Introduction – the ‘dark-side’ of work and trust

The management of risk has become the over-riding concern of late-modern social policy (Kemshall, Reference Kemshall2002) in the UK and nowhere is this more evident than in the English National Heath Service (NHS). Notions of risk and corresponding uncertainty are at the very core of medical practice – in its application of evolving technology and expertise in seeking to alleviate morbidity and mortality. Whilst prima facie this management of the risk of illness to the patient is scientific, many have noted the ‘artistry of medicine’ (Schön, Reference Schön1983), emphasising the extent to which medical knowledge is acquired and honed through hands-on practice (Nettleton et al., Reference Nettleton, Burrows and Watt2008). Yet this craftsmanship of dealing with the risks of illness has been seen to produce its own risks. Sennett (Reference Sennett2008: 6) notes the potential for work to become an end in itself to the extent that the over-arching effects of the craft come to be of secondary importance. Such unacceptable practice became all too evident within the infant cardiac surgery department at Bristol in the 1990s where mortality rates were far higher than should have been the case: ‘A tragedy took place. But it was a tragedy born of high hopes and ambitions, and peopled by dedicated, hard-working people. The hopes were too high; the ambitions too ambitious’ (‘Learning from Bristol’ inquiry report quoted in Alaszewski, Reference Alaszewski2002).

The Bristol disaster, along with other dysfunctions within the NHS in the late 1990s has led to an increasingly stringent form of risk management. Whilst medicine, as already stated, is inescapably about managing uncertainty, an increasingly systematic and expansive form of risk governance has been rolled out across the NHS in order to avoid repeats of situations where clinical craft loses sight of patient safety and quality of care. Whereas trust, on the one hand, deals with the inherent unknowableness of the future (Keynes, Reference Keynes1921) by assuming away aspects of uncertainty (Möllering, Reference Möllering2006), risk management seeks to bring a certain degree of measurability to expectations, even though certainty about the future is impossible. In this way, risk reflects how ‘the nature of modern culture, especially its technical and economic substructure, requires precisely such “calculability” of consequences’ (Weber, Reference Weber and Runciman1978: 351).

Accordingly, an audit culture, centring round clinical governance, has enveloped the NHS as a means of holding practitioners accountable for the secondary consequences (patient outcomes) of their clinical work (Scally and Donaldson, Reference Scally and Donaldson1998). The origins of this culture, and indeed clinical governance, are wider, more complex and date back further than Bristol and the mid to late1990s. Geographical variations in care as underlined in the Calman and Hine (Reference Calman and Hine1995) report, increasing litigation costs and a longer-running interest in performance monitoring, accountability and audit – as epitomised in New Public Management – were fundamental in the emergence of the audit society. Yet the risk-focused rhetoric and format by which clinical governance and associated policies have been implemented has been specifically politicised and driven as a means to warding against the risks clinicians have been perceived to pose (Alaszewski, Reference Alaszewski2002) – both to patients and an effective, efficient NHS.

From a ‘transaction cost economics’ perspective – where exchanges of goods, services or knowledge can be facilitated within a market (via price), organisation (through control systems) or via trust – there has been a move away from this latter mode due to the levels of uncertainty it assumes away. Hence, although trust is by far the most efficient of the three approaches (Arrow, Reference Arrow1974), acting as a lubricating force (Fukuyama, Reference Fukuyama1995), the risks of leaving clinicians to their own ‘work’ have come to be perceived as too great. Correspondingly distrust becomes a means of ‘guarding against the abuse of power’ (Gilson, Reference Gilson2006: 370) and in a late modernity, where danger and uncertainty are less tolerated (Furedi, Reference Furedi1997), governance frameworks offer a means of pursuing certainty and managing risk through calculability and stipulation.

