Richard Lehman’s journal review—16 January 2017

richard_lehmanNEJM  12 Jan 2017  Vol 376

PROMS and PROs
I first went to the Proms in 1966. I enjoyed the queuing, the atmosphere, and the music: young Barenboim playing Beethoven with Boulez, Kertesz conducting Dvorak, Heather Harper singing something English and so forth. So long ago. Now the word prom seems to mean a passing-out ritual (in both senses of the word) by school-leaving teenagers, or Patient Reported Outcome Measures. The Americans prefer PRO for the latter, but that is equally confusing. PROMS/PROs have become something of an industry over the last ten years or so. The title of this Viewpoint says it all—”Patient-Reported Outcomes—Harnessing Patients’ Voices to Improve Clinical Care.” They are a way of harnessing patient voices. Once harnessed, these can be converted into metrics and tick-boxes and studied quantitatively at a distance. All this is well-described in this article, which ends by saying that “There is sufficient scientific rationale and understanding of implementation methods to expand collection of PRO data in clinical care. Doing so could turn the rhetoric regarding ‘patient-centered care’ into a reality.” But there is a much better, quicker way of doing that. It consists of  unharnessing patients’ voices and actually listening to them. This is called Experience-Based Co-Design and it works in real-time cycles where patients have the power to change things. No wonder it hasn’t caught on.

Drugs fail for teenage migraine
While I’m reminiscing about my teen years, I may as well tell you that they marked my first experience of appalling one-sided headaches preceded by flashing lights. I had no idea of what they were but imagined that they must be due to my general unfitness for human life. Aspirin and paracetamol, with or without codeine, had no effect on them. Fifty years on, we are no further forward, except perhaps for triptans. Preventive drugs are generally useless: “There were no significant differences in reduction in headache frequency or headache-related disability in childhood and adolescent migraine with amitriptyline, topiramate, or placebo over a period of 24 weeks. The active drugs were associated with higher rates of adverse events.”

JAMA  10 Jan 2017  Vol 317

Global BP trends and outcomes
A massive survey links blood pressure data from 154 countries with deaths from hypertension-related causes in 195 countries between 1990 and 2015. Yes, you read that right. There is extrapolation going on, which we are assured is OK because this study employs complex, state-of-the-art methods including spatiotemporal Gaussian process regression. In fact, as it modestly states near the beginning, “This analysis supersedes all previous global burden of disease study results for SBP because all data from 1990 to present have been re-analyzed using consistent methods.” It doesn’t get any easier, and although this article is open-access, I defy you to read through it without wondering what half of it is about. The analytic methods may be state-of-the art and consistent, but that makes it all the harder to interrogate the assumptions which lie behind them. The general drift seems to be that although levels of “high” systolic BP (beginning at 110!) are rising in most countries, the mortality associated with them is tending to decrease. The editorialist who comments on this analysis also shows some signs of struggle but concludes that “the project works to be transparent about its complex methods, which remain challenging for most readers, but it also works to encourage global participation, which has increased since the project began. Even if the extensive amounts of data are fuzzy and imperfect, they provide valuable estimates of current global disease burden.”

JAMA Intern Med  Jan 2017

Wrong either way
“Clinicians rarely had accurate expectations of benefits or harms, with inaccuracies in both directions. However, clinicians more often underestimated rather than overestimated harms and overestimated rather than underestimated benefits. Inaccurate perceptions about the benefits and harms of interventions are likely to result in suboptimal clinical management choices.” This is the massively important finding of a systematic review of physicians’ estimates of the benefits and harms of treatments, tests and screening. Everybody needs to read this—in full, as it is open access. There is a particularly trenchant paragraph near the end which asks how we can really share decisions with patients when we are often in the wrong ball park ourselves. Tools for shared decision making, based on continuously updated best evidence, are at last going to appear in quantity over the next year or two. They’ll be designed to be used in real-time consultations to help patients decide on treatment choices. But I suspect their main value will be as knowledge vectors for doctors.

Weekend warriors
A weekend warrior is a person who gets most of their physical activity done at weekends. This is a highly British pattern, and its merits are vindicated in a study of 11 cohorts from the Health Survey for England and the Scottish Health Survey, who completed questionnaires about physical activity between 1994 and 2012. Data from 63 591 respondents of mean age 58 were then linked with mortality registers to see what patterns of physical activity seemed to postpone death most. “Inactive” participants reported less than 150 minutes a week of moderate activity and less than 75 minutes of intense activity. Compared with the more active groups, their mortality was 30% higher overall. But above that level of exertion, it doesn’t seem to matter how active you are or how you apportion it. “Weekend warriors” did just as well as people who spread their activity through the week.

