidal ideation compared to midazolam or placebo, with few side effects. Other interventions (drug or non drug) may be successful, but evidence was too limited to say. 

There are some inherent problems in mashing up new evidence with historical meta-analyses. The quality of included studies was very variable and whereas drugs are drugs, people can mean different things when it comes to non pharmacological treatments such as cognitive behavioural therapy. Finally, data on drug treatments and their potential harms was hard to tease out. But with suicide rates at 13.5 per 100,000 people in the US (10.1/100,000 in the UK), and a rise in reported suicidal thoughts, attempts and planning among 18-25 year olds over the past decade, research in this area is needed.

Small and large brain lesions—a marker for cognitive decline?

Do structural brain changes seen on MRI in healthy people, correlate with subsequent cognitive decline? This longitudinal cohort study studied people aged 50 and over who hadn’t ever had a stroke, but had an MRI of their brain in 1993-5 and up to five cognitive assessments over the subsequent 20 years. The MRIs showed either no infarcts (86%), small ones of <3mm (3%), larger ones of 3-20mm (10%) or a mix of smaller and larger (2% or 35 participants). The good news is that smaller and larger infarcts weren’t associated with more cognitive decline than those with none. But curiously, the admittedly small number of participants with mixed small and large infarcts did show steeper decline. The authors suggest that the combination may represent more pervasive disease as both small and larger arteries are affected at the same time, possibly by different processes. It could be a marker of risk for cognitive decline and provide opportunities for prevention.


Too much vitamin D?

Has enthusiasm for vitamin D gone too far? A reported 3% of US adults take over 4000 IU of vitamin D a day, presumably on the basis that if deficiency is bad for you, then an excess must be good. But does higher dose vitamin D supplementation improve bone mineral density (BMD) and strength? This small randomised trial of 311 healthy adults found that it doesn’t; bone density (radial) was significantly lower in people who were given higher doses (4000 or 10,000 IU/day) compared with those taking the lower dose of 400 IU/day. Tibial BMD was also significantly worse in the group that took 10,000 IU/day compared with those taking 400 or 4000 IU/day. There was no significant difference in bone strength at the tibia or radius between the three groups. The question raised by this study is; does high dose vitamin D actually have a negative effect on bone? Further study of that important question and other potential harms is needed.

JAMA Internal Medicine

Vegetarianism associated with long term health

This large prospective cohort study provides further confirmation that people who get their protein from plants rather than animals live longer—especially compared with people who eat red or processed meat. Researchers followed 70 696 Japanese adults for a mean of 18 years and found that swapping some animal protein for plant or fish protein was associated with lower all cause, cancer related, and cardiovascular disease related mortality. The authors admit that: “plant protein intake may represent a healthy eating behaviour; although adjustment for several lifestyle factors showed little difference in the overall results, the possibility of residual confounding in the association between plant protein and mortality remains.”

Ditch the urine dipping?

This retrospective cohort study of 2733 hospitalised adults with asymptomatic bacteriuria (ASB) found that 82.7% were given inappropriate antibiotic treatment, especially if they were elderly or confused. Patients with ASB who received antibiotics had a 37% longer hospital stay (4 versus 3 days) compared with those who weren’t treated. “Despite common misperceptions, positive urinalysis or urine culture results do not define a urinary tract infection (UTI) or necessitate antibiotic therapy,” say the authors, echoing US national guidelines. Clearly this sort of study can’t prove causation; patients with ASB may have volunteered symptoms of a UTI that weren’t recorded and despite the authors’ best efforts, confounding factors may abound. But treating ASB is contrary to guidelines, associated with potential harm, and widespread in a hospital setting. Urinalysis in asymptomatic, stable patients may be unnecessary.


Ann Robinson is an NHS GP and health writer and broadcaster