This paper will look at the turn away from trust within the NHS and assess the effectiveness of its successors. Whilst Bristol and other disasters have highlighted the ‘dark side of trust’ (Connell and Mannion, Reference Connell and Mannion2006) – where the legitimacy and motivations of professionals have been questioned within a context of wider economic constraints – moves towards the increased use of control and quasi-markets have their own ‘dark sides’, which have received much less attention and scrutiny from policy makers. The problematic nature of organisational and then marketised approaches will be considered in turn. Le Grand (Reference Le Grand2007), in his case for the utility of the latter, argues that what is actually required is a combination of choice/competition, control and trust in the running of the NHS. A third section will consider this position, contesting the existing paucity of trust applied within the NHS and underlining the dangers of the other two modes.

Command and control: medicine by numbers

The New NHS: Modern, Dependable (DoH, 1997) and key follow up white papers (DoH, 1998, 2000, 2001) have a particular understanding of how things have gone wrong within clinical practice, which is similar to the concerns raised by Sennett (Reference Sennett2008) already referred to. Left to individuals and small clinical teams, the practice of medicine in its parochial, craft-like state was too dependent on local expertise; hence, the potential for losing sight of best practice and optimal patient outcomes. The solution proposed by these policies was essentially the setting and monitoring of protocols, targets and performance measures in order to achieve a much more standardised, nationally homogenous practice of medicine across the NHS.

This move towards a more scientific-bureaucratic form of clinical practice (Harrison, Reference Harrison2002) has, therefore, sought to remove elements of uncertainty and variability associated with the craftsman-type model (Harrison and Smith, Reference Harrison and Smith2004) – in seeking to mitigate perceptions of risk associated with this. To the extent that medicine is not an art or craft, and that it is purely about the instrumental application of abstract knowledge, this move can be said to be purely positive. However, a number of critiques have highlighted the potential limitations of this bureaucratisation of medicine (Flynn, Reference Flynn2002) due to its inability to account for the wider, more human aspects of patient outcomes. Healthcare is inescapably about suffering, emotions, social values and norms, which the instrumental rationality of science and bureaucracy are unable to account for (Brown, Reference Brown2008a). Thus, the argument develops that bureaucratised medicine is a poorer, reductive, least-worst quality of medicine, in that the aspects which are immeasurable, but are nonetheless exceedingly important to patient experience, are neglected. So whilst command and control may offer a more modern and dependable medicine, this might polemically be understood as leading to a one-dimensional (Marcuse, Reference Marcuse1964) NHS: inhumane and valueless.

To return again to some of the notions raised by Sennett, whilst the possibility exists that the craftsman loses himself in the work to the extent that ‘nothing else matters’ (Reference Sennett2008: 6), so too may notions of the human become lost in the midst of bureaucracy to the extent that only ‘convention, efficiency and “internal coherence”‘ (Marcuse, Reference Marcuse, Bronner and Kellner1989: 121) come to be of significance. Such an account is a classic tale of the Weberian iron cage where, as with craftsmanship, man-made things inherently develop dysfunctional tendencies. In the interests of patient outcomes, the fundamental assumptions behind the efficacy of command and control are (a) that the stipulations/protocols and targets/performance management are capable of accurately representing quality patient care, and (b) that these same protocols and targets are capable of achieving high levels of compliance/achievement amongst practitioners (Brown, Reference Brown2008b).

Regarding the first of these suppositions, Rothstein (Reference Rothstein2006), amongst others, sets out a compelling argument of how the internal logic of risk regulation inherently tends towards the minimisation of risk posed towards the institution, rather than towards the individuals served by this institution. Risk is applied as a basis of governance due to its apparent incontrovertibility and probabilistic acknowledgment of the potential for failure in attempting to manage these problems. The more effectively and precisely such governance is carried out, the more sensitive institutions become towards the potential negative consequences of failing to successfully manage risks – ‘“good governance” gives rise to risk itself’ (ibid: 217). Transforming a wide array of social problems into problems of risk assists the institutions with the accountability and liability of their decision making (institutional risk) though may compromise their ability to deal effectively with the practice on the ground (societal risks). Hence the effects of this defensive process are essentially two fold – an increasing ‘risk colonisation’ of institutions and policy on the one hand, and equally worryingly a tendency within this growth of governance to seek to minimise ‘institutional risks’ ahead of those risks faced by society on the other (Power, Reference Power2003).

