Intended for healthcare professionals

Views & Reviews From the Frontline

Bad medicine: the rise of duloxetine

BMJ 2014; 348 doi: (Published 17 January 2014) Cite this as: BMJ 2014;348:g139
  1. Des Spence, general practitioner, Glasgow
  1. destwo{at}

I like to understand how things work. Complexity makes me wary because it often conceals a lack of understanding. We are told that medicine is incredibly complex yet paradoxically we offer the crude simplicity of drug interventions. Some drugs are used in multiple conditions, and this so called indication creep broadens pharmaceutical companies’ potential for profit—certainly good business if not good medicine.

Duloxetine is such a drug and the number four top seller, with projected sales of $4bn (£2.4bn; €2.9) in the United States this year.1 It is licensed for depression, anxiety, fibromyalgia, neuropathic pain, and incontinence. Its maker, Lilly, is actively seeking a licence for arthritis and back pain, opening up a potential market of 45 million Americans.2 3 In the United Kingdom duloxetine sales have increased about fourfold, to some £28m, in just five years.4

Duloxetine prescribing is endorsed by a Cochrane review,5 but this psychoactive drug is used in conditions that are largely medically unexplained, without any strong intellectual or biological basis. So is this story, of a humble antidepressant turned superstar blockbuster, a triumph of medicine or marketing?

Chronic pain, the most common medically unexplained symptom, with wide cultural variation unexplained by genetics, is a key market.6 The drug industry is intent on medicalising all reported pain, evoking the emotional rhetoric that pain management is a “human right.”7 Pain research uses myriad soft rating scores that patients report themselves. These are converted into numbers to enable the production of statistically significant outcomes—even if they are not clinically detectable or relevant.

The research is dense and complex. Predictably, the placebo response is high and sustained in pain research, and the actual benefits of duloxetine over placebo are of the order of 1 point on a 10 point rating score.8 9 10 11 12 13 Research over-represents women, has high drop-out rates, and lacks intention to treat analyses. Worse still, follow-up has been a mere 3 to 6 months, which is unacceptable for a drug that will potentially be taken indefinitely.

Predictably, the expert academic guardians are steeped in conflicts of interest.14 Where are the hard objective data of long term benefit? And should duloxetine be taken with other psychoactive drugs, for example, opioids and gabapentinoids?

Duloxetine is associated with considerable discontinuation syndrome. Stories choke the internet: “I went to the doctor [and] tried to explain my symptoms and he thought I was crazy.”15 Once patients start duloxetine it is not so easy to stop. Also the opportunity cost is huge, with other Cochrane reviews finding effective non-drug alternatives.16 17 18 Doctors have typical therapeutic myopia: the rapid rise of any psychoactive drug for multiple medically unexplained symptoms surely looks like bad medicine.


Cite this as: BMJ 2014;348:g139


  • Competing interests: I have read and understood the BMJ Group policy on declaration of interests and have no relevant interests to declare.

  • Provenance and peer review: Commissioned; not externally peer reviewed.

  • Follow Des Spence on Twitter @des_spence1


View Abstract