Microsoft word - leask.doc

Evolutionary Change within the UK Pharmaceutical Industry: A Cladist
Approach.
Graham Leask
Aston Business School
Aston University
Draft Paper to be presented at DRUID
January 2002
Abstract
This paper seeks to examine the effect of a decade of rapid change within the UK Healthcare environment upon Pharmaceutical Companies. Between 1990 and 1999 the UK Pharmaceutical Industry had to contend with a series of seismic shocks that radically changed the competitive landscape and introduced marked new pressures on participant companies. Prompted by the Health Service need to curb costs the UK NHS reforms were the most profound since 1948 resulting in a marked shift in industry methods of working and the relative success of their strategies. The evolution of the top 10 pharmaceutical corporations is examined using cladistics a technique from the biological school of classification. A technique based upon studying the evolutionary relationship between companies with reference to the common ancestry of the group. The concept of the corporate genome driving path dependant behaviour is introduced.
Introduction

The UK Pharmaceutical industry received a series of seismic shocks during the last
decade as the government and ultimate monopoly purchaser of pharmaceutical products sought to curb rising drug costs. Rapid dramatic change to the industrial landscape driven via implementation of two key Government White Papers. Working For Patients (1989) and The New NHS (1997) that emanated from successive Conservation and Labour administrations. Changes that significantly altered the marketing environment for Pharmaceutical companies operating within the UK. This research adopts an evolutionary perspective examining the progress of the 10 leading Pharmaceutical Corporations that by the end of 1999 represented 46.4% of the UK market in terms of sales. [See Table 1] Sales chosen as an objective output of the firm avoiding the problems associated with profit or return on equity calculations. The UK Pharmaceutical Industry was chosen for three key reasons. First a number of previous researchers (Cool & Schendel, Bogner, Thomas & McGee) had pointed to the differences within strategic groups found within the US Pharmaceutical Industry. Second, the UK operating environment changed radically during the 1990’s. Third availability of reliable sales data for the period made such a comparison feasible. Each corporation examined both collectively and as part of its constituent companies or divisions for as pointed out by Cool and Schendel [ref 1] rivalry between firms is conducted at the operating level of the firm. Where the company or division known to Porter [p129 ref 2] defines a strategic group as a group of firms in an industry following the same or similar strategy along strategic dimensions. While Porter also states [p132 ref 2] that individual groups are differentially affected by the five main competitive forces. In fact according to Porter [p149 ref 2] formulating competitive strategy in an industry can be viewed as the choice of which strategic group to The pattern of firms evolution within the Pharmaceutical industry can thus be illustrated by mapping visible strategic moves requiring the ability to classify individual firms and the groupings to which they belong. An aspect of research where biological sciences has a considerable pedigree in providing a variety of classification methods. Of which Cladistics appears to provide a transparent and highly visible way to illustrate the pattern of evolutionary relationships between species or strategic group. What is Cladistics and how is it useful
Cladistics a term coined by Mayr [ref 3], derived from the Greek word for branch is a form of classification originally developed by Hennig a German Entomologist. A system of portraying patterns of evolutionary relationships in branch diagrams based on synapomorphies or sister relationships. The definition of Cladistics [ref 4] is “a method of classification of animals and plants that aims to identify and take account of only those shared characteristics which can be deduced to have originated In the common ancestor of a group of species during evolution, not those arising by A methodology that has recently enjoyed somewhat of a renaissance driven first, by the ability of Cladistics to display numerical data where molecular DNA type analyses provide a whole new opportunity to discover evolutionary relationships between morphologically dissimilar organisms. Second by the advent of the computer and the ability of programs to store, manipulate, and rework vast databases providing the ability to explore evolutionary relationships via a host of new scenarios. In parallel has been the recognition that species may be analogous to firms together with use of biological metaphors to explore evolutionary relationships between firms. Here researchers such as Nelson and Winter, Hannan and McKelvey provided much Hennig’s phylogenetic system was based on classification to sister groups, a collateral not solely dependant on ancestry. Species being part of a sister group if two individuals are more related to each other than either is to a third taxonomic group. Hennig proposed two rules. First that