This highly devolved, abstracted management promotes a closed, circularity of quality (Lindeberg, Reference Lindeberg2005), where the importance of the patient is lost under the multi-tiered bureaucracy of the system (Power, Reference Power1997). This means that assumption (b) (above) of this form of performance and risk management is more likely to be flawed (Brown, Reference Brown2008b). Command and control suggests just that; yet as a growing literature has shown, both empirically (e.g. Public Administration Select Committee, 2003; Som, Reference Som2005) and theoretically (Gallivan and Depledge, Reference Gallivan and Depledge2003; Bevan and Hood, Reference Bevan and Hood2006), there is a limit to the extent to which a culture of targets and audits can assure compliance and, therefore, performance. The undermining of control mechanisms is likely to increase where the new frameworks are seen to compromise the craftsmanshipFootnote ¹ or artistry of medicine (Brown, Reference Brown2008b), as it is these aspects which give meaning to practitioners’ actions and it is meaning, values and norms which govern individuals’ actions (Habermas, Reference Habermas1987).

Choice and competition: incentivising control

In summary, command and control approaches are based on the proposition that simply trusting in healthcare professionals to carry out their craft opens up the proverbial Pandora's Box (Sennett, Reference Sennett2008) of what happens when ‘work’ becomes separated from, and more important than, patients’ best interests. Yet, as has been seen in the preceding section, the systems designed to safeguard against this happening are, by their very nature, subject to their own means of diverging from patient interest and, through their powers of compulsion, may engender larger numbers of professionals being distracted from optimal patient outcomes. For even where a lack of meaning within bureaucratic stipulations and measures compromises their power to compel compliance in a substantive sense, the bureaucratic burden necessitates a ‘blind self-preservation’ (Habermas, Reference Habermas1984: 398) form of superficial compliance within these structures. This distraction remains significant (Royal College of Nursing, 2008).

The limitations of command and control, and trust, lead Le Grand (Reference Le Grand2007) and others to advocate systems that seek to assure and improve healthcare quality primarily through choice and competition – the market. This model, it is argued, ‘fulfils the principle of autonomy, and promotes responsiveness to users’ needs and wants; it provides incentives for providers to provide both higher quality and greater efficiency; and it is likely to be more equitable than alternatives’ (ibid: 42). Thus, whilst craftsmanship and bureaucracy are flawed through their human involvement and design, it would seem that the market – in its supra-human nature – is beyond these vulnerabilities.

Le Grand accounts for both those workers whose motives are keenly focused on the patient (knights) as well as those who have been distracted away from these interests (knaves): ‘Knavish providers will want to attract users because their livelihoods and thus self-interest depends on it. But knightly ones will also want to stay in business so that they can continue to provide a service of benefit to users. So they too will want to provide services that attract users’ (ibid: 44). Latent within this basis of arguments for market incentives is the assumption that the ‘providers’ who benefit from the market are synonymous with those ‘providing’ the healthcare.

Of course, this is far from the case and those who run the healthcare trusts (the local healthcare provider organisations in the UK) have a somewhat distanced relationship from the professionals on the ground. Nevertheless, as Le Grand describes, the incentive is there for the organisations to maximise the quality and efficiency of their provision. The dilemma remains as to how those responsible for the local providers (chief executives and senior management) ensure that practitioners on the ground practice in a way that offers sufficient quality and value for money. The answer is through control mechanisms and, therefore, we return to the initial problematic nature of performance frameworks, merely at a different organisational level. For the ‘invisible hand’ is a human hand after all.

