PSSST! Wanna sell a dodgy drug? If you are a fledgling pharmaceuticals marketer in search of new tricks, Ben Goldacre’s book offers a handy guide. This is not, of course, his intention. For readers other than the unethical pharma neophyte, it is an intense and a depressing read — both for the facts outlined but also the limited nuance.

The author lists many ruses by which pharma companies leave us all "exposed to avoidable suffering, pain and death". Anyone familiar with the industry’s tactics will recognise Goldacre’s description of how drugs are cooked up. First, assemble the best ingredients; then give the experimental drug to an unrepresentative group of patients in studies that are too small and too short or too long; then compare with an alternative against which the new product can shine.

If the result is disappointing, vigorously whip up the data; change the agreed criteria for success; analyse the trial patients again to play down side-effects or exaggerate efficacy. Present the most engineered version to regulators; hide negative findings in inaccessible data, or forget to submit them to scrutiny.

Last, market by hiring ghost writers to report highlights in academic journals lubricated by the prospect of drug company reprints, and spinning the findings under bylines of academics for hire. Send copies to physicians, along with gifts.

However, the problem facing investigators is that the drug kitchen is a messy place, and not all the cooks have set out to poison customers. Ideally, new medicines would be tested at the right dose against the best comparator. The results would be analysed by independent third parties and disseminated in digestible form for prescribers. But that would require the huge sums put by investors into private drug development at their own risk to be made good by governments, without the drawbacks that would entail, and for academics working on the science to have fewer biases, including against clinical trial "bureaucracy".

Goldacre is at his best in methodically dissecting poor clinical trials. He stresses the importance of transparency so third parties can pore over the data. His examples include a freedom-of-information request on a drug that was so redacted that nothing was left but the letterhead.

He is less strong in explaining the complex background reality, such as the general constraints and individual slips of regulators and pharma companies’ employees. Nor is he able to assess the net effect of the current system: how many lives have been improved compared with those hurt by the present, if imperfect, system, which has led to extraordinary new treatments in recent years. For all the flaws legitimately and valuably highlighted by Goldacre, "bad pharma" is not entirely evil.

© 2012 The Financial Times Ltd