all included in higher taxa must be descended from a single ancestral species. Second that all species descended from this ancestor must be included in a single higher taxon. Both conditions must be met. Mayr a contemporary of Hennig agreed with only the first of these principles coining the term Cladistics from Hennig’s single-minded attention to the order in which species branch With regard to evidence Hennig proposed that synapomorphies i.e. the presence of characters that the first two taxa possess but the third lacks should be subject to “reciprocal illumination” i.e. a relationship hypothesised from one source of information should be tested by bringing another source of evidence to bear upon it. If both sources support the hypothesis he viewed it as confirmed. Views formed in part through agreement with some observations of the ideal morphologists a group contemporary with Hennig in Germany. Alternatively termed typologists the morphological viewpoint stated that classification should be theory neutral and based upon a compendium of observations. That discrete patterns exist in nature awaiting discovery, that we do not need to know the origin of to discover. A view that has merit with the mapping of strategic groups where strategies remain unknown until after the event to an external observer. Hull’s [ref 5] reviews the place of Cladistics within modern evolutionary theory before discussing the important influences that the advent of computers and the growth in the science of molecular biology had upon the recent revival in the use of Cladistics. Starting from the work of Sneath (1957) who wanted to devise taxonomic methods usable by computers a key point of relevance is the statement by Hull that Sneath regarded weightings used by taxonomists as largely intuitive and therefore devised classifications based upon numerous equally weighted unit characters clustered via a mathematical algorithm. Sneath believed classification should reflect the scientific method by becoming empirical and repeatable. Work paralleled by Sokal whose chief objection to traditional taxonomy was the use of unequally In conclusion Hull makes a key point reporting the views of Platnick, that Hennig’s phylogenetic systems strength is its method. Method that has remained unaltered since its inception; only interpretation and data selection has changed. A point that augurs well for the use of Cladistics in a study of industry strategic groups as the method despite being tempered in the fire of strong debate remains unchanged. McCarthy and his colleagues [ref 6] describe a methodological assessment of the construction of a formal classification of manufacturing systems using Cladistics. Previous research, according to McCarthy fails to incorporate the benefits of the science of taxonomy an omission that he describes as reducing the usefulness, stability and accuracy of the classification. According to McCarthy lessons should be drawn from biological taxonomy based upon the discipline and rules regularly applied Addressing the question of why choose Cladistics? McCarthy gives three clear reasons. First, the authors regard Cladistics as theoretically superior with a deep philosophical underpinning that phenetic and other evolutionary based systems lack. Second, Cladistics produces classifications based upon the natural order of characters as distinct from artificially constructed more subjective systems. Third, by representing relationships in a cladogram the data employed is clearly illustrated along with all decisions and assumptions thus making the classification process transparent. Of these the final argument appears the most compelling and is the key McCarthy assumed first, that manufacturing systems evolve and have ancestors, a point clearly paralled by the Pharmaceutical Industry where although companies are broken up or merged yet constituent products or genes can be traced. Second, systems speciate. Third, manufacturing systems like companies are subject to natural selection so only those most fitted to their purpose survive. McCarthy proposed a mechanism for natural selection based on four key principles. One, there is natural and randomly occurring variation within a population. In critique of which the concept of random occurrence within a system designed for a specific purpose to a specification would appear unlikely. Two, heredity exists between systems i.e. offspring or subsequent systems resemble their parents and systems within the industry are more similar than external systems. “Genetic material” being passed on by people, communication, industry norms and standard operating procedures. Somewhat analogous to the proposal here that products represent company genes passed on complete with accumulated knowledge, history and tacit actions. Three, systems with clear advantages become more frequent within a population as industry adopts “best practice”. A point perhaps parallel to the follow my leader approach adopted by some companies with regard to mergers. See for example McGahan’s interpretation of Merck’s acquisition of Medco in the US a move followed rapidly by SKB and Lilly.