The alternative to such intra-organisational or intra-departmental command and control would be to invoke further internal competition and performance-related pay. Were this an industry dealing with the straightforward manufacturing of consumer durables, or aircraft maintenance (DoH, 2001), then such rewards/penalties would be straightforward. However, difficulties surrounding the comparability of cases and measuring ‘quality’ in healthcare make this massively complex. The choice of what to reward, and to what value, becomes a significant bureaucratic exercise in itself. Due to the humanity of not only the subjects but also the objects of this work, the likelihood of trade-offs between efficiency and safety and/or quality of patient experience remain high. The immeasurability of patient experience makes it inevitable that neither the market nor control mechanisms are sufficiently able to adequately ensure that what matters is rewarded or audited (Kennedy, Reference Kennedy2004: 162) – such is the intricacy of health, illness and the subjective human disposition that experiences it. So whilst the choice and competition agenda might well be popular with the electorate (Le Grand, Reference Le Grand2007), this should not be confused with popularity amongst patients. The vulnerability which separates the latter from the former necessitates a trust which is won through patient-focused care and embodied expertise (Calnan and Sanford, Reference Calnan and Sanford2004) – two qualities which are not easily commodifiable.

The new GP contract and the associated Quality and Outcomes Framework (QOF) represent one apparent success story to the contrary. For in their largely self-employed position and the linking of ‘performance’ to pay (at least in certain respects) through QOF, the problematic distance between provider-as-contractor and provider-as-carer is minimal, and therefore there is no command and control required to link one to the other. Initial research in this area does suggest though that even in this context there are moves towards a ‘top-down’ approach within larger practices, with correspondingly negative effects on motivation (McDonald et al., Reference McDonald, Harrison and Checkland2008). Moreover, the initial diagnosis/referral setting, public health goals and management of chronic illness typical within General Practice are arguably more amenable to standards, reward and risk-management protocols than is the case in acute healthcare. This means that GPs are better able to work within new governance frameworks, whilst maintaining a narrative of ‘holistic practice’ (Checkland et al., Reference Checkland, Harrison, McDonald, Grant, Campbell and Guthrie2008).

This does not seem to be the case for hospital-based professionals (Brown, Reference Brown2008c). In terms of the riskiness of acute medicine, Schön (Reference Schön1983) observes the frequency of anomalous situations that require a reflective, creative approach. Such scenarios for which protocol is insufficient (Alaszewski and Brown, Reference Alaszewski and Brown2007) underline the danger of an over-reliance on standardised (through bureaucracy or the market) means of delivering healthcare – for ‘routines tend to become increasingly dysfunctional over time: not only do they fail to adjust to new circumstances but “shortcuts” gradually intrude, some of which only help professionals to cope with pressure at the expense of helping their clients’ (Eraut, Reference Eraut1994:112). It is in this sense that the case for craftsmanship re-emerges.

Markets, hierarchy and trust: towards the lesser of three evils

To err is human, as indeed it is to fall ill. So far, this paper has noted the potential for pathological tendencies to develop within trust, hierarchies and markets due to their human components and therefore the likelihood of distractions away from optimal patient care – through too overt a focus on the craft itself, or through the need to satisfy performance measures/market pressures which are unable to adequately represent ‘quality’. Given that each of the options when assessed separately is so clearly flawed, it is of little surprise that the current NHS employs a mixture of all three.

Indeed Le Grand (Reference Le Grand2007), in his treatise on the merits of choice and competition acknowledges the necessity and inevitability of a mixture of means of managing transaction costs, though he argues of course in the direction of markets. Similarly, Darzi (Reference Darzi2008), in his NHS review, recommends the increased use of choice in driving NHS quality and in so doing fulfils Le Grand's criteria that such choice should be as free and informed as possible. Hence the proposition of hospital dashboards which would present neat accounts of a hospital's performance in areas such as Accident and Emergency waiting times, or infection control.

Media coverage over the last decade has made these two issues, perhaps above all others, fundamental to how hospitals are judged and critiqued. Ensuring that patients can choose to avoid an MRSA-ridden hospital would obviously send signals, via the invisible hand of the market, that serious action ought to be taken by managers. These managers would therefore, in consulting experts on infection control, realise that alcohol was the most cost-effective way of minimising MRSA and so administer a compulsory hand-washing scheme. This laissez-faire and risk-responsive system sounds ideal – until it is discovered that the distraction away from soap and water, combined with over-crowding to minimise waiting times, proliferates Clostridium-difficile.

Similarly with waiting times, it has become apparent that A&E staff, left to their own craft, develop inefficient tendencies which mean that patients wait overly long for treatment. By ensuring that patients wait no longer than a stipulated time (four hours), and attaching financial incentives/sanctions to this, efficiency is improved. Yet this also means that ‘gaming’ develops around this four hour time (Bevan and Hood, Reference Bevan and Hood2006) and that a patient with a minor complaint who has been waiting three hours 45 minutes is seen ahead of a more recently arrived patient with chest pains (posing very serious risk). Later on, this ‘heart’ patient is moved temporarily to a respiratory ward in order to clear beds, before later being finally moved to the appropriate unit. This bed shifting puts the patient at greater risk due to the lack of appropriate care in the respiratory ward and further fosters C-diff through the increased moving of patients.

These two anecdotes, based on interviews with professionals (Brown, Reference Brown2008c), beg the question: Do the distractions (or risks) generated by markets and hierarchies develop to become more grievous than the initial risks associated with craftsmanship? Proponents of markets and hierarchies would no doubt acknowledge these weaknesses but point to the substantial efficiency gains, for example in reducing long waiting times. Yet the gaming referred to in the previous paragraph questions the validity of this evidence. That one third of A&E doctors admit to regularly fabricating figures (BMA, 2007) suggests the possibility of an epistemological fallacy surrounding the success of performance frameworks and/or market incentives.

The two examples also add credence to Rothstein's (Reference Rothstein2006) analysis suggesting risk-management practices tend towards deflecting institutional risk (towards political capital as a result of media criticism over waiting times or infections) rather than necessarily minimising risk to individuals in society – in this case the patient. Because institutional risks are not able to be reconciled with societal risks, in that the former are constructed by the media and accordingly distorted, this inevitably leads to a divergence which in cases may well exacerbate risks to patients. However the need to be seen to be ‘doing something’ is one aspect where markets and hierarchies are strongly preferable, in that relying on trust as a means of fending against uncertainty inevitably tends, prima facie, towards doing very little. Risk management, on the other hand, is a visibly active (if often dysfunctional) way of seeking to deal with the ‘unknowable’ of the future (Keynes, Reference Keynes1921) – as applied explicitly through bureaucracy or as incentivised via choice and competition.

This tendency towards policy ‘meddling’/risk deflecting, combined with the epistemological fallacy over the positive effects of market/control incentives, means that the efficacy of trust is over-looked and under-estimated. Thus, whilst a means of organising transactions facilitated purely by trust would be unfeasible and dangerous, the undervaluing of trust in current policy and proposals means the ‘mixture’ is skewed towards markets and hierarchy. As noted in the introduction, trust is cheaper than markets and hierarchies due to the lack of bureaucracy entailed, with the costs of the latter two mechanisms becoming further apparent when their multiplier effects are considered. Because trust and control are not simply alternatives but dialectically linked (Gallivan and Depledge, Reference Gallivan and Depledge2003), imposing control (or market) mechanisms erodes the trusting and altruistic behaviour of those working within these frameworks, changing the way they relate to these frameworks and thus undermining the application of the control mechanisms. As attempts at control (to combat this) are increased further, the more these mechanisms are seen as illegitimate and circumvented or subverted (Brown, Reference Brown2008b). Whilst Le Grand (Reference Le Grand2003; Reference Le Grand2007) acknowledges this possibility of knights being made more knavish, the seriousness of its impact is seemingly underestimated due to the epistemic problems already described.

