I cannot agree with Dr Spence’ evaluation on pharmacoeconomic studies because he misunderstands some important pharmacoeconomic methodologies.1 First, the National Institute for Health and Clinical Excellence (NICE) guidance appraised the cost-effectiveness analysis (CEA)2 rather than cost-benefit analysis (CBA).1 Though the CBA is well known since the 1970’s, the beginning of economic evaluation of public health programmes, they are different types of pharmacoeconomic studies. Second, pharmacoeconomics, including CBA, CEA and cost-utility analysis (measured by quality-adjusted life year, QALY, usually regarded as a subtype of CEA), is not a pseudoscience. Fraught with assumptions in pharmacoeconomic study is resulting from the uncertainties everywhere in the real-world clinical practice for various study subpopulations and their diversified responses to treatment alternatives (no treatment also an alternative).

The analysis in his first paragraph is a market forecasting with biases. The “astounding” monetary value, £4bn in the UK for 10 years, is showy and actually a pseudoscience. Why? As he states in the subsequent text, “In the real world warfarin clinics will still run,” no one drug will take 100% of the market share on which his extrapolation was based. Warfarin’s average monthly cost of £1 is not realistic either without taking blood monitoring cost into account. It is also inappropriate to compare a monthly cost per person to a 10-year cost of all patients, which would be definitely avoided in a good pharmacoeconomic report. And in fact the £4bn estimation for 10 years is incorrect without discounting for inflation. Please think about the following statement, “When the NHS was launched in 1948 it had a budget of £437 million (roughly £9 billion at today’s value).”3 His analysis lacks a thorough consideration of uncertainty and needs sensitivity analysis. Moreover, this budget impact would be primarily the NHS’ concerns. While the NICE works closely with the NHS, I believe a budget impact analysis has been conducted and the threshold has been accepted.

Admittedly, pharmacoeconomics still has limitations and thus is controversial, just as Dr Spence is concerned about.1 The problem is that we should evaluate a pharmacoeconomic study correctly and comprehensively while critically. After all, minimising bias is the groundwork of any quality research.

1. Spence D. Beware the cost-benefit analysis. BMJ 2012;345:e5040.
2. NICE. Rivaroxaban for the prevention of stroke and systemic embolism in people with atrial fibrillation. 2012. www.nice.org.uk/nicemedia/live/13746/59295/59295.pdf.
3. NHS. About the NHS. 2011. http://www.nhs.uk/NHSEngland/thenhs/about/Pages/overview.aspx.

Competing interests: None declared