ریسک های نوآوری برون سپاری در توسعه محصول جدید دارویی

New product development, which until relatively recently was a core activity within the pharmaceutical industry, is increasingly being outsourced (Howells et al., 2008). This is symptomatic of the transformation of product development towards a more decentralised, network-based and alliance-rich function. We have selected the pharmaceutical industry for this paper because it offers an excellent example of a research and technology intensive industry where outsourcing has led to problems in the innovation and new product development process. We examine the impact of the increase in outsourcing on the way firms manage the clinical trials process. The increasing role of Clinical Research Organisations (CROs) has dramatically changed the nature of the innovation process. CROs initially provided a limited service in clinical trials management, but have gradually evolved into organisations that have expertise across a much wider spectrum of the clinical development process. The fragmentation of the innovation process, which has arisen from the increased role of CROs, creates innovation risks for the pharmaceutical companies. In this paper we discuss the risks associated with knowledge losses for research and technology intensive industries that may arise from lack of integration of new product development activities and develop a number of propositions for further research.

The pharmaceutical industry is facing major productivity challenges. For example, over a 10 year period the number of new drugs reaching the market declined by 20%, while research and development (R&D) expenditure increased by 60% (CMR International, 2006). The cost of bringing a new drug to market has increased dramatically and is estimated to be in excess of $1.2 billion (DiMasi et al., 2003 and DiMasi and Grabowski, 2007). Competition remains high and lead times have fallen, with the average period of market exclusivity down to less than two years (DiMasi and Paquette, 2004). A key development has been the rise of the importance of the contract research organisations in response to large scale outsourcing of research activities around clinical trials. The pharmaceutical industry provides lessons for other research and technology intensive industries that have not yet developed their outsourcing activities to the same extent. The purpose of this paper is to examine the impact of the increase in outsourcing on the way research and technology intensive firms manage their new product development process and to discuss the potential harm that this may have for such industries. Discovery research, the scientific laboratory research aimed at identifying new drug targets and novel pharmaceutical agents, is seen as the main driver of innovation. Indeed, much attention has been given to renewing technologies in this area, particularly in the fields of biotechnology, genomics and associated technology platforms (Siedlok et al., 2010). Organisational changes occurring as part of this transformation of the discovery research base are well described in the literature, with networks (Powell et al., 1996 and Huston and Sakkab, 2006), clusters (Chiaroni and Chiesa, 2006 and Casper, 2007) and strategic alliances (Deeds and Hill, 2006 and Gutierrez de Mesa Vazquez et al., 2006) all contributing to a move towards ‘virtual’ research forms (Chesbrough and Teece, 1996, Jones, 2000 and Gassmann and Reepmeyer, 2005). As argued by Chesbrough, 2003a, Chesbrough, 2003b, Chesbrough, 2003c and Chesbrough, 2006, the process of innovation has shifted from one of closed systems, internal to the firm, to a new mode of open systems involving a range of players distributed up and down the supply chain. Less well described are the changes occurring in the later stages of the development process, more specifically within the management of the human clinical trials required for proof of drug efficacy and safety before regulatory approval and marketing. Clinical development is the longest and most expensive stage of the pharmaceutical new products process, requiring more than 10 years for completion and representing 50% of total R&D spending (Parexel, 2004). The clinical trials stage is characterised by relatively low levels of external technology sourcing or out-licensing (Bianchi et al., 2011 and Malik, 2011). However, the management of clinical development has seen a move towards large-scale outsourcing to innovation service providers (rather confusingly for our purposes termed ‘contract research organisations’ (CROs), but the clinical research being part of late-stage ‘development’ as the term is more broadly understood). The CRO industry has grown to represent over 25% of the total spend on clinical trials, the leading CROs alone managing more than 23,000 clinical trials per annum and having revenues in excess of $5 billion (Milne and Paquette, 2004, Parexel, 2004, Tufts Center, 2006a and Getz, 2007). Outsourcing of product development can be seen as an attempt to manage a diverse range of technologies across different product classes (Coombs et al., 2003). It may also speed product development processes, thereby reducing time to market (Chatterji, 1996, Bayus, 1997 and Mahnke et al., 2006). Such innovation outsourcing might reduce costs, particularly for incremental innovations (Piachaud, 2005 and Huang et al., 2009). Outsourcing may also give access to global knowledge sources, so increasing development productivity (Gassmann et al., 2010). However, these potential benefits of outsourcing product development are by no means certain. Despite the trend towards increased outsourcing of development activities and a growing market for external product development services (Chiesa et al., 2004) there is limited recent evidence to suggest that firms may be beginning to take back development activities in-house because of the difficulties in managing outsourced innovation processes (Zirpoli and Becker, 2011). There is a growing literature on the innovation risks associated with large-scale outsourcing, particularly where non-peripheral business functions are concerned (for a general overview see Quelin and Duhamel, 2003). Alexander