The NHS has been discussed thus far largely in terms of its service provision, yet it is of course very much a learning organisation as well – this aspect being central to its ability to provide low-risk, high-quality healthcare. Although the inevitability of a mixture of markets, hierarchies and trust has been acknowledged, Pilgrim and Sheaff (Reference Sheaff and Pilgrim2006) point out that in many cases the contradictory combination of increased marketisation and extended bureaucratisation leads to a paradoxical environment where ‘organisations are permitted, nay encouraged to learn, but not too much and not too openly’ (ibid: 10). Whilst trust may (again) be the most effective of the three bases for facilitating mutual learning within the ‘knowledge economy’ (Adler, Reference Adler2001), clinical governance – which has been placed at the heart of recent knowledge and quality strategies within the NHS – fails to harness this trust, relying instead on the bureaucratisation of medicine (Flynn, Reference Flynn2002). The rhetoric around clinical governance would suggest it is a newer form of ‘soft’ (Ferlie and Geraghty, Reference Ferlie, Geraghty, Ferlie, Lynn and Pollitt2005) ‘new public management’, yet in practice – through its use of monitoring and surveillance and its perceived muddling with wider efficiency/performance drives – it manifests itself as a harder, more traditional form of NPM (Brown, Reference Brown2008c).

Discussion: promoting inclusion, consultation and ownership

Choice is an increasingly advocated basis of healthcare provision in several European welfare states (e.g. Blomqvist, Reference Blomqvist2004; Propper et al., Reference Propper, Wilson and Burgess2006). In England, the recent review of the NHS also conceives choice as a key guiding principle to assure the quality of healthcare provision (Darzi, Reference Darzi2008). The Darzi review (ibid) further underlines the importance of listening to professionals and giving them an increased say in how services are organised and run. The arguments and evidence presented in this paper would suggest that these two tenets, whilst well intentioned, may lead to conflicting aims and muddled policies. For whilst choice and corresponding competition may, prima facie, offer patients better services (though evidence for this is limited and highly contestable), the dysfunctional attributes of command and control are inevitably incentivised through allowing patients to choose.Footnote ²

These tensions become further evident when trust is brought into the equation. Recent UK policies have recognised the importance of the concept (e.g. DoH, 2007) not only in terms of patient experience, but also equally in facilitating professional autonomy to practice their craft. Yet this rhetorical stance is not reflected in the substance of policies which, through their monitoring, surveillance and requirements for revalidation, demonstrate the deficiency of trust in these same practitioners. The pressures exerted through the system in order to verify good craftsmanship and ensure competitiveness correspondingly affect trust between managers and professionals, professionals and patients and amongst professionals themselves (Calnan and Rowe, Reference Calnan and Rowe2008). The transaction and checking costs are most visible in the bureaucratic and organisational distractions away from the patient. So whilst the patient may have greater choice in being able to avoid institutional risks and poor performance, because their individual experience is profoundly shaped by trust, which in turn is dependent on communicative interactions with their professionals (Brown, Reference Brown2008a), their actual experience of healthcare is not necessarily enhanced. Rather it may in fact be compromised by choice (albeit indirectly), where ‘choice becomes a “new paternalism” in which the availability of a patient-centred service, a high priority for users, does not figure on the agenda of providers’ (Taylor-Gooby, Reference Taylor-Gooby2008: 165).

Good healthcare, in order to be efficient, safe and effective must harness trust to a greater extent than is currently the case. It has been argued here that whilst pure reliance on trust may be ineffective at managing risk, performance and knowledge – the alternatives may be more problematic still. Hence, good governance – that which embraces trust – will be able to optimise efficiency (Arrow, Reference Arrow1974) and learning (Adler, Reference Adler2001) through inclusion, consultation and therefore ownership. Such means of regulating the craft of medicine will be more effectual in reaching these goals through being more able to win the concordance and facilitate the cooperation of professionals due to the increased legitimacy of these frameworks (Brown, Reference Brown2008b).