[ref 7] Four, variation provides advantage and systems adapt to conditions or change over time to survive. McCarthy suggests selection of the best cladogram should be the result of statistical methods applied to the pair wise data and methods of parsimony. Where the simple rule - the longer the tree the poorer the fit. The shortest tree that requires the smallest number of evolutionary changes is adopted. McCarthy suggests two applications for Cladistics and several benefits from use of cladistic method. First through systematic and comparative method Cladistics permit the comparison of general attributes of complex systems. Second, these analyses may provide a comparative index of assistance in elucidating organisational process. McCarthy regards functional studies and Cladistics as complementary. Combination of environmental forecasts with the Cladistics could in McCarthy’s opinion provide a blueprint of which organisational species will dominate or languish. “Regardless of industrial sector organisations could use cladistics as an evolutionary analysis technique to determine their origins and current position relative to their peers.” Barthelme et al discuss the difference between phenetic and phylogenetic evolution before proposing that since technological concepts are subject to evolution a parallel between knowledge classification and species classification seems valid and appropriate. A point directly relevant to the accumulation and development of knowledge so critical to successful pharmaceutical research Cladistics is chosen to model the evolutionary memory. First, because phylogenetic theory traces the evolution tree by applying rigorous criteria, which in the author’s opinion fit with the technological concepts, proposed. Second, in their opinion Cladistics describes the rationale, argument and chronological progress of the knowledge based system through time giving cognitive meaning to evolution. Third, Cladistics allows organisational system species to be clearly grouped according to their evolution, directly in support of what is proposed here. Radical Change Within the UK Market
Prior to 1990 the UK market could have been described by the following key 1. GP led but heavily influenced by hospital opinion through the vehicle of referral and local opinion leader meetings. 2. Non-price sensitive with little awareness of the cost of individual 3. Lower than average access to customers with larger companies dominating customer contact through muscle marketing via large 4. Each GP operating as an individual practitioner with sales calls targeted on the individual doctor, generally a 1 to 1 discussion of product features and benefits of 8 to 10 minutes duration. (Individual field force observation supported by numerous GP discussions) 5. Little or no Health Authority influence upon how practice is run. A situation that had changed little since the establishment of the National Health Service (NHS) by Bevan in 1948 with the tenet the provision of a comprehensive health service delivered free at the point of need and funded out of taxation. [ref 8] But times had changed and the UK Government were facing increasing problems of 1. As the demographic makeup of the population changed with an increasing elderly population, a lower proportion of working population and less recruitment to the working community as people 2. People’s expectations of healthcare provision increased with wider access to education, the growing consumer movement and influence of 3. As productivity of pharmaceutical research and development declined with less attractive targets of high population with high-unmet medical need and increased competition the cost of new products and new 4. The growth in age related diseases such as Alzheimers. Thus the gap between demand and planned funding began to widen and the Thatcher Government in 1989 became the first administration willing to tackle this thorny and Working for Patients published in 1989 completely changed the rules of engagement for the Pharmaceutical Industry within less than two years rendering ineffective the adaptive evolutionary processes adopted by a rather traditional industry up to this point. In dramatic moves these reforms decided to manage cost by a multi pronged 1. Bringing strategic control for both GP (Community) and Hospital spending 2. Introducing the concept of competitive tendering for healthcare through an internal market consisting of purchasers (Health Authorities and Fund holding GPs) and Providers (Hospitals and GPs).[ref 9]. 3. Simultaneous introduction of measurement of pharmaceutical costs together with targets and incentives to reinforce desired behaviour. For example PACT (Prescribing Analysis and Cost) developed in Lothian was introduced to the NHS as part of the Thatcher reforms. PACT also provided a benchmark to compare Health Authorities or practices with UK norms. 4. Increased funding for computerisation of general practice. 5. The appointment of third party influencers to drive desired changes at a local level. Several hundred Pharmaceutical and Medical Advisers were employed to change GP prescribing behaviour as part of the Thatcher reforms. These changes announced with the publication of Working for Patients were implemented in April 1991 and marked a rapid and profound change in the UK The following table illustrates the radical nature of this change. Pharma Environnent pre 1990