and Young (1996) highlight the risk of supplier dependency. Barthélemy (2001) identifies a number of hidden costs of outsourcing, and Earl (1996) suggests potential problems in selecting service providers with the necessary capabilities, particularly in environments of rapid technology change. Hoecht and Trott (2006) draw attention to risks of information leakage and possible losses of technology-based core competencies. Similarly, Zirpoli and Becker (2011) suggest that outsourcing design and engineering tasks in new product development might lead to loss of architectural or whole-system knowledge. There is also the possibility that innovation outsourcing may hinder a firm's ability to capture the full potential for new product ideas. The innovation management literature informs us that the capacity to absorb new knowledge, to transfer and diffuse knowledge and the ability to learn through interaction are crucial success factors in innovation (e.g. Cohen and Levinthal, 1989 and Zahra and George, 2002). In a recent review of innovation outsourcing research Stanko and Calantone (2011) demonstrate the many unknowns in terms of the performance implications of outsourcing. Hsuan and Mahnke (2011) propose a number of pressing research needs in this area, including greater focus on the management of innovation outsourcing processes, analysing these processes at multiple levels, and exploring the role of service organisations in creating value in the outsourcing process. Based on a number of key informant interviews with innovation managers from pharmaceutical companies and innovation service providers this paper explores the potential risks to a firm's innovation capability. In particular we focus on the use of CROs to undertake part of the late-stage clinical development process. We address three broad research propositions. Firstly, innovation service providers (the CROs) have become key drivers of clinical product development in the pharmaceutical industry. Secondly, the absorptive capacity of pharmaceutical firms is the key to their success in product development in a business environment where the CROs have taken on many of the key tasks. Thirdly, the pharmaceutical firms need to maintain the integration mechanisms necessary to uphold their absorptive capacity, diffuse knowledge and support interactive learning, and so avoid the risk of losing control and creativity opportunities for new products. The paper is structured as follows: Section 2 characterises the clinical product development process in the pharmaceutical industry and the role played by CROs as innovation service providers; in Section 3 we review the literature on innovation outsourcing and develop research propositions; Section 4 explains the research approach used; Section 5 analyses the potential risks of knowledge losses, and Section 6 concludes.

This paper has explored the outsourcing of clinical development to innovation service providers within the pharmaceutical industry. Our exploratory research provides support for the proposition that the scale and scope of outsourcing to innovation service providers (here termed CROs) has reached a level where they have become key drivers of product development success. We find no evidence to support the fact that firms are beginning to take back development activities in-house (Zirpoli and Becker, 2011). Large-scale outsourcing is apparent at the level of individual functional tasks, clinical trial projects and drug development programmes. Also, there is an emerging trend for ‘preferred supplier’ relationships with smaller numbers of CROs, and global outsourcing of particular functional disciplines such as clinical trials management. Time and costs were not seen as critical to the outsourcing decision, but rather flexible resourcing to cope with the changing capacity requirements arising from the technology uncertainty in new product development programmes was a major factor. There were also clear indications of pharmaceutical companies making strategic outsourcing decisions with the aim of developing new areas of technological know-how (Coombs et al., 2003 and Gassmann et al., 2010). We do find some support for the proposition that the absorptive capacity of pharmaceutical firms is the key to their success in product development in a business environment where the CROs have taken on many of the key tasks. The CROs appear to vary in terms of innovation potential, from those preferring to operate within strict parameters established by the client pharmaceutical company to those, often smaller CROs and partner biotechnology companies, where greater operational autonomy is used to stimulate innovations in service provision. This supports the idea that such project-based firms may not always be sources of innovation (Blindenbach-Driessen and van den Ende, 2010). It would be important to understand the factors influencing the creative climate within and between these organisations (Sundgren et al., 2005) but this is beyond the scope of this exploratory study. We found some indication that loss of technological know-how (Hoecht and Trott, 2006) might be an emerging problem as a result of full-scale functional outsourcing of particular role and task areas, often on a global basis. In such situations the innovative capacity of the service providers would be even more critical. The scope of outsourcing was widespread, with a high degree of internationalisation (Tijssen, 2009). The highly structured and regulated nature of the clinical development process seems to allow a degree of modularisation (Mikkola, 2006), and there is, as might be expected, a high level of focus on objectified knowledge in the form of databases of trial outputs. Information and communication technologies are increasingly global, but remain most well developed within single organisations and within particular functional areas (Howells, 2002). There was, however, a relatively low level of understanding between informants concerning how the broad-based knowledge of the overall product ‘architecture’ is maintained (Baldwin and Clark, 2000 and Zirpoli and Becker, 2011). The more the development process is segmented the more it becomes