Rather than a mere reversion to old ways, a new form of ‘conditional’ (Calnan and Rowe, Reference Calnan and Rowe2008) and ‘reflective’ (Adler, Reference Adler2001) trust (i.e. informed rather than ‘blind’) – earned rather than assumed – would enable a more patient-centred form of care. Such trust fosters ownership, responsibility and community and is therefore ‘increasingly recognized as the organizational principle most effective in generating and sharing new knowledge’ (ibid: 220). Late modernity's heightened awareness of the fallibility of expertise, which has given rise to the governance and choice agendas, is precisely that which can enable this more reflexive (self-confronting) and risk-aware form of trust; one which is actively qualified in relation to specific risks, rather than blind (Lewis and Weigert, Reference Lewis and Weigert1985) or ambivalent towards uncertainty.

As with any form of risk management, a more thorough analysis must include an understanding of the way risk has been politicised and applied in the first place. The use of Bristol and other dysfunctions in constructing a political problem around the dark side of trusting in/relying on the craftsmanship of medicine has created its own Pandora's box, which, due to the manner in which risk management works to minimise institutional criticism (Rothstein, Reference Rothstein2006), is less open to being politicised itself. Thus, there exist more fundamental questions towards medical regulation and professional autonomy than simply providing, or being seen to provide, the least ‘risky’ form of medical care. ‘Thus tensions and conflicts of interests are deeply embedded in the demands on professions, and release ongoing dynamics into the health system and the health workforce. These tensions cannot simply be overcome by individual actors, whether professionals or service-users. They can, however, be reduced or reinforced by different models of state regulation’ (Kuhlmann, Reference Kuhlmann2006: 33).

These tensions, however, may have created, as Freidson (Reference Freidson2001) has argued, a form of professionalism where the professional acts as a mediator officiating over the interests of the state and serving the needs and demands of the public. The benefits of such a role, according to Freidson (Reference Freidson2001), are to ensure trust in public services and reduce the costs of governmental action and control. Thus, in the context of general practice in the English NHS a new type of GP may be emerging, such as a street-level bureaucrat (Checkland, Reference Checkland2004), who mediates between the managerial world of guidelines, evidence-based medicine and performance indicators and the professional practice of every-day medicine, or a public service entrepreneur, who adopts the values of the market to meet the needs of patients, while not being driven by the profit motive (Boyce, Reference Boyce, Kuhlmann and Saks2008). The latter would represent a hybrid of contrasting positions in the literature on professionalism, illustrating the view of occupations as driven primarily by self-interest and the need for power, status and material wealth and by altruism and the need to put the interests of the patient first (Saks, Reference Saks1995; Calnan et al., Reference Calnan, Silvester, Manley and Taylor–Gooby2000). This may account for the continued high level of public trust in clinicians, despite the change in the nature of trust relations between doctors and their patients (Calnan and Rowe, Reference Calnan and Rowe2008).

In this light, the analysis presented here − in recognising the artistry (Schön, Reference Schön1983), embodied learning (Nettleton et al., Reference Nettleton, Burrows and Watt2008) and craftsmanship of medicine − suggests that ‘to ask why societies incorporate their knowledge in professions is thus not only to ask why societies have specialised lifetime experts, but to ask why they place expertise in people rather than things or rules’ (Abbott, Reference Abbott1988: 323). Rules and mechanisms are clearly necessary, yet unable to account for, or ensure, the communicative aspects that are so crucial to positive patient outcomes. So whilst people are surely fallible, the increased application of a new form of late-modern trust – qualified and reflexive – may help attenuate the more significant risks of its alternatives.