Pharma Environment post 1991
GP prescriptions almost always written as Computerised practices generate Hospital consultant exerts influence upon Move towards practice decision therefore Each GP follows own treatment regimens Introduction of practice and health In 1997 following the defeat of the Conservative Government Labour introduced their White Paper – The New NHS. Reform that abolished the division of General Practice into Fund holding and Non-Fund holding in its place introducing the concept of GPs working co-operatively within Primary Care Groups which bore more than superficial similarity to the Health Maintenance Organisations that operate throughout the US. The focus of these new groups who work “seamlessly” with local hospitals, Health Authority and Social Services remained on extraction of value for money from NHS suppliers. Primary changes still more severe with peer pressure supplementing prescribing advisers, local retail pharmacist involvement in Primary Care Groups to carry out drug utilisation review and construct local formularies and further concentration of buying power into the hands of less than 500 Primary Care Groups. Thus in less than a decade the Pharmaceutical Industries primary customer had been reduced from 36,000 independent practitioners to collective decision making and in many cases purchasing via under 500 Primary Care Groups. The pharmaceutical environment changed markedly as a result of these reforms that introduced structural, functional and cultural change to the National Health Service. 1. The Pharmaceutical Industries access to key customers. 2. The customers “independent” decision making process 3. The relative importance of the respective customer groups 4. Increased barriers to new product entry 5. The average length of the product life cycle with GP brand loyalty being replaced almost overnight by a switch to computer generated generically Methodology
Corporations were split into their constituent companies or operating divisions. First, because pharmaceutical firms ultimately compete at the level of the individual physician seeking to displace current choices with their products. It is at this level that reputations are built. Second due to the Pharmaceutical Industry being subject to a number of mergers and acquisition ownership of the firm may change during the period and acquisition is clearly a strategic option for Pharmaceutical Corporations. Each company was examined in terms of their product portfolio at sub-therapeutic class level not therapeutic class, an approach that differs from previous researchers (Bogner, Thomas & McGee, Cool & Schendel) The rationale that companies compete via substitutable products thus R3A inhaled bronchodilators are not substitutable by R3D inhaled corticosteroids. The first is for acute relief, the second for chronic control, often used in concert but not generally as substitutes for one another. Relationships between companies were traced through using strict cladistic principles with product “genes” being mapped from company to corporation, where a therapy sub-class is solely served from two merging companies this node is coded a 1.1 as against 0.1 when emanating solely from one parent. The approach here differs from that adopted by McCarthy in relying almost solely upon proven evolutionary “product gene” relationships rather than phenotypic observed characters, which were applied later to confirm group selections as per Hennig. Age of the portfolio was also taken into account first because new products may be expected to receive a greater degree of promotional time, second in order to examine the relative number of new product introductions. New market introductions chosen as the measure for refreshment of the corporate portfolio rather than patents first because this removes the confusion over which patent is relevant as when different isomers were patented as with Zantac, [see Lynn p 165] or the plethora of process patents that can surround some NCE’s. Second, patent expiry is taken into account here due to focussing on sales as the ultimate measure of corporate success. An approach that also removes the need to consider market discounts, such as those given by some companies to Dispensing Doctors, as IMS audits report gross sales into Finally companies were placed into strategic groups based on observable patterns of behaviour their market phenotype. Where phenotype is defined as the set of observable characteristic of an organisation resulting from the interaction of its Discussion
