Lundbeck, based in Denmark, put escitalopram (brand name Cipralex) on the UK market last June. Escitalopram is made by splitting the active molecule in Lundbeck's best-seller citalopram (Cipramil).

Such practice is increasingly common in the fiercely competitive industry. It enables a company whose drug is about to go off-patent and will be copied and sold cheaply by generics companies, to secure a 20-year monopoly on what is marketed as a new drug, but is merely the active component of the old one.

Lundbeck promoted the new drug as more effective. The Prescription Medicines Code of Practice Authority, a self-regulatory body, has ruled that the claim breaches the industry's code. Lundbeck must now change its advertising in medical journals.

Eight complaints were brought against the company by David Pyle, a psychiatrist in Wales. He objects to NHS doctors being urged to prescribe an expensive new antidepressant when cheap generics of the old one are available. The authority upheld five of Dr Pyle's complaints.

Lundbeck was found in breach for claiming that "Cipralex is significantly more effective than Cipramil in treating depression". Lundbeck appealed, but lost.

Its claim was based on three studies comparing the two drugs. Lundbeck argued that although each study does not show statistical improvement, when the three are pooled together there is evidence that the new drug works faster. This meta-analysis was carried out by Jack Gorman of Columbia University, New York, who declared $5,000 (£3,500) of consulting fees from Lundbeck in the American Journal of Psychiatry in January last year.