health economics
Dr Paula Lorgelly (left) and Dr Charlotte Davies (right).
The purchasing behaviour of the UK’s state-funded National Health Service (NHS) has been attracting considerable attention of late with several key reports, including those by British billionaire businessman Sir Philip Green in 2010, the National Audit Office (NAO) in 2011 and financial services company Ernst & Young in 2012 , identifying wide variation in the purchase price paid by NHS hospitals for consumables (ranging from bed sheets to MRI diagnostic equipment).
More specifically, the reports by the NAO and Ernst & Young have also identified that purchasing decisions are taking place at the fragmented, hospital level, somewhat counterintuitive to what one might expect when considering a large buyer such as the NHS.
However, hospital purchasing is only part of the picture. The behaviour of the suppliers and the nature of their relationship with the NHS is also relevant and the fragmentation in purchasing within the NHS. This raises the possibility that the NHS could be leaving itself open to so-called “seller power”, whereby sellers are able to extract anticompetitive prices by cornering a majority of a particular market sector. This is well researched and understood in the pharmaceutical sector. However, medical devices have attracted much less interest until now.
The authors used the prostheses implanted in primary total hip replacement surgery as a case study to investigate whether the supplying industry has the structural features necessary, although not sufficient, for the exercise of market power both at the national and hospital level. Further to this, they also focus their analysis on the pattern of purchasing at the individual hospital level in order to determine the level of specialisation hospitals have in their procurement of hip prostheses.
In on-going research the authors are using a large national patient level dataset derived from the National Joint Registry (NJR) for England and Wales and the Hospital Episode Statistics (HES). From these data, a hospital level panel for each year 2004 to 2008 was constructed with the hospital year (of surgery) as the unit of observation. This resulted in a total of 2,281 hospital year observations, for 278,063 patients.
They identified that at the national level, only five manufacturers have a market share of consistently over 5%. However, more startling was the finding that two of the manufacturers consistently account for 69% of the market over each of the years 2004 to 2008. From these results they conclude that the national market is highly concentrated (very oligopolistic), with no significant entry of new firms and with very stable concentration and stable market shares of the leading manufacturers. At the hospital level, they find that procurement is typically very specialised.
The authors' regression analysis of the determinants of hospital procurement concentration reveals that hospitals have become more specialised in their purchases since 2007, when the previous government issued a healthcare spending policy, payment by results, which rewarded hospitals with budgetary resources in level with a range of performance criteria.
Procurement specialisation was also found to be higher in NHS foundation trusts (which have financial autonomy) as opposed to NHS trusts (which are financially managed by local health authorities). All regions were more specialised in their purchases than the UK’s capital city, London, with some indication of increased concentration especially in localities around manufacturer headquarters.
Finally, larger hospitals (in terms of number of procedures) are less concentrated in their purchases, albeit only slightly so.
They confirm the findings by the NAO and Ernst & Young, that procurement in the NHS is not at all uniform. These results also raise further questions, most notably, why do hospitals have such concentrated purchasing patterns for hip prostheses?
Are hospitals exercising their buyer power by negotiating scale discounts (bulk purchases), or is it the result of targeted marketing or even market sharing, by the manufacturers?
Clearly, it would be beneficial to have information on price, however, this is not available within the dataset used and nor is it readily available within the wider literature.
Future work will include an indepth investigation into the decision making process regarding the procurement and purchase price of hip prostheses, involving both the surgeon and hospital procurement departments.
By Dr Charlotte Davies and Dr Paula Lorgelly.
Dr Charlotte Davies is a health economist at Norwich Medical School at the University of East Anglia, Norwich, UK. Her research interests are the competition, regulation and performance of medical devices. These interests are pursued mainly through the analysis of large individual patient level panel data-sets using econometric techniques. She is a post-doctoral fellow, funded by Arthritis Research UK and has a PhD in health economics from the University of East Anglia and an MSc in health economics from the University of York.
Dr Paula Lorgelly is a health economist at Monash University, Melbourne, Victoria, Australia.