For let us be clear about one thing: Prozac is a marketing success, not a scientific one. This is relevant to the current debate about rationing, Viagra, the price of drugs and whether the National Health Service as we know it is any longer affordable. Science can bring new drugs to the surgery, but only marketing can make them sing.
Let me explain. For decades, depression, which causes enormous suffering and thousands of deaths (by suicide) every year, has been under-treated. So any development that encourages people to seek treatment must be a good thing. The arrival of the selective serotonin re-uptake inhibitors (SSRIs) in the late Eighties, of which Prozac is the best known, was certainly a good thing.
These were antidepressants without (unpleasant) side effects. That was their unique selling point. They were no more effective at lifting depression than the tricyclics which preceded them, but they did it without causing the dry mouth, constipation and drowsiness that were frequent side effects of the older, dirtier, drugs.
This was crucial to improving the acceptability of anti-depressant drugs. Prozac and its relatives were claimed to be the first "clean" anti-depressant - a mood brightener that did not leave you with a leathery tongue, a sluggish gut and cotton wool between your ears.
Except that they were not the first. That accolade belonged to the "new" tricyclics, such as lofepramine (brand name, Gamanil), developed in the Sixties and Seventies, which were refined, cleaned-up versions of the early tricyclics discovered in the Fifties.
Most of the research evidence on which Prozac has built its extraordinary worldwide success - an estimated 35 million prescriptions have been written worldwide and global sales now bring its manufacturer Eli Lilley $1.7bn dollars a year - is thus based on a little white lie, a sin of omission. It compared Prozac with the old tricyclics over which it has the significant advantages I have outlined. I know of no study which has demonstrated similar advantages for Prozac or the other SSRIs over the new tricyclics such as lofepramine, other than in specific conditions such as obsessive compulsive disorder.
So why have the makers of lofepramine not taken to the roof-tops to challenge Prozac's dominance of the market? Simple. It is out of patent so it belongs to nobody. No single company could expect to benefit from its success.
The significance of this is that Prozac costs the NHS twice as much as lofepramine at pounds 20.77 for a month's supply compared with pounds 9.84.
Prescriptions for Prozac have risen from 365,000 in 1991 to more than 2.9 million in 1997, an eightfold increase. The question which Prozac supporters such as Claire Rayner, chairman of the Patients Association, should be asking is not whether the drug brings worthwhile benefits (it certainly does) but whether the benefits it offers over and above those offered by rival drugs, if any, are worth the price differential of pounds 10 per patient per month - or more than pounds 100m a year.
This puts the row over Viagra, a drug which has undisputed advantages over rival treatments, and spending on which was estimated to rise to between pounds 50m and 100m a year even if every man who wanted it got it on the NHS, in perspective.
It is the kind of issue that the new National Institute for Clinical Excellence will be examining when it starts work on April 1.
But I don't envy Professor Sir Michael Rawlins and his team their job.
Who could have predicted, when Prozac received its licence in 1989, that it would become a household name across the world? Yet without the clout of the marketeers behind it, millions might have been denied the release from their misery that the drug has brought.
The story of Prozac demonstrates that there is more to health care than scientific proof of effectiveness and economic proof of value. Health care is about more than science and cost, but Sir Michael and his team will require the judgment of Solomon to ensure that the NHS is neither broken by excessive demand nor weakened by unreasonable denial.