Thursday, September 9, 2010

New Review of Ghostwriting for Selling HRT

My e-mail has been jumping with people praising the new article in PLoS Medicine by PharmedOut leader Adriane J. Fugh-Berman at Georgetown:

My own reading of the paper has a few qualifications, however.

Generally the paper is based, yet again, on secret drug company documents released through litigation--in this case, people who developed breast cancer while on Wyeth's Prempro and later sued. Dr. Fugh-Berman notes that while Prempro for menopausal symptoms and/or prevention was touted at the time as "hormone replacement therapy" or HRT, that presumes that menopause is a disease of hormone deficiency, which is itself a disputed model of menopause that favors drug treatment; so a better label would be HT.

The focus of the study is the role of DesignWrite, a medical education and communications company (MECC), in managing a portfolio of articles to be placed in medical journals touting the virtues of Prempro, including trying to extend its use to off-label indications such as preventing Alzheimer's and Parkinson's disease and macular degeneration, and producing younger looking skin. Fugh-Berman was able to track about 50 papers that DesignWrite produced that were eventually published. I wish she would have addressed the question, as David Healy did in a similar article about psychiatry and ghostwriting, as to just what proportion of the total literature this represents. I did a quick PubMed search on English language articles on conjugated estrogen-progestin therapy published between 1997 and 2003, Fugh-Berman's target period, and came up with 253 cites; which suggests that as many as 1 out of 5 papers on the topic were ghostwritten by marketers.

The procedure was that DesignWrite's writer produced a draft and then Wyeth reviewed and kibbitzed about the draft. Only then was the draft sent to an academic physician who was asked to appear as the supposed author. Fugh-Berman did not discover among her documents what was paid to the academic physicians; for major papers DesignWrite received $25,000. She gives examples of academic docs meekly signing off on "their" articles, but the occasional old-fashioned physician actually taking the role of author seriously and insisting on editing and changing the text--which DesignWrite, to mollify the doc, tried to permit so long as Wyeth's main marketing message was retained. (One recalcitrant academic even tried to acknowledge the role of DesignWrite's hired ghostwriter in print, but that effort was firmly squelched.)

It's certainly valuable to have this information in the public record, and to know that "HRT" was yet another area manipulated by ghostwriting. My qualifications come from having followed this discussion fairly closely at the time, and conferring with some of my evidence-based colleagues on what advice to give to our patients when the RCT's were announced showing that "HRT" increased both cardiovascular and breast cancer risks. My worry is that we are engaged in some revisionist history right now on the "HRT" question. A few points to recall:
  • Fugh-Berman mentions only in passing that one clear-cut benefit of HT, which I don't believe has ever been denied, is prevention of osteoporosis. As we proceed today to vilify HT, with the benefit of 20-20 hindsight, we tend to forget the plight of the 80-year-old woman with a broken hip, who was in danger of dying or being stuck for life in a nursing home as a result, and the strong desire to prevent such an occurrence if possible.
  • Also minimized today is the reality of severe hot flashes and the fact that many patients, when we tried to stop their HT, came back complaining bitterly of the lowered quality of life, and in some cases even frank incapacitation this caused.
  • Then, consider the actual "number needed to harm" as revealed, not in industry-sponsored commentaries, but in the raw numbers contained in the supposedly reliable RCTs that first revealed the true risks. As best as I can recall, for conditions like breast cancer and other cancers, the number of patients who would have to take HT for a year for 1 to end up with a cancer was in the range of 500 to 1200. That is a real risk, but it's within a numerical range that a woman with severe hot flashes, or at high risk for osteoporosis, might rationally decide to continue HT even knowing the risk.
In hindsight, we know now that HT was oversold as a preventive for cardiovascular disease, and this seems mainly the result of relying on observational studies before RCT results were available. The RCTs clearly showed increased, not decreased cardiac risk. It's good to now know that this misimpression was not completely accidental but was aided and abetted by ghostwritten marketing articles. The complete picture is a bit more nuanced than this one article conveys, IMHO.

No comments: