NEJM 5 Feb 2015 Vol 372
519 Refractory angina seems to be common in cardiac clinics but not in primary care. When all the drugs have failed, and revascularization is not an option, device makers like to get inventive. The latest gizmo is an hour-glass shaped expandable metal object which sits at the portal of the coronary arteries and “creates a focal narrowing and increases pressure in the coronary sinus, thus redistributing blood into ischemic myocardium.” My inclination is to say “Yeah, right,” but these things do seem to work, somehow or other. “The device was also associated with improvement of at least one Canadian Cardiovascular Society angina class in 71% of the patients in the treatment group (37 of 52 patients), as compared with 42% of those in the control group (22 of 52) (P=0.003).” The control was a sham procedure.

528 Seeing yet another paper about magnesium—this time for stroke—I was going to suggest that somebody set up a special journal for medical papers devoted to this noble ion. But looking on Google, I see that I am way too late: “Magnesium Research, the official journal of the international Society for the Development of Research on Magnesium (SDRM), has been the benchmark journal on the use of magnesium in biomedicine for 16 years.”

The FAST-MAG study concluded: “Prehospital initiation of magnesium sulfate therapy was safe and allowed the start of therapy within two hours after the onset of stroke symptoms, but it did not improve disability outcomes at 90 days.” Sadly, another magnesient flop.

JAMA 3 Feb 2015 Vol 313
483 Here’s a paper that will become a classic of the surgical outcomes research literature. It looks at a half a million operations performed in hospitals belonging to the American College of Surgeons National Surgical Quality Improvement Program and finds that 5.7% of patients are readmitted to hospital, almost all with complications that they did not experience in the immediate post-operative period. The commonest reasons are surgical site infection (SSI) and ileus. The authors thoughtfully examine the causes in relation to each class of procedure and propose some solutions, for example the out patient management of SSIs, even when this requires debridement and intravenous antibiotics. And they are keen to point out that hospital readmission may be essential and should not be disincentivized to the detriment of patients. “Readmissions after surgery may not be an appropriate measure for pay-for-performance programs but rather better suited as measure for hospitals to track internally.”

496 So the answer should lie in better audit of surgical outcomes and cycles of improvement, right? At first sight, this notion seems to be questioned by the next two papers. The first looks at surgical outcomes and Medicare expenditure in relation to participation in the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP). “Enrollment in a national quality reporting program was not associated with the improved outcomes or lower Medicare payments among surgical patients. Feedback on outcomes alone may not be sufficient to improve surgical outcomes.” But hang on. The intervention here was not “Feedback on outcomes alone” but a specific program of feedback, the ACS NSQIP. And surgical outcomes did improve—it’s just that they improved at the same rate in all hospitals, whether or not they used the ACS NSQIP. All that this study shows is that the ACS NSQIP is no better than “usual care,” i.e. surgeons learning by other means how to improve their outcomes.

505 This paper does the same thing but looks specifically at inpatient complications and mortality. Again, the intervention being observed is ACS NSQIP, and the “control” group are hospitals who chose not to adopt this particular form of audit and feedback. Again, there was no difference. It may be that surgeons are already very good at improving their outcomes and that this exercise by their would-be leaders is futile, like so many exercises by people who think they know better than their working peers.

469 A high point in my non-career was the opportunity to present some slides about the Boston pioneer of surgical outcomes reporting, Ernest Codman, to a distinguished audience including Donald Berwick. Actually, I think he arrived late and missed my presentation, but never mind. Here is a piece in which he reflects on these supposedly negative studies with the title “Measuring Surgical Outcomes for Improvement: Was Codman Wrong.” Like everything Berwick writes, it is thoughtful and masterly. But Codman was not wrong: it’s just that learning using his end result method has become so much part of surgical practice that continuous improvement is evident everywhere, not just where the writ of the ACS NSQIP holds sway.

Ann Intern Med 3 Feb 2015 Vol 162
167 I had a fond delusion that Italian nursing homes might be nice places to fade out, where at least you were guaranteed a decent diet based on pasta, cheese, and fresh fruit and where pressure ulcers were uncommon. Alas, here is a study that disabuses me of any such notions: the Italian researchers found no difficulty in recruiting 200 malnourished patients receiving institutional care or home care packages who had pressure sores. They randomized them to receive an energy-dense, protein-rich oral formula enriched with arginine, zinc, and antioxidants (400 mL/d) or an equal volume of an isocaloric, isonitrogenous formula for 8 weeks. Unfortunately I can’t get the full paper here in Seville, so I don’t know what these “antioxidants” might be – flavonoids, ascorbic acid, vitamin E, allopurinol? The word should be banned. And, without the full paper to scrutinize, I feel sceptical that the feeding formula enriched with these “OligoElements” really improved the rate of healing. Personally I would prefer to receive traditional Italian invalid foods such as minestrone, zuppa pavese and zabaglione.

175 Speaking of fading out, the outlook is getting worse. Dying (in America at least) is getting to be more rather than less distressing, according to this study of pain intensity and symptom prevalence during the last year of life from 1998 to 2010. It relies on proxy reports, mostly from family members, relating to over 7000 deaths in a representative US population cohort. Reported pain increased slightly but the increase in reported depression and confusion was 26-31%, which is a clear signal that something is going badly wrong. It suggests overmedication with the wrong drugs, and lack of the right kind of social and spiritual support.

JAMA Intern Med Feb 2015
OL “Mechanical ventilation. Internal cardioverter defibrillators. Hemodialysis. Extracorporeal membrane oxygenation. Cardiopulmonary resuscitation. Percutaneous gastrostomy feeding tubes. Multidrug chemotherapy. Dying used to be less complicated when unaccompanied by decisions about high-tech interventions. We forestall mortality but at the cost of increasing end-of-life complexity. Clinicians help patients manage this complexity with open communication about prognosis, patients’ values and goals, and the impact of life-sustaining therapies. Since the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment1 of the mid-1990s, however, we have known that seriously ill patients rarely communicate end-of-life preferences to physicians.”

OL So who should do the communicating? In the commentary I quote from above, James Kirkpatrick suggests that nurses should play a key role, though only 10% have received any training in end-of-life conversations. The study he is referring to is a based on questionnaires sent to nurses and hospital residents and staff physicians. “The following family member related and patient related factors were consistently identified by all three clinician groups as the most important barriers to goals of care discussions: family members’ or patients’ difficulty accepting a poor prognosis, family members’ or patients’ difficulty understanding the limitations and complications of life-sustaining treatments, disagreement among family members about goals of care, and patients’ incapacity to make goals of care decisions.” But who is to take on the final responsibility to see that these barriers are overcome? In the UK, I would suggest that it should always be the GP. And we should not accept the decision of “experts” and multidisciplinary teams as binding, but question any decisions that might cause patients and their families to go through unnecessary distress.

Lancet 7 Feb 2015 Vol 385
OL In the course of my clinical working life, I had as little to do with drug companies as possible, and nothing whatever to do with device manufacturers. It came as a jaw-dropping surprise to me when Medtronic decided to let its full data on a commercial product be analysed by two independent teams. I was at Yale University when it happened and I got some trickle-down money from that project, so I must declare an interest. The other thing that impressed me about Medtronic was that it performed a sham-controlled trial on the Symplicity renal denervation device which it had bought from another company on the back of a massively successful open-label trial in so-called resistant hypertension. The Symplicity-2 trial, using a sham procedure as control, failed to meet its prespecified end-point, and rather than make excuses, Medtronic accepted that it should not be promoted thereafter. Now along comes another open-label trial comparing the Symplicity device with stepped drug treatment, paid for by the French government: and Symplicity wins by 6mm Hg.

Bizarrely, the accompanying editorial in effect criticizes Medtronic for not pushing its own product, pointing out various flaws in the Symplicity-2 trial and alleging “an important question is whether a sham procedure is ethical and practical in patients with resistant hypertension often accompanied by other comorbidities, leaving these patients at high risk of cardiovascular and renal disease progression and potentially adding to treatment costs.” What nonsense. We have no idea what these devices do for long term outcomes and sham-controlled trials are the only ethical way to find out.

OL While my dealings with Medtronic persuaded me that they were genuine in their commitment to honesty and openness, I can’t say that Roche have been very convincing on that score. Their repeated failure to provide the Cochrane investigators with promised data about oseltamivir (Tamiflu) is well documented. Moreover it led Tom Jefferson and his team to conclude that published data are quite unreliable sources for the assessment of many clinical interventions of marginal benefit, and that only full clinical study reports and individual patient data will suffice. Even then, manufacturers have ample scope to redact as they please. So when in the face of the largely negative Cochrane review about the supposed effectiveness of Tamiflu, Roche funds an independent study of its data sets which comes out with a more positive conclusion, what are we to think? Only that these data sets should be available for everybody everywhere to examine and reach truly independent conclusions.

For more on this, see Paul Roblin’s BMJ blog.

BMJ 7 Feb 2015 Vol 350
Great people ride my hobby horses this week. Harlan Krumholz gets to write about data sharing in the wake of a long-overdue Institute of Medicine report suggesting that future clinical trials should make their data available for independent analysis. Well, what about the past trials on which we base so much of current practice? I’m so glad too that Harlan highlights an important point which is too seldom made: “One of the peculiarities of clinical research is the challenge of replicating a study. A tenet of the scientific method is reproducibility. And yet, for clinical trials, reproducibility is often precluded by cost or logistic, ethical, or business reasons. As a result, direct replication by repeating the experiment in a different laboratory is not an option. This constraint makes it even more important that independent scientists can study the raw data.”

On the whole I think the BMJ Confidential series is an awful idea, but when it features a fellow-spirit like James Le Fanu I can’t forbear to quote:

“What single unheralded change has made the most difference in your field in your lifetime?”
“The 2004 General Medical Services contract, which financially remunerated good doctors for practising bad medicine.”

And, for a change, you could almost recognize him from the sketch of his face.

I was once co-opted by the UK National Council for Palliative Care to represent the heart failure interest, but after a year or two I felt that somebody younger and more specialized was needed. It looks as if the cause still needs pleading. Mayur Lakhani is now the chair, and writes a short piece with the title “If I Ruled the NHS: Palliative care for everyone with advanced progressive incurable illness.” Hurry up guys: it’s the fate that awaits most of us.

Plant of the Week: Coussapoa dealbata

Don’t try this one at home, but do go to admire it in Seville, where it towers majestically to 50 metres, putting out aerial roots and spreading its ground roots in great ridges for tourists to sit on. To begin with I mistook this rubber tree for a huge version of Magnolia grandiflora. It has similar grey bark and dark evergreen leaves with brown felty undersides.

Since Seville experiences the odd light frost, this native of South American jungles may be hardy enough to survive in some tiny enclaves by the sea in Cornwall. I don’t have my Bean handy to check.
But – o joy! – I look online and find the following:
“For the past century W J Bean’s Trees and Shrubs Hardy in the British Isles has been the pre-eminent reference to woody plants hardy in Britain and Ireland, usually referred to by gardeners simply as ‘Bean.’ First published in 1914, it was updated regularly until the Eighth Edition, brought out in four volumes during the 1970s. The text from this edition and its Supplement by Desmond Clarke (1988) is being made freely available here by the International Dendrology Society.”

But there is no mention of Coussapoa, even under its old name, Ficus dealbata.