important to maintain the absorptive capacity of the commissioning organisation (Cohen and Levinthal, 1990; Zahra and George, 2002). It is argued that the social integration mechanisms for making sure that potential knowledge sources are fruitfully exploited and realised within and between companies are key to innovation performance. Given the importance of the social integration mechanisms we believe that it merits specific further research to find out whether these mechanisms still work for the pharmaceutical firms that have outsourced an increasing number of very important functions to the CROs. In the traditional business model of the pharmaceutical industry the pharmaceutical firm itself is at the centre of the network and it has clear organisational boundaries around its in-house activities and co-ordinates its relationship with the different actors in the medical sector (Lane and Probert, 2007). This provides the firm with a strong position in terms of both in-house and external learning opportunities and reduces the risk of information leakage to competitors (Hoecht and Trott, 2006). With the reliance on CROs in undertaking key aspects of the clinical trials the whole issue of risks in the maintenance of business relationship appears even more crucial (Powell et al., 1996). This is particularly the case in a technology-intensive environment such as the pharmaceutical industry, where the phenomena of technological and market uncertainties, competitive volatility, high R&D expenditure, rapid obsolescence and presence of network externalities prevail (Mohr et al., 2004 and Yadav et al., 2006). Companies rely on inter-firm relationships and the knowledge they encompass to achieve technological innovation and bring that innovation to the market (Harryson et al., 2008 and Easingwood and Harrington, 2002). The new emerging outsourcing based model of development in the pharmaceutical industry, resembles a diffused innovation, loosely coupled system. In this system the pharmaceutical firms cease to have their own individually tailored networks and so may be less able to act as ‘hub firms’ and extract innovation value (Dhanaraj and Parkhe, 2006). Hence, we find tentative evidence to support the proposition that the pharmaceutical firms need to maintain the integration mechanisms necessary to uphold their absorptive capacity, diffuse knowledge and support interactive learning, and so avoid the risk of losing control and creativity opportunities for new products. 6.1. Management implications The above findings raise a number of implications for Clinical Development Managers within the pharmaceutical industry. Most important is the need to manage the potential knowledge losses from outsourcing within a complex matrix of inter-related development phases, technical functions, and activities at project, programme and system levels. The conduct of clinical development programmes is often non-linear, with emergent knowledge and iterative learning cycles (Powell et al., 1996). Clinical Development Managers seek to manage inter-functional knowledge by externalisation and combination of explicit knowledge in the form of clinical trial protocols, project and programme plans, good practice guidelines and regulations, and the medical-scientific data from the clinical trials themselves. However, knowledge losses may still occur due to lack of interoperability of IT systems, lack of transparency, and lack of absorptive capacity within and between functions. Although these potential knowledge losses may be mitigated to some extent within the boundaries of the firm, R&D Managers need to consider use of common IT platforms, incentives for openness and transparency, and the level of technical and contextual understanding when outsourcing to CROs. Integration of tacit knowledge through internalisation and socialisation may be improved by use of cross-functional teams and staff secondments, but here too there are significant issues in relation to CROs. Functional experts from the pharmaceutical companies and CROs need direct access to each other in order to build the relationships required for tacit knowledge sharing. The relationship management approaches on both sides need to be sufficiently robust to achieve some stability and continuity of interaction. Creation of ‘clinical outsourcing teams’ within increasing numbers of pharmaceutical companies provides a broader focus for managing clinical trials operations. However, the success of these initiatives will depend upon the breadth of outcome measures employed, and the extent to which they include intangible outcomes such as learning and knowledge creation. Lack of strategic input by some CROs highlights the need for operational staff to be involved in clinical trial design, and for Clinical Development Managers to ensure a good knowledge of clinical trial management issues at all levels of decision-making within the project team. Stakeholder and relationship management issues are particularly important for Clinical Development Managers at the programme level, where external experts and steering committees have to be managed, changes to country-specific regulatory standards need to be understood, and the implications of consolidated findings from clinical trials must be considered. At this level, issues of absorptive capacity appear to be increasingly important, and large-scale outsourcing to CROs can create imbalances between programme participants. For example, our informants note that smaller pharmaceutical companies often lack statistical and biometrical expertise, and so are less able to critique and challenge what the CRO has delivered. Informants also mentioned the need to avoid a blame culture between the programme level participants, allowing greater transparency and earlier resolution of issues. Clinical Development Managers should carefully design suitable relationship governance structures and relationship management instruments (Oxley and Sampson, 2004), but also need to consider the extent to which trust and control approaches to risk management (Hoecht and Trott, 1999) are possible. The ability to integrate and manage the functional, operational and programme-level