Article Text

Download PDFPDF

A critical look at Lean Thinking in healthcare
  1. T P Young1,
  2. S I McClean2
  1. 1
    School of Information Systems, Computing and Mathematics, Brunel University, Uxbridge, UK
  2. 2
    School of Computing and Information Engineering, Faculty of Computing and Engineering, University of Ulster at Coleraine, Coleraine, UK
  1. Professor T P Young, School of Information Systems, Computing and Mathematics, St John’s Building, Brunel University, Uxbridge UB8 3PH, UK; terry.young{at}


Background: With healthcare, Lean Thinking encounters a world, not devoid of value, but awash with sophisticated and mutually unconnected concepts of value.

Design: Given a shortage of systematic analysis in the literature, this paper provides a preliminary analysis of areas where the read-across from other sectors to healthcare is relatively well understood, based on a broad review of its impact on care delivery. It further proposes areas where conceptual development is needed. In particular, healthcare, with its many measures of value, presents an unusual challenge to the central Lean driver of value to the customer.

Conclusion: We conclude that there is scope for methodological development, perhaps by defining three themes associated with value—the operational, the clinical and the experiential.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

The last century saw revolutions in production and operations management (see Hopp and Spearman,1 for an historical overview, and Berwick23 and Locock4 for healthcare perspectives). Of these, Lean Thinking emerged with Taiichi Ohno at Toyota.58 Laursen et al9 describe how a broad family of quality improvement concepts impacted first on operations management, with service management following from around 1984, and medical management from around 1996. They further identify Lean Thinking as formally emerging in operations around 1992, in service around 1996, and in the medical arena in the early 2000s.

Lean offers five stages of improvement, founded on the concept of value to the customer, created by the producer, and defined “in terms of specific products with specific capability offered at specific prices through a dialogue with specific customers.”8Value stream mapping identifies waste in the end-to-end process (as activities that do not add value), while the concept of flow stipulates that products move smoothly from process to process without waiting or waste. Flow may be a physical matter—such as realigning machinery in a factory—as shown with the five Ss (sort, straighten/simplify, shine, standardise, sustain) system adopted by the NHS.10 However, it also embraces the idea that a system accepts each product variant as it comes along. Flow eliminates obvious waste (for instance, repeat activities or remedial work) and hidden waste (such as the cost of managing inventories, queues, delay or travel). Customer pull creates a new production dynamic away from batches and queues. Finally, the search for perfection is a reminder that Lean embraces a continuous-improvement mentality. With global healthcare expenditure soaring above $3.2 trillion,1112 and with healthcare systems increasingly challenged to deliver better care to more people using less resource, the quest to explore the promises of Lean Thinking is compelling.

The first half of this paper reviews the scene and notes the profound gap between the medical approach of randomised trials and other academic methodologies on the one hand, and that of the improvement methodologies on the other. The second half provides a preliminary analysis of the many views of value driving healthcare delivery and proposes a framework to consolidate the scene.


A short experiment using the keyword “NHS” on Google, followed by the phrase “Lean Thinking” using the “search within results” function revealed 565 hits (experiment performed January 19 2006). Searching within the results, with the string “NHS Trust,” reduced the hits to 239, and listed a wide range of newsletters, websites and other grey literature related to Lean Thinking in hospitals, ambulance services, PCTs, Strategic Health Authorities and other NHS networks. The Modernisation Agency proposed a set of 10 critical improvements,13 in which the impact of Lean Thinking can be seen. For instance, Change No. 5 (“Avoid unnecessary follow-ups”) is classical Lean elimination of waste to save an estimated £100 million in follow-up DNAs (Did Not Attend).

The literature, academic and grey, has grown considerably since the start of 2006. For instance, repeating the January 2006 experiment reported above on 17 October 2007 increased the hits to 37 100 (as opposed to 565) and the “NHS Trust” hits to 554 (as opposed to 239). Typical of the guides now available is that produced by Warwick Business School.14 Evidence of effectiveness generally takes the form of case studies. Spear,15 for instance, provides evidence of safer, higher-quality, more cost-effective care through a series of examples. The NHS Confederation commissioned a report,1617 with a set of exemplar improvements, while Ben-Tovim1819 reports from Flinders Medical Centre in Australia. Findings from these sources include the following:

  • delays from emergency care to surgery reduced to 1.7 days from 2.4 days and bureaucracy cut by 24%;

  • reduction in the number of steps in the pathology process from 309 to 57 and reduction in turnaround time from 24–30 h to 3 h;

  • a hospital halved the number of insurance-related safety incident reports over 3 years, and has moved healthily into profit without staff cuts on the Lean journey.

The practice of Lean and other manufacturing philosophies in the NHS was clarified when one of the authors (TY) was invited to share in two interviews by David Bensley, Operational Research Programme Manager at the Department of Health, who was leading a study into hospital improvement on behalf of the Department of Health and the NHS Modernisation Agency. Preceding interviews had investigated the extent to which hospitals had followed a rigorous philosophy, such as Lean Thinking. Two NHS process improvement experts, Dr Kate Silvester20 on 8 October 2004 and Paul Walley21 on 18 October 2004, were then interviewed. (Both wish to have it acknowledged that this was a retrospective analysis and that their thinking has continued to develop; moreover, they would stress the importance of a change of mindset in adopting such methods.) A picture emerged in which both saw that PDSA (plan, do, study, act) was the only improvement methodology that had bedded down to any extent in the NHS culture. PDSA is a traditional manufacturing improvement method in which new ideas are first planned (P) then implemented, or done (D), studied (S) and then applied (A), tested further or discarded, depending on the findings.2122

