“Sorry, I Told You So”

I was brought up never to say, “I told you so.” My parents and teachers drummed it into both my id and ego that being right offered its own satisfactions. As such, there was no need to show off. I do think I have internalized this injunction reasonably well and held off on trumpeting my wisdom to the whole world—at least until now. And it may well not have been my acquired modesty that has always kept me so quiet. Perhaps I did not have too many opportunities to be any smarter than the next guy. (Speaking of modesty, it can apparently be taken too far. As Churchill famously quipped about one of his parliamentary rivals, Clement Atlee, “He is a very modest little man who has a good deal to be modest about.”¹) But sometimes a man’s gotta do what a man’s gotta do. In this case, I have recently been proven right about something, and some of my friends and colleagues must be a little surprised—especially given their firmly held opinion of two of my more modest qualities: IQ and my clinical skills. It turns out that, as I had predicted, the anti-dementia drugs such as donepezil may not be all they’ve been cracked up to be. In a recent beautifully designed and executed study, the AD2000 Collaborative Group published its findings in The Lancet.² I strongly suggest that for further details the interested reader read the article as well as the informative accompanying editorial.³ Essentially, the piece supports previous research that donepezil does indeed offer some very modest improvements in patients with Alzheimer’s disease (AD). However, reminding us of Clement Atlee, these changes are so modest that no one, certainly not the patient, nor the family, nor the treating physician, when blinded to which agent (active or placebo) the patient is taking, can usually see any difference between the two groups. To make matters worse, the drug is not cost-effective, does not save anyone—either family or the state—any money, and neither reduces nor delays long-term institutionalization. These are claims we have long heard from a disinterested party: the drug companies that produce the cholinesterase inhibitors. Out for my daily walk the other evening, I started thinking how I could have gotten it so right. In any case, how did I even know I really had been skeptical and was not simply rewriting history? Well, it turns out that I actually have my opposition to it in writing in the archives of Israel’s Ministry of Health. For much of the 1990s, I was head of its Division of Geriatrics, and among my many duties was to give the Minister my best professional advice. One fine day in 1997 I received a memo asking me to prepare a position paper on these new anti-dementia drugs. Did they work? How much did they cost? What proportion of the elderly population suffered from Alzheimer’s disease, and how many of these would likely benefit? Above all, should they be included in the national formulary at state expense? Fortunately, for posterity, but unfortunately for me at the time, I had just finished reading a wonderful book on the subject of how to provide evidence-based health care.⁴ I say “unfortunately” because the Minister, a real populist, expected only positive answers. As one can gather from the title of this little primer, it applies the principles of evidence-based medicine to the art of policy-making in a most clear and elegant manner. I did then what I thought I was supposed to do: draw up and sign a scientific report that reviewed the evidence, pro and con, offering the honorable MP my most informed recommendation. But how did I come up with such an unwelcome negative one? First, my reading of the literature in the mid-1990s convinced me that while the anti-dementia drugs did offer statistically significant differences, these were hardly clinically significant. What, after all, is an “improvement” of 1-2 points on a 30-point Mini-Mental State Examination (MMSE)? As well, I have long been generally skeptical of the usually overly optimistic claims of drug companies. In this case they really poured a lot of dough into the bread of the argument. I also saw a lot of my colleagues being bought by these same firms (in some cases consciously, in some more naively). For example, they would be invited to “speak” on the subject, supported by a generous honorarium in various exotic venues. As well, the companies would try to caress our egos not just via payment but by calling us “opinion leaders,” either local or national. In addition, the folks in the Alzheimer’s societies around the world weren’t much help. Justifiably desperate for a cure or at least a palliative, these groups, the members of which otherwise involve themselves with all kinds of good works, did not want to be bothered with my kind of medical skepticism. To make matters worse, many of the doctors who advised them just fed the families false hopes. Why listen to me when they had a whole chorus of prestigious “opinion leaders” on the other side? Most of those “positive” papers published before the Lancet study suffered badly from something that has long bothered me about research in the elderly: the mismatch between the clinical “cleanliness” of those patients chosen to participate in the drug trials and those whom I look after.⁵ For good scientific reasons, in an efficacy trial (Does the drug do more good than harm in those who take it?) one needs subjects with the disease in question, AD, but with little other comorbidity. These people may be old, they may score 20 over 30 on an MMSE, but heaven forbid they should be illiterate, poor, hypertensive, depressed, diabetic, or noncompliant like so many of my patients. As well, they must have loving, supportive caregivers, which are not always, in my experience, a dime a dozen. In theory, positive efficacy trials are meant to be followed up with effectiveness studies (Does the drug do more harm than good in those to whom it is offered?) Now that’s a different story and no surprise that no one had yet sponsored an effectiveness trial with these drugs. That is, until the publicly funded Lancet study, which has come very close. And what do we read in the paper’s discussion? It turns out that, while the experiment tried to conscript approximately 3000 subjects, the researchers were only able to recruit just under 600 souls, in part because of the drug companies that tried to scotch the experiment. As well, the trial had a delayed start “because of difficulties in securing placebo and drug supplies from the relevant pharmaceutical companies who opposed the trial.” And the study organizers had actually asked to buy the drugs from the company, not to receive them as a donation. (By the way, I realized that my medical skepticism is not always justified. For example, for many years, I thought that the move to encourage the prescription of statins was a plot cooked up by the ice cream companies to enable more of their product to be sold. It turns out, however, that with this issue I was back in my old role of being wrong. And in expiation, I am myself taking a low dose of atorvastatin. At least in my case, it really does seem to work.) But back to donepezil. I may have been wrong about beta-blockers, mini-dose aspirin, ACE inhibitors, coumadin, and statins, but at least I know that for once in my clinical life I can crow out “I told you so.” And as the Lancet article so elegantly demonstrates, I was this time (unfortunately for the patients) right.