Trust me,I write about doctors

Hard v soft writing

The highly accomplished English had razor-sharp news values and arguably made the Mail the most formidable of the nationals. But his views on hard/soft writing were misplaced, I believe. He argued that if something was worth writing about, it was worth writing “hard”. There is nothing wrong with bare-knuckle journalism if it involves a black-and-white distinction between, say, a truth and a lie. But medicine and science is full of different shades of grey. Bare-knuckle reporting can knock the stuffing out of critical nuances, encouraging subs to harden up stories, sometimes reducing them to worthless fiction.

P-values

My guide to reporting clinical trials for the Medical Journalists’Association in 2014, started with: “No reporter should be exposed for the first time to the intricacies of p-values and confidence intervals in the pressure cooker atmosphere of the newsroom, or at a lone freelance desk.” I was such a reporter. There has been, in the British journalistic tradition, an understandable emphasis on words and writing, but almost every element of modern life is measured in statistics, from the quality of the air we breathe and the health of the economy to sporting performance. If I were starting out today, I would do a master’s in data journalism or statistics.

Research

Perhaps because I was statistically illiterate, I accepted far too readily reports in highly prestigious medical journals such as The Lancet or the New England Journal of Medicine. There was a naive temptation to believe that it “must be right because it is in The Lancet”, but this was the journal that published the infamous article by Dr Andrew Wakefield suggesting that the MMR vaccine might cause autism.

Was this an isolated example? Alas, no. In 1994, the late Professor Doug Altman, one of the world’s leading medical statisticians, estimated that only one per cent of medical research was flawless. In 2006, Dr Trisha Greenhalgh, author of How to read a paper, claimed that only 10-15 per cent of published scientific research would be of lasting value. In 2009, Sir Iain Chalmers, a leading campaigner for evidence-based medicine, and his Australian colleague Professor Paul Glasziou, estimated that as much as 85 per cent of research investment was wasted.

If I had my time again, I would develop this area as a sub-speciality. What better way to make a name for yourself than by focusing on something reportedly squandering billions of dollars of public and private money?

Sub-specialising has huge potential. Medicine and surgery has become increasingly specialised, with sub-specialities of sub-specialities. Medical journalism should follow suit, with more people sub-specialising in, for example, niche areas such as mental health and cancer.

Peer review

But surely, independent experts peer-review journals like those cited above? Shouldn’t this ensure good quality research? Peer review is far from foolproof. In a randomised control trial, a study with eight areas of weakness was sent to 420 reviewers on the database of The Journal of the American Medical Association (JAMA). Only 10 per cent of the 221 respondents (53 per cent) identified four or more weaknesses; 16 per cent failed to identify any. The mean number of weaknesses commented on was two.

Fraud

Bad peer review encourages fraudulent research. The British Journal of Obstetrics and Gynaecology is highly acclaimed, but its peer reviewers did not spot the spectacular fraud by Malcolm Pearce, of St. George’s Hospital, London. He claimed to have successfully relocated a five-week-old ectopic pregnancy into the womb of a 29-year-old woman. I was one of hundreds of journalists around the world to report the story, even though one of his patients was dead at the time he claimed to have operated on her. How were we to know otherwise?

Was this an isolated example? Alas, no. Fraud and Misconduct in Biomedical Research runs to more than 300 pages. Built up around breathtaking case histories, this title is essential reading for aspiring medical and science journalists. Reading it left me wondering how often my colleagues and I have been duped.

Absolute risk reduction and relative risk reduction

Yes, more about figures and more about duping - this time, the duping of readers by reporters, sometimes deliberately, sometimes unwittingly because of misleading press releases.

Absolute risk reduction (ARR) is used to compare two therapies in terms of actual numbers of people treated. For example, in a migraine drug trial, two out of 100 people taking the drug have a migraine, compared to four out of 100 taking a placebo, giving an ARR of two per cent (two per cent versus four per cent.)

Relative risk reduction (RRR) describes the same difference in effectiveness by comparing proportional differences between groups. Thus, the relative risk reduction level for those taking the drug having a migraine is 50 per cent lower than in the placebo group – again, two per cent versus four per cent.

Quoting a 50 per cent reduction might make good headlines, but quoting RRR without ARR can be highly misleading and is a common criticism of medical reporting. Such stories can exaggerate benefits of treatment and downplay risks.

Percentages involving small numbers can be highly misleading. For example, a university registry formally criticised the medical course for failing 50 per cent of first-year Afro-Caribbean students. There were, however, only two students – one passed and one failed.

Balance

I have written about this topic before (BJR September, 2015: vol. 26; no 3), but it is also an integral part of this piece - and one of the most important things I have learned as a medical journalist. I had been taught the conventional wisdom that balance was the hallmark of good journalism, a fair way of telling the story, a means of being objective and preventing bias.

Balance can certainly work well in a straightforward story in which, for example, a government minister and their shadow are given equal time or space to state opposing views. But trying to balance some medical stories can distort research, generate bogus controversy and result in groundless fear and false hope.

For example, there is no scientific rationale for assuming that homeopathic remedies - which are devoid of pharmacologically-active molecules - can produce clinical effects. Any benefits must be restricted to placebo. Yet if I were writing a piece about homeopathy, convention would demand that I balance the story. A homeopath would probably steadfastly insist that science didn’t know everything and that he had hundreds, if not thousands, of satisfied patients.

Would he or she have a point? Surely homeopathy is harmless? The placebo effect can have significant benefit, can it not? It can, but I still vividly recall the series of pictures I was shown of women with fungating breast cancer by Michael Braum, now professor emeritus of surgery and visiting professor of medical humanities at University College, London. They had all been treated with homeopathy. By the time it became apparent that this was useless, it was too late.

Of course, the right kind of journalistic balance is essential. The late American astronomer and author Carl Sagan called on science journalists to balance their genuine sense of wonder about apparently new developments with old-fashioned scepticism. In this respect, we are at exactly the same stage as we were when I began my accidental foray into this intriguing world.