On average 80% of the UK’s top 10 Corporations Sales come from products over 7 years on the market. Sixty percent from products over 10 years old. [See Table 2] Given the relative importance of these older products to corporate success it seems inconceivable that they do not drive the majority of corporate priorities. An investment path that leads back approximately twelve years before launch in many cases. [Ref 10] These facts together with the risk of research support the hypothesis that market conditions are not planned for a significant period pre launch but product market decisions are driven by the product therapy area, the nature of the product and the companies accumulated knowledge within that therapy area. Therapy area at sub-therapy class level determines the competitive arena into which the product is launched. The number and nature of competitors. The relative weighting between hospital and community influence, balance between acute and chronic therapy, key customer groups, their priority and a number of related critical decisions. Such decisions are therefore from this point on at least in part path dependant Decisions build upon one another as pointed out by Carroll and Hannan The company’s current priorities are therefore significantly affected by the mix of product sub-therapy areas, the number of products within each one and their relative weighting as regards sales both actual and potential. The hypothesis presented here is that this set of sub-therapy classes within the company is analogous to a corporate genome that sets priorities as to product mix and a myriad of product related decisions. The number of products that the company has within that sub class determines the number of times that the gene is expressed and relative importance in terms of sales determines the dominance of the gene. In evolutionary terms company research could be serendipity or due in part to random chance or guided towards a set objective. As Barthelme et al [ref 12] point out such selection processes could be explained by two conflicting models used to explain the process of evolution. Darwin proposed variation based on pure random chance with fitness to the environment via a natural selection process, whilst Lamarck believed variations to be of a directed intentional nature. Either, evolution is gradual, incremental guided by chance and affected by natural selection. The survival of the fittest concept. Alternatively evolution may be either continuous or occur in a series of jumps or saltations, a process guided by intelligence. A Lamarckian viewpoint that genetically transmitted traits originate by adaptation, such inheritance the result of goal directed behaviour. Barthelme proposes that a corporate knowledge system in many ways analogous to the corpus of research knowledge, where objectives are clearly defined fits a Lamarckian approach better as development of a goal directed system does not result from a random process. Similarly accumulation of knowledge is based on relevance. The principle of acquired characteristics. Once a product enters development however, the accumulated cost of bringing it to market some $700m dollars according to latest McKinsey estimates [ref 13] ensure that decisions to refresh the product portfolio are not taken lightly and the cost of The corporate genome also affects the fit that the company has with its environment first and foremost as customers attached reputations to products and companies that promote them. Second as companies build strength in specific areas they develop both research and market franchises. Market franchises because success in an area must increase the companies willingness to complement its portfolio by licensing in related products, secondly companies wishing to licence products or co-market their product would be expected to find related market success attractive. Thus companies are driven by their focus upon specific therapeutic areas that can either be the reward of a long-term research commitment or an opportunist action to in licence a specific product. Allen & Hanbury present a classic example of this with a long-term commitment to respiratory medicine as illustrated by examining their corporate Similarly companies do not compete at the corporate level or even at the individual company level but at sub-therapeutic area level where one product must displace another in the mind of the physician. A point brought out in stark relief through the use of the corporate genome analogy, where the intense rivalry between Astra and Glaxo first drawn attention to by Lynn [ref 14 p 165 & 229] who states that before launch Astra rattled Glaxo with Losec a head on challenge to the company’s most successful product and that “with its control of the ulcer market to protect, Glaxo set about systematically rubbishing the new product”. A situation very clearly illustrated by examination of the relevant sections of the two companies genomes at a specific The examination of such genetic relations can also be used to track the evolution of corporations through a series of mergers, disposals and acquisitions. In table 5 part of the base data for a cladogram illustrating the Astra Zeneca merger clearly shows the combination of the two entities and a genetic audit trail that can be used to distinguish between genes in the new entity that build on existing strength that may be called complementors, genes that remain isolated from the main backbone of the organisations activities and that may be sold off to complement other organisations and products that were important to the organisation many years ago but now are vestigial characteristics of the previous organisation. With the use of a cladistic approach based upon the companies genetic make up a number of different strategic groups are discernable. 