Richard Lehman’s journal review—3 April 2017

richard_lehmanNEJM  30 Mar 2017  Vol 376

Rivaroxaban vs aspirin long after VTE

The longer you give anticoagulants to people following venous thromboembolism, the fewer subsequent thromboembolic events they will have. At some point you have to decide whether to go on or to stop, and this Bayer-funded trial has recruited participants who “had completed 6 to 12 months of anticoagulation therapy and were in equipoise regarding the need for continued anticoagulation.” Bayer produces the direct oral anticoagulant rivaroxaban which has plenty of patent time left and costs about £800 per patient per year in the NHS. The trial compared fixed doses of rivaroxaban (10mg and 20mg) with 100mg of aspirin daily. Not surprisingly, rivaroxaban proved better at preventing recurrent VTE over the course of a year than aspirin: absolute rates were 1.5% (20mg), 1.2% (10mg) and 4.4% for aspirin. It was a typical industry trial involving 3396 patients from 244 sites in 31 countries with Bayer as the data holder and main analyst. But that’s not my beef today: what puzzles me is the lack of a definition of the “equipoise” that these patients were supposed to be in. This isn’t specified in the article or the supplement, and the table of patient characteristics shows a very heterogeneous picture. I do hope Bayer will make the individual participant data available without delay.

Quadrivalent HPV vaccine & pregnancy outcomes

A few weeks ago, I unintentionally strayed into a minefield on the subject of the safety of quadrivalent vaccines to prevent human papillomavirus infection. This time I will tiptoe around the circumference very carefully. In a nationwide case-control study from Denmark, women who had vaccine exposure during the prespecified time windows were matched for propensity score in a 1:4 ratio with women who did not have vaccine exposure during the same time windows. In these matched analyses, exposure to the quadrivalent HPV vaccine was not associated with significantly higher risks than no exposure for major birth defect, spontaneous abortion, preterm birth, low birth weight, small size for gestational age, or stillbirth.

Inotropic “support” during cardiac surgery

When people with heart failure are given drugs to stimulate the heart, they tend to die faster. If they feel better in the meantime, I would say that the choice should be theirs. But in this trial of levosimendan the patients were mostly asleep. The inotrope was given to patients just before heart surgery and the hope was that it would help to reduce post-operative low cardiac output syndrome. The composite end-point of the trial was a very generous mix, but despite that and a well-powered design, levosimendan failed to make any difference.

How open data gets used

Five years ago, when I had the amazing fortune to be in at the start of the Yale University Open Data Access project, data sharing was a bold and threatening idea which provoked much overt and covert opposition. Now it has become motherhood-and-apple-pie, and the best that its opponents can come up with is that mothers are nice and so are apple pies, but do we really need to divert so much attention to them when mum is in a comfy nursing home and there are so many other pies on the supermarket shelves. Let’s not forget that even now, many pharmaceutical companies only allow very limited access to their data and very few academics show enthusiasm for sharing. And interest in working on full data sets is relatively limited as yet, since it’s not a major draw for funders. But a survey of data requests made to the US National Heart, Lung, and Blood Institute data repository shows what I hope is the beginning of a surge. From January 2000 through May 2016, a total of 370 investigators requested data from 1 or more clinical trials. Requests for trial data have been increasing, with 195 investigators (53%) initiating requests during the last 4.4 years of the study period. Although I would argue that every trial with important consequences for clinical practice and/or health spending should undergo independent individual participant analysis, very few people are doing this. “Demand for trial data for secondary analysis has been increasing. Requesting data for the a priori purpose of reanalysis or verification of original findings was rare.” Just look back at that Bayer trial of rivaroxaban—wouldn’t it be good to delve into the individual data before committing to spending £800 per patient per year for hundreds of thousands of people?

JAMA 28 Mar 2017  Vol 317

Vit D & calcium: no effect on cancer incidence in older women

There is so much conflicting observational evidence about vitamin D that I felt I had to give honourable mention to this randomised controlled trial, even though it tells us little. Here is what: Among healthy postmenopausal older women with a mean baseline serum 25-hydroxyvitamin D level of 32.8 ng/mL, supplementation with vitamin D3 and calcium compared with placebo did not result in a significantly lower risk of all-type cancer at 4 years.

Hysterectomy type in early endometrial cancer

Surgical triallists are among my heroes, and I follow their work from time to time in these reviews, and sometimes by chatting with Peter McCulloch about his wonderful IDEAL Collaboration. Here’s a classic example of a trial which directly informs clinical practice: the Laparoscopic Approach to Cancer of the Endometrium (LACE) trial. It was conducted between 7 October 2005, and 30 June 2010, in 20 tertiary gynaecological cancer centres in Australia, New Zealand, and Hong Kong. 760 women with stage I endometrial cancer to either total laparoscopic hysterectomy or total abdominal hysterectomy, and follow-up ended on March 2016. So over 12 years from setting up the trial to publication in JAMA. The two methods of hysterectomy resulted in equivalent disease-free survival at 4.5 years and no difference in overall survival.

Plumbing the effects of lead in children

The city of Dunedin in the South Island of New Zealand was mapped out in advance using rulers on a map in Edinburgh (syn Dunedin), prior to mass migration there by members of the Free Church of Scotland. Presbyterian rulers are hard and straight, and so in consequence were the roads of Dunedin NZ, which take no account of local topography. Anyone who has tried to drive in Dunedin will know what I mean. Chugging and stalling on the nearly vertical slopes, the cars and lorries of yesteryear would have belched out the fumes of barely combusted leaded petrol. Children gasping up these mountains inhaled measurably more lead than those in most other cities. Although the dangers were known at the time, amelioration came too late for many. In a cohort of just over 1000 children whose lead levels were tested at the age of 11 in 1972-3, high exposure was associated with lower cognitive function and socioeconomic status at age 38 years, and with declines in IQ and downward social mobility.

