Financial history is full of bubbles, driven by “our innate inclination to veer from euphoria to despondency.” As I read an account of how bubbles happen in Niall Ferguson’s excellent book The Ascent of Money, I was reminded of how medicines are also prone to bubbles. Indeed, I thought of other human bubbles and decided that we are prone to bubbles throughout our lives.

The first—or at least most famous—bubble was the tulip bubble of the 17th century. Tulips first arrived in Europe from the Ottoman Empire at the end of the 16th century and became symbols of wealth and luxury because of the intensity of their petals, an intensity much greater than any other flower known in Europe at that time. The Dutch threw off Spanish rule in the first half of the 17th century and proved themselves great businessmen, inventing many new financial instruments. One of their inventions was a futures market that allowed goods, including tulip bulbs, to be bought and sold without being delivered.

The scene was thus set for the tulip bubble. Had you been around in 1636 and had the money you surely would have wanted to buy tulip bulbs, which were clearly increasing dramatically in value. French speculators entered the market in 1636, and the bubble seems to have begun in earnest in November 1636. By the end of January 1637 40 tulip bulbs cost around 100 000 florins, enough to buy 1000 tons of butter or to pay 600 skilled labourers for a year. It’s hard to relate 17th century money to present money, but 100 000 florins is about one million Euros, so one tulip bulb cost about 25 000 Euros. The collapse began on 3 February 1637, and by May tulips had lost more than 90% of their value.

Since that bubble we have had hundreds of  bubbles, and most readers of the BMJ will have lived through the dotcom and housing bubbles. Indeed, Ferguson ends his book by suggests that those of us in Europe and the United States may be living in a gigantic bubble that may be about to collapse.

The first requirement for a bubble, writes Ferguson, is “displacement,” a change in economic circumstances that creates profitable opportunities. The equivalent for medicines might be nothing more than the arrival on the market of a new drug—but particularly one that seems to have a new action and work on a common health problem.

Then comes “euphoria” or overtrading, a feedback process where the promise of big profits drives up prices. With medicines this is the phase of  “the best drug ever” that I learnt about at medical school. Next comes “mania,” when naive investors are prey to swindlers. This is the phase when ordinary prescribers, perhaps driven by intense advertising, reprints from medical journals, and pressure from patients, begin to prescribe the drug in large numbers.

Eventually, often quickly, we come to “distress” when insiders begin to realise that the reality does not justify the exorbitant price and begin to sell. This stage arrives with medicines when reports of side effects begin to appear. The final stage is “revulsion” when prices fall and investors “stampede for the exit.” In medicine this is the point when the drug is seen as “the worst ever” and may be banned by regulators. At medical school I was taught that if the drug wasn’t banned then it would eventually reach its rightful place in the pharmacopoeia, a place well below that reached in the mania stage.

Benoxaprofen and rofecoxib (Vioxx) are the two drugs that I most associate with bubbles, and I had walk-on parts in the drama of benoxaprofen. It was a non-steroidal anti-inflammatory drug developed in Britain by Lilly and launched in the UK in 1980. Around that time I won the Medical Journalist Association’s prize for young journalists, which was funded by Lilly and had to be spent on travel to undertake research. I was interested in compensation for medical injury, and as Lilly had been heavily involved in compensating women damaged by diethylstilboestrol (DES) it made sense to visit their headquarters in Indianapolis. I was put up in luxury far greater than I had ever experienced before, and part of my seduction was to be shown  promotional films for benoxaprofen. These I thought wildly over the top as they suggested that the drug didn’t just relieve symptoms but reversed the disease.

Soon afterwards the drug was launched onto the market in Britain and described by the Liverpool Echo as a “miracle drug.” Prescriptions soared, and probably because so many patients were put on the drug so quickly side effects soon became apparent. In May 1982 the BMJ published a cluster of papers describing side effects of benoxaprofen, including one that suggested that the drug had killed 12 people though kidney and liver failure. (I remember a very tense meeting with representatives of Lilly where Tony Smith, deputy editor of the BMJ, said to them “if that’s your attitude we’ll see you in court.”)  In August the drug was suspended in Britain, and the Committee on Safety of Medicines reported that it had had 3500 reports of adverse effects, including 61 deaths. The bubble had burst, and Lilly withdrew the drug from the market. In retrospect the drug probably had no more serious side effects than other non-steroidal anti-inflammatory drugs—and was killed  by its bubble. In this case the bubble probably resulted from Lilly overmarketing the drug. “Those who live by the word shall die by the sword,” said Stephen Lock, the editor of the BMJ.

Vioxx suffered a very similar fate, but the other human bubble that occurred to me as I read Ferguson’s book was the less dramatic but recurrent bubbles that we created at the BMJ when trying to accept the right number of research articles to be published. In those days we had about 100 submitted a week and had room for about six in the days when we were restricted by paper. If we accepted too many then the time to publication became too long, but if we accepted too few we would have nothing to publish. We used to swing backwards and forwards from one possible fate to the other, compensating by increasing the number of pages if we had too many or rushing papers through the system if we had too few.

Why did this cycle happen, we wondered. We had a consistent number of papers submitted each week, and the quality of the papers had a bell curve distribution with no great variation from week to week. We concluded that we were introducing the instability ourselves through worrying about having too many or too few papers. We decided to stop worrying, and the cycle disappeared.

We were exhibiting what George Soros, the enormously successful financier, calls reflexivity. “There is,” says Soros,  “a two-way reflexive connection between perception and reality which can give rise to initially self reinforcing but eventually self defeating, boom-bust processes, or bubbles. Every bubble consists of a trend and a misconception that interact in a reflexive manner.” Soros believes in the irrationality of people, whereas the “geniuses” who created credit default swaps and other complicated financial instruments designed to eliminate risk were sure that rationality and advanced mathematics would triumph. Soros has made billions while the rationalists lost their shirts. My advice is to always bet on irrationality, even in this age of evidence based medicine.

Richard Smith was the editor of the BMJ until 2004 and is director of the United Health Group’s chronic disease initiative.