knowledge domains within pharmaceutical clinical development is itself a form of system-level architectural knowledge (Baldwin and Clark, 2000 and Zirpoli and Becker, 2011). This highly specialised “know-how” is tacit and learned through practice and experience (Nonaka, 1991). Historically, this tacit knowledge was built up within pharmaceutical companies through structured career development and careful consideration of social interaction through use of teams and the built environment. Our informants identified examples of highly sophisticated organisation development approaches within pharmaceutical companies, including use of ‘holistic’ computer simulation software in group workshops. However, Clinical Development Managers need to be aware of the impact of outsourcing on this system-wide knowledge. With clinical trials activities distributed across a network of participants, creating and communicating this form of tacit knowledge becomes more difficult. Recent trends towards outsourcing full clinical development programmes create a focus for developing such knowledge within the CROs. Similarly, one key informant describes a virtual development organisation established by a major cancer charity, where even the system-level knowledge and expertise was outsourced to a group of highly expert consultants. One further implication for Clinical Development Managers emerging from this discussion is the need for awareness concerning the changing network position of their organisation at each of the three levels of knowledge generation; operational, programme and system-wide. Consideration of clinical development outsourcing through this multi-dimensional framework would seem to offer interesting and valuable opportunities for further study. 6.2. Strategic implications For many years the management of clinical trials was conducted by pharmaceutical companies’ in-house ‘medical’ departments. These in-house departments developed and maintained the full range of medical/scientific, statistical, trials management, regulatory and drug safety expertise required for the conduct of clinical trials programmes. Such ‘closed’ innovation systems offer a number of advantages, allowing for bureaucratic control (Hoecht, 2004), minimising information and knowledge leakage (Norman, 2004 and Oxley and Sampson, 2004), establishment of organisational routines (Garsten and Grey, 1998), and development of core capabilities (Barney, 1991) and ‘enduring idiosyncratic differences’, which create competitive advantage (Cockburn and Henderson, 2001). External collaborations were largely limited to medical investigators and regulators. There was, therefore, some degree of ‘distributed innovation’ (Coombs et al., 2003). The medical community, with direct relationships to the pharmaceutical medical/scientific specialists, were collaborative participants in the ‘research community’ of the clinical trial and drug programme, creating a network for explorative knowledge creation (March, 1991 and Powell, 1990). It has been argued that a firm’s success crucially depends upon its ‘centrality-position’ in inter-organisational networks and the experience gained in managing these networks (Powell et al., 1996). The larger pharmaceutical companies were both well positioned and experienced in managing relationships within these pharmaceutical-medical networks. The move away from the integrated in-house model may have been a rational response to pressures in the industry but is problematic from the innovation perspective in that it may reduce knowledge creation and learning opportunities that arise from the closely integrated development process. Furthermore, activities that may seem peripheral and mundane at first sight can turn out to be very crucial in terms of learning opportunities that they provide. For example, the evaluation of the statistical data in Pfizer's famous drug trials led to the opportunity for Viagra (Trott, 2008, pp. 275–281). Reinterpretation of results can be a major source of new insights leading to new product ideas, such as new application possibilities. Similarly, the collation of results from the different parts of the studies commissioned to CROs present opportunities for evaluation and reassessment that can easily be missed if companies rely on other parties to summarise the results Linked to the findings of this paper is another possible unintended consequence of the decision to outsource parts of the new product development process. It may be exactly the fragmentation of the process into separate elements that may have lowered the barriers of entry for firms to enter the industry. In the traditional network the large pharmaceutical firms controlled the entire new product development process and in so doing created a high barrier to entry. The new emerging network model with a fragmented development process has enabled CROs to enter the industry and offer their expertise (Piachaud, 2002). Furthermore, pharmaceutical companies may also be losing the ability to appropriate innovation from the network within which previously they were central. This finding is supported within the innovation network literature (Dhanaraj and Parkhe, 2006 and Coenen et al., 2006). Finally, it is worthy of note that there may be lessons for firms beyond the pharmaceutical industry. Twenty years ago the key companies in the pharmaceutical industry could not have foreseen the changes in the industry structure and the impact on the research process that arose from outsourcing what was then considered mundane research activities to the CROs. Firms in other research intensive industries may also be tempted to outsource research activities to specialist providers. For example, commercial software developers such as SAP might be tempted to outsource software development applications to other specialist software developers. Such a move may appear to offer efficiency gains but would deprive them of potentially valuable product design insights. This might also lead to inadvertently nurturing future competitors in the wider industry just as the CROs in the pharmaceutical industry have gained industry importance at the expensive of the large pharmaceutical firms.