JAMA Intern Med  Mar 2017  Vol

Does acid suppression increase C diff recurrence?

Like many doctors, I’ve been told that gastric acid suppression increases the risk of recurrent Clostridium difficile infection. This systematic review shows how weak the basis for that assertion is. Nobody has done a prospective randomised trial. OK, that would be difficult, but our knowledge at present is derived from 16 heterogeneous observational studies with a total of 7703 patients with CDI, of whom 20% developed recurrence. The rate in patients taking acid suppressors was 22.1% compared with 17.3% in those who were not. Taken at face value, that’s a statistically significant difference, but there is so much scope for bias and residual confounding that I wouldn’t attach any meaning to it.

The Lancet  1 Apr 2017  Vol 389

Multiple sclerosis

Finding little to tempt me in the research papers of this week’s print Lancet or the website, I decided to dive in to their reviews sections on multiple sclerosis. Let me test your patience with my reflections on MS over 45 years. The positives first: diagnosis has improved immeasurably. Classification has improved a little. Treatments abound, but the choice is difficult in relapsing/remitting MS and non-existent for progressive MS. The thing that has improved most is personal care for patients, thanks to community specialist nurses and dedicated MS centres.

Now for the negatives. MS still leads to severe disability in 80% of patients, with a mean loss of 10 years of life. Although there is now a choice of 14 drug treatments, they all work on inflammatory pathways and cause a variety of serious or lesser adverse effects, roughly in proportion to their efficacy. The progressive forms of MS are probably driven by other mechanisms and no adequately tested therapy has shown any effect.

These would be good teaching papers for any course on the shared understanding of medicine, should such a course ever be offered. The diagnosis paper illustrates all the dilemmas of real-life diagnostic pathways and how we should conceptualise and communicate about them. The treatment paper shows how difficult it is to avoid excessive reliance on surrogate outcomes which relate very poorly to clinical outcomes. How can you do conventional shared decision making when you have an offer of 14 variously tested drugs, plus non-drug interventions like bone marrow transplantation? Can we expect patients to understand these when we don’t fully understand them ourselves? How much should be left to clinical expertise, and how can we bring kindness into the equation?

The BMJ  1 Apr 2017  Vol 356

Steroids in pregnancy

Antenatal corticosteroids gave birth to evidence-based medicine as we know it. They are also good for premature babies. Forty years since Iain Chalmers began his transformative work in this area, here is a prospective cohort study of the effect of corticosteroids on infant outcomes, based on data from 300 neonatal intensive care units in the United States. “The effect on mortality and survival without major morbidities of exposure to antenatal corticosteroids seems to be largest in infants at the lowest gestations, including infants at 23 weeks’ gestation who were not included in randomized controlled trials.” The study also found that antenatal exposure to corticosteroids was associated with lower rates of mortality and major hospital morbidities at most gestations for which steroids are currently recommended. This study supports the administration of antenatal corticosteroids in women with threatened preterm labor from 23 to 34 weeks’ gestation.” (As usual with The BMJ, American spellings remain in American papers.)

Fewer antimalarials, more antibiotics?

I have never worked in Africa, but those who have tell of moribund feverish children carried for hours or days to remote medical centres, where many of them die. What would you do in that situation? I’m sure I would give them a cocktail of whatever antimalarials and antibiotics happened to be in the cupboard, plus fluids by whatever route was feasible. How deplorable this sounds in the comfort of air-conditioned offices thousands of miles away. Here’s a paper by nearly 40 authors which assesses the effect of introducing near-patient rapid malaria testing in 8 randomised trials and one observational study, in Afghanistan, Cameroon, Ghana, Nigeria, Tanzania, and Uganda. In 522 480 children and adults with acute febrile illness, use of antimalarials fell, but antibiotic use remained high at 69% of those with a negative malaria test and 40% of those with a positive test. The authors state that this “probably represents overprescription. This included antibiotics from several classes, including those like metronidazole that are seldom appropriate for febrile illness, across varied clinical, health system, and epidemiological settings.” Perhaps. But first prove that attempts to reduce antibiotic prescribing do not lead to more dead children and adults in these settings.

Insulin: pumping vs injecting

“Adding pump treatment to structured training in flexible intensive insulin treatment did not substantially enhance educational benefits on glycaemic control, hypoglycaemia, or psychosocial outcomes in adults with type 1 diabetes.” That’s really all you need to know about the REPOSE trial. You might like to send this paper to your local diabetes centre the next time they prescribe a pump to one of your patients. You might also like to find out if they receive money from a manufacturer of these pumps.

Plant of the Week: Pulmonaria Opal “Ocupol”

Early spring is the time for plants which are named after internal organs. If you are lucky, you will have enjoyed the dainty flowers of hepaticas over recent weeks, and will now be admiring your pulmonarias. In each case, the leaves are what caused our forefathers to call these plants liverworts and lungworts, or their equivalents in Latin: the old leaves of hepaticas are a liverish colour and shape, while pulmonarias have leaves that are spotted like cut lungs.

Pulmonarias come in a patriotic range of reds, whites, and blues. Left to their own promiscuous devices, they tend to have baggy leaves and flowers that vary between pink and blue on the same stem. But selectors and hybridizers have done sterling work over the last 15 years and there are now some wonderful pulmonarias available in every garden centre. “Opal” is one of them. The leaves are positively emphysematous, with more white than green, but all the more beautiful for that, while the flowers are open and of the softest clear blue.

All pulmonarias are tough native perennials by nature, though I have found that the red ones don’t like hard winters. You can split them at will, and now that there are so many good clones, I would suggest you stick with doing that, while banishing any miscegenous seedlings to the further reaches of your garden or your compost bin.