1. Consolidators – these companies illustrated within this sample by Novartis and Sanofi Winthrop exhibit the following phenotypic characteristics. First over the last decade they have acquired several companies with different modes of operation. Novartis for example was formed through the merging of Ciba and Geigy two operating companies of Ciba Geigy strong in sales of gynaecological and rheumatological products respectively and focused on the GP or Community market. These two being merged with Sandoz a company with a strong hospital orientation. Through examining the respective corporate genomes there appears to be little overlap or immediate complement between the three. Sanofi was the combination of Sanofi and Lorex Synthelabo GP oriented companies. Second, on merging the companies identities were immediately subsumed which may for example have obliterated decades of accumulated market reputation and loyalty. 2. Glaxo acquired Wellcome via hostile takeover in 1995, Zeneca merged with Astra in 1998. A key characteristic of these organisations is that while they may merge internally for efficiency in research for example externally each operating company is kept separate with its own product focus. Smith Kline could also be placed within this group alongside American Home Products. These companies also have a history of using licensed products to exploit opportunities and plug portfolio gaps. Zeneca licensing in lisinopril (Zestril) which complemented strength in cardiovascular medicine and constituted 12.1% of Astra Zeneca’s sales the 2nd most important contributor to sales. SKB acquired the licence for paroxetine (Seroxat) when they merged with Beecham in 1989. Seroxat SKB’s largest product represented 38.5% of sales at year-end 1999. Similarly American Home acquired the UK rights to lansoprazole (Zoton) when they purchased Lederle, a product that represented 45.4% of sales by the end of 1999. [Source IMS BPI MAT 1999]. 3. MSD, Lilly, Pfizer and Warner Lambert represent research led companies that have remained committed to organic growth via strong operating companies focused upon defined product therapy areas. In conclusion the use of Cladistics may provide a systematic method to analyse and organise knowledge about industrial populations. A structured process for the study of diversity allowing identification of patterns and relationships that may explain the existence and evolution of different strategic groups. The resulting cladogram a visual and transparent intellectual vehicle of practical value for interpretation and explanation. A snapshot of the evolutionary history of a company and constituent firms comprising an industry. In short a method that in combination with the corporate genome concept may allow timely and relevant comparisons to be made between different pharmaceutical groups tracking evolutionary relationships and clearly discerning the specific locus of competition. Acknowledgement
The Author wishes to thank Gerry Williams of IMS UK for permission to use data References.

1. Cool, Karel. & Schendel, Dan. (1988) Performance Differences Among Strategic Group Members. Strategic Management Journal. 9, 207-223.
2. Porter, Michael E. (1980) Competitive Strategy. New York. Free Press.
3. Mayr, E. (1965) Classification and Phylogeny. American Zoologist 5, 165-
4. The New Oxford Dictionary of English (1998) Oxford University Press. 5. Hull, David L. (2001) The Role of Theories in Biological Systematics. Stud Hist Phil Biol & Biomed Sci. 32, No 2, 221-238.
6. McCarthy, Ian., Ridgway, Keith., Leseure, Michel., & Fieller, Nick. (2000) Organisational diversity, evolution and cladistic classifications. Omega 28, 77-
95.
7. McGahan, Anita. M. (1994) Focus on Pharmaceuticals: Industry Structure and Competitive Advantage. Harvard Business Review. November-December 115- 124. 8. Powell M.A.(1997) Evaluating the National Health Service. State of Health 9. Jones T (1994) The Structure of the National Health Service. Publishing 10. Janzen, William. P. (1996) High Throughput Screening as a discovery tool in the Pharmaceutical Industry. Laboratory Robotics and Automation. 8, 261-
265.
11. Carroll, Glenn. R. & Hannan, Michael. T. (1995) Organizations in Industry. Strategy, Structure and Selection. Oxford University Press. New York & Oxford. 12. Barthelme, Francoise., Ermine, Jean-Louis., & Rosenthal-Sabroux, Camille. (1998) An Architecture for knowledge evolution in organisations. European
Journal of Operational Research. 109, 414-427.
13. Lynn, Matthew. (1991) The Billion-Dollar Battle. Merck v Glaxo. Heinemann. 14. Bastianelli, Enrico.,Eckhardt, Jurg., & Teirlynck, Olivier. (2001) Pharma: Can the middle hold? The McKinsey Quarterly 1, 117-125.
The Top 10 UK Pharmaceutical Corporations Corporation Market
SKB 3.8 Pfizer 3.5 Novartis 2.7 Lilly 2.3 Warner Lambert Product Portfolio Age Profile Top 10 Corporations Years from Launch
SKB Pfizer Novartis
1999 1999 1999 1999 1999
1999 1999 1999
Source IMS BPI Data MAT 1990, 1994 & 1999 Why Astra and Glaxo Were Intense Competitors Products Therapy % of Sales Products Therapy % of Sales Products Therapy % of Sales Example of Part of Base Data for Astra Zeneca Cladogram Red = At least 1 product (gene) from each parent Yellow = Astra synapomorphies

Source: http://www.druid.dk/conferences/winter2002/gallery/leask.pdf