Ann Intern Med  10 Jan 2017 

Pay for performance
Here’s a systematic review of “the effects of pay-for-performance programs targeted at the physician, group, managerial, or institutional level on process-of-care and patient outcomes in ambulatory and inpatient settings.” That’s a fair spread. We’re not talking about a single thing here, like the UK Quality and Outcomes Framework and how that developed over time. I’ve said a fair bit about that over the years, and to me it’s shown that you really can’t discuss such things without close attention to context and system dynamics. It started with good intentions, but over time it became the worst of perverse incentives. I’m quite prepared to believe that some incentive systems sometimes help in particular contexts – those Medicare hospital readmission penalties the Yale team analyzed the other week, for example. In the end, the reviewers admit the limitations of their analysis: lack of rigour in the studies, heterogeneous interventions and populations. Meaning they can only weakly conclude that incentives can sometimes improve process-of-care measures in ambulatory care, but that there is no clear evidence of patient benefit in any setting.

Overdiagnosing breast cancer
When my wife received an invitation for mammography screening recently, the letter ended with a sentence about this being “to reduce deaths from breast cancer.” Well, that’s certainly the aim, but does it actually do that, and if so, at what cost? Peter Gøtzsche was the first to ask these questions in depth and 20 years later he is still at it. Here his team look at trends in breast cancer in Denmark between 1980 and 2010. Estimates of overdiagnosis varied according to screening policies, but overall there was an increase in breast cancer diagnoses of 48% over this period, but no change in the incidence of advanced breast cancer.

How to do a BP lowering trial
How do you do a blood pressure lowering trial? Come on, most of you have done it. You measure someone’s BP. It is high. You ask them to get some more readings done by the practice nurse. You may give them a home monitor, or they may buy their own. You put them on one drug. They come back and it’s still not as low as you decide it should be. You give them a second drug. And so on. But most randomised controlled trials of BP lowering are not quite like that. Some still randomise patients to a single new agent versus placebo. Some randomise to a particular level of BP, comparing “intensive” with “less intensive” treat-to-target regimens. Within this second group there are large heterogenities of effect, meaning that very similar trials can sometimes come up with opposite results. Here’s a paper that deals mostly with how this can be addressed by better design and analytic methods. It’s a good article, and required reading for those in the field. But it’s not a lot of help to Dr. Coalface and his patient who should really be talking together about the next possible stage of long-term treatment. After all, there is no such disease as “hypertension,” just a mechanistic risk factor which can be looked at from a population standpoint, as in trials, or an individual standpoint, as we do in our everyday practice. Looking at elevated BP from the individual’s point of view, we should start with an overall assessment of cardiovascular risk and what could be done about each element of it. Then we should be able to present some information about the known benefits and harms of each BP lowering agent, in terms of actual event reduction. Then we should leave the decision with the individual. This may even happen one day.

Lancet 14 Jan 2017 Vol 389

Washing blood products from brain ventricles
The management of intraventricular brain haemorrhage is happily a niche topic. Patients who suffer this kind of bleed have 50% risk of death or severe disability. The temptation to do something is great, and in this study there was no non-intervention group. The 500 patients had extraventricular drains which were irrigated either with saline or with a solution of alteplase every 8 hours for 4 days. Mortality was lower in the alteplase group, but functional outcomes were slightly worse.

Stressing the amygdala and the heart
Reading a book on China the other day, I learnt that the word for “being busy” is mang and its pictogram consists of a heart, and beside it death. This is a very simple character and may therefore be at least 4,000 years old, meaning that the connection between stress and cardiac death is nothing modern at all. Sages at all times in human history have concluded that being too busy is bad for the heart, while being insufficiently busy is just as bad. They have drawn charts of energy, developed systems of humours, invoked the power of the stars and planets, and of late they have brought neuroendocrinology to bear on this important subject. Even the bone marrow gets a look in: “In this first study to link regional brain activity to subsequent cardiovascular disease, amygdalar activity independently and robustly predicted cardiovascular disease events. Amygdalar activity is involved partly via a path that includes increased bone-marrow activity and arterial inflammation. These findings provide novel insights into the mechanism through which emotional stressors can lead to cardiovascular disease in human beings.”

