Analysis

Post-marketing observational trials and catastrophic health expenditure

BMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e3987 (Published 12 June 2012) Cite this as: BMJ 2012;344:e3987
  1. John S Yudkin, emeritus professor of medicine
  1. 1Department of Medicine, University College London, London WC1E 6BT, UK
  1. j.yudkin{at}ucl.ac.uk

Over the past decade, one of the world’s largest insulin manufacturers, Novo Nordisk, has worked hard on its social responsibility profile. In 2001, the company launched an equity pricing scheme for its insulins in 50 of the world’s poorest countries.1 This might have been triggered by the unease felt across Denmark at the company’s participation, with 38 other pharmaceutical companies, in an infamous court case in South Africa attempting to use trade agreements to block the government’s use of generic antiretroviral drugs.2 Nevertheless, by 2008, in the Access to Medicines Index, a ranking of the world’s largest drug companies on their efforts to increase access to medicine, Novo Nordisk came second, with an aggregate score of 3.9 out of 5.3

In 2010, however, Novo Nordisk invoked negative headlines by threatening to withdraw all its products from Greece because of a governmental order to cut all drug prices by 25%.4 In the UK, Novo Nordisk has faced protests about its withdrawal of its Mixtard insulin formulation, still used by 90 000 people.5 In the 2010 update of the Access to Medicines Index, the company’s score dropped to 2.1 and its ranking to eighth.6

The company has been working hard to recapture its market share, exploiting “emerging markets” with a rapidly rising prevalence of diabetes. There are suspicions that some of the company’s recent observational studies are predominantly serving a marketing purpose.7 Such studies can play a role in defining a drug’s efficacy and safety: Novo Nordisk’s stated aim for its PREDICTIVE study was “to evaluate the incidence of adverse events while using Levemir under normal clinical practice conditions.”8 9 The company paid doctors in 26 countries to enrol 47 565 people with diabetes for initiation on insulin detemir, Novo Nordisk’s long acting insulin analogue. A PubMed search for “PREDICTIVE” with “Levemir” or “detemir” shows that the study has generated some publications, all from the 11 European countries in the study. The other 15 countries, which contributed 27 914 study participants, have not yet generated a single publication.

The payment to participating doctors was for work involved in converting someone from tablets or human insulin to a new drug regimen. After the six month study, most patients would probably remain on the new analogue rather than going through another, uncompensated, change of drug regimen. The sponsor did not provide the study insulin; instead its cost fell to the health system, insurance company, or patient. The PREDICTIVE study did not recruit doctors or subjects in the United States.8 9 The latest observational study on insulin analogues launched this year by Novo Nordisk, A1chieve, recruited 66 726 subjects in 28 countries, none of them in Europe or North America.10 11 12

The 26 countries in the PREDICTIVE study ranged in wealth from Luxembourg (gross domestic product (GDP) per capita $105 197) to India (GDP per capita $1410 (£917; €1130)), where 3435 patients were recruited.13 The retail price for insulin detemir in India (100 U/mL) corresponds to $17.77 for a 3 mL insulin pen. For a patient using 30 U insulin daily, the annual cost of insulin would be $648.60, compared with $74.32 for a generic human insulin (among the lowest prices in the world, thanks to India’s thriving generic drugs industry). What might justify an eightfold price difference? The recommendations of the National Institute for Health and Clinical Excellence (NICE) are that it is more cost effective to target the use of the long-acting analogues to people with type 2 diabetes whose lifestyle is significantly restricted by symptomatic hypoglycaemia.14 Human insulins should be the initial treatment of choice, particularly when resources are limited.15

In the UK and most other high income countries, the state or health insurance company pays the cost of treatment. In other situations, particularly in low and middle income countries, patients must cover drug costs. The impact of a diagnosis of diabetes on household spending can be dramatic: in India in 2005, median spending on diabetes care was $227 in urban and $142 in rural areas, comprising, respectively, 10% and 17% of household income.16 Among the poorest quarter of individuals, this proportion was between 25% and 33%.

The term “catastrophic health expenditure” describes the potential impact of ill health, often an acute event, on the ability to pay for essential household needs.17 In Asia the diagnosis of diabetes increases the risk of catastrophic health expenditure from 16% to 23%.18 Median household income in India is around $3000 a year,19 so converting from human insulin to insulin detemir would increase a person’s expenses by around 20% of this sum. If the aims of PREDICTIVE and A1chieve are predominantly those of marketing, then Novo Nordisk’s strategy may be driving households into catastrophic health expenditure.

Novo Nordisk can be justifiably proud of some of its social responsibility initiatives over the past 10 years, but it now needs to reclaim the moral high ground. It is hoped that, when the next Access to Medicines Index is published this year, the company will again be among the league leaders.

Notes

Cite this as: BMJ 2012;344:e3987

Footnotes

  • doi:10.1136/bmj.e3974
  • , doi:10.1136/bmj.e3990
  • Competing interests: All authors have completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years, no other relationships or activities that could appear to have influenced the submitted work.

References