The BMJ 14 Jan 2017 Vol 356

Screening the end bit of the bowel
When screening for colonic cancer using flexible sigmoidoscopy was first proposed, someone commented that it was like screening for lung cancer by taking an x-ray of the left lung. But oddly enough, screening by full colonoscopy hasn’t shown a massive advantage – if you want a short account of all the bowel screening alternatives, there’s one in this week’s NEJM. This BMJ article is a pooled analysis of all the randomised trials of screening by flexible sigmoidoscopy alone. Considering that sigmoidoscopy examines only about 50 cm of the distal colon, it’s generally a surprisingly effective form of screening in men and younger women. In women over 60, however, it misses a lot of cancers, because they tend to be in the more proximal colon.

Migraine and risk of perioperative stroke
As I mentioned, it’s over fifty years since I first started getting migraines, and I still get them in an attenuated form of frequent auras sometimes followed by headache. According to most studies, this puts me at a higher general risk of stroke. What if I need major surgery? Oh dear, oh dear: I’m more than twice as likely to get a perioperative stroke. The absolute risk, according to a study from the Massachusetts General Hospital, is 6.3 per thousand for migraine with aura, as compared with 3.9 for migraine without aura and 2.4 for people without a history of migraine. “Migraine should be considered in the risk assessment for perioperative ischemic stroke.” the authors say. To what end? Maybe a trial of perioperative aspirin in migraineurs undergoing surgery may be needed. It would have to be very large, but then there are a lot of us, and millions of operations are done every day. APSOMS: Aspirin to Prevent Strokes in Operations on Migraine Sufferers. 50K participants, 1,000 centres, instant randomisation, costless intervention, simple outcome measures of stroke and bleeding. In the twenty-first century, this thing should be doable and published within 6 months.

Plant of the Week: Daphne x transatlantica “Eternal Fragrance”

The correct name for this hybrid daphne is apparently “Blafra,” but you can see why garden centres prefer to call it by its showier name. I don’t know how eternal it will prove to be, but it is certainly fragrant, and it’s inclined to flower at all sorts of times. Ours is having a go now, but not with any great conviction.

Normally I would be singing the praises of the Daphne bholua at this time of year. The best daphne of all, it flowers from early December to late March, and its fragrance is among the loveliest in the plant world. But the small grafted plant we managed to find last summer is still not flowering. Once you could buy the variety “Jacqueline Postill” (still the best) in lots of garden centres, but I’m told that the recipe for micropropagating this difficult plant was passed on by its discoverer and then lost. So a commerce of tens of thousands of plants has fallen to a few hundred. I am sure some Dutchman will soon put this right, but in the meantime buy “Eternal Fragrance” instead. It doesn’t guarantee the wintertime bliss of bholua but it does make a nice healthy dome of evergreen foliage, and when it is in flower, it is definitely worth kneeling to worship its eternal fragrance.

  • kidmugsy

    “The general drift seems to be that although levels of “high” systolic BP (beginning at 110!) are rising in most countries, the mortality associated with them is tending to decrease.” The key word there is “associated”. What can it mean? How does one associate mortality rates with high BP? Linear correlations inferred from epidemiological data? RCTs? Frequently repeated assertions? Confusion of the risks of very high BP with risks at somewhat lower BP?

    From time to time I try to understand what evidence lies behind BPophobia but I don’t seem to find much success. Any help would be gratefully received.

  • George

    I notice in the discussion of Mang and CHD that emotional stressors are something a little different from being busy. Indeed, busy people who valued business have often lived long lives being busy, sometimes under great pressure, but without resentment – one thinks of GB Shaw (94) or Colin Wilson (84). The emotional stress comes perhaps not from being busy per se, but resenting being kept busy – or, indeed, resenting being kept idle, seeing that inactivity or sedentariness is very much linked to heart attacks, yet monks who seek it out from love of quietude may not court added risk.
    We might say that this link is a matter of adjustment, or integration, which takes us back to old-fashioned ideas about chronic disease, such as those of Michael Balint, with which you may well be more familiar than I, as he seems to be credited with having coined the phrase “patient centred care”.

  • Mark Aley

    re migraine: the only preventers they used were amitrip and topiramate, so no BB’s, CCB’s, etc