Interestingly, both improvement specialists articulated their own methods in terms of a central, strongly heuristic methodology—consistent with PDSA. However, when it came to deciding what measures to take in the next PDSA iteration, each would consider the next move from a variety of perspectives, using Lean, other philosophies, and trends showing up in statistical process control charts.23

In summary, there is evidence of widespread familiarity with Lean, and accumulating evidence of benefit when it is applied, especially in the areas of safety, delay and cost-effective delivery of care. However, there is evidence that even those advising clinicians and management would apply their own judgement when selecting Lean approaches alongside other approaches within a broader methodology, such as PDSA. Ironically, one might argue that this is exactly what Ohno6 originally advocated—but such a discussion is beyond the scope of this paper.

So, how has Lean been applied, and what is to be the prime driver of value?


Industrial and academic methods have produced very different corpuses of literature24 (p. 396). Hopp and Spearman1 (p. 42) capture something of this dichotomy between the industrial and commercial worlds: “It is apparent that business schools and corporations have swung far apart since the Ford and Carnegie studies of 1959 . . .” Qualitatively, the industrial improvement scene is characterised by champions who promote methods and promulgate success stories within an appealing intellectual framework, requiring buy-in and commitment to implement change within that framework. As Hill and Wilkinson25 observed with respect to TQM, “the original ‘gurus’ of quality management have been long on prescription but shorter on analysis and, moreover, have differed among themselves.”

The uptake of improvement methods into healthcare has not always been smooth. Blumenthal and Kilo26 reported on the US healthcare scene: “It remains too easy for health care organisations to talk a good game . . . and yet to leave their daily operations virtually unchanged.” Moreover, the messages of mixed methods being combined under a single banner, of rhetoric winning over reality, and a mixed message over impact, recur in that literature, too.2728

Many of these general observations resonate strongly with the healthcare scene we have just described, where there is strong evidence of the activity of champions, the role of success stories and the promotion of “how to” guides. The historical experience of improvement methods in industry and healthcare raises some concerns about the extent to which what is hailed as Lean is genuinely Lean in practice. Finally, we must recognise that there is a tension between the rigour of an academic approach (especially in terms of controlled trials) and the improvement ideal, which starts with what exists in real life and seeks to improve upon that within the context of everyday practice.

Our contention is that the improvement agenda has been driven by champions and by analogy with the industrial and service-sector experience. The intuitive link between value streams and patient pathways is compelling, and in fact, it is exactly this link that drove Ben-Tovim et al19 through the study in which they report their own, successful, improvement process. Moreover, examples of benefit are now common. However, it is more difficult to see exactly what has happened to the founding concept of value. Given that clinicians have developed many ways of measuring value-type concepts, such as quality, it is worth examining how the simple concept of “value to the customer” is faring in the world of healthcare.


That Lean can continue systematically and dramatically to improve a sector as significant as healthcare is still a matter of belief, rather than proof, and there are some obvious reasons why healthcare may differ crucially from other sectors. The staggering, global, scale and complexity of healthcare provision set healthcare naturally apart from manufacturing, perhaps from all other service sectors. Although dual lines of authority, clinical and managerial, need not be a unique factor, they are certainly a sensitive issue at present,29 but it is perhaps in the realm of the customer that differences are most obvious.

Shah and Robinson’s30 classification of medical device users shows several levels of users, including carers and healthcare workers, each exhibiting characteristics that might make them analogous to the customer. By extension, there are many people who might, at the same time and with the same particular patient in mind, have a role as customer for that product or service, and hold widely different views as to the value of that product or service. Alongside the personal “customers” are those who specify or procure31 attempting to buy “the best value for many services to achieve the maximum health gain for those most in need.”32 In the UK and US, such parties represent “an organised attempt by a private or public sector purchaser to ensure quality and to improve health outcomes.”33 Some communities have been considering value-type questions for a long time, and there are sizeable literatures around the sociology of healthcare and, of course, the business of defining and measuring clinical and healthcare quality.

A case study is offered to demonstrate the potential for clashes due to different concepts of quality. In May 2001, one of the authors (TY) ran a workshop on industrial methods in the NHS to explore how a cross-disciplinary team of managers, doctors and nurses might apply improvement thinking. A Lean Consultant was duly hired to gather information, part of which was obtained during an observation of a haematuria outpatient clinic on the afternoon of 24 February 2001. The patient pathway is captured in fig 1, and the Lean Consultant timed a few patients through the system (see table 1).

Figure 1 Pathway of movement of patients through a haematuria clinic from waiting on arrival1 to departure.12
Table 1 Summary of Lean Consultant’s observations related to patient movements around the clinic illustrated in fig 1

The clinic starts smoothly, but a queue builds as the clinic proceeds because patients wait to see the Consultant. Both the Consultant and the clinic Manager attended the workshop, and a critical contrast of views emerged as to how best to “Lean” the system, in this case for those patients diagnosed as not having cancer. The Manager, who came from a nursing background, took the view that the clinic had performed its purpose by this stage and that waiting time could be eliminated by giving patients the good news that they did not have cancer and sending them on their way. The Consultant, on the other hand, felt that the problem of diagnosis remained—even in the light of a negative finding for cancer—and therefore wanted to ensure a further consultation, even if this meant a queue building up. One wonders what choice patients might have made—and whether that choice might have changed with diagnosis.

The question of what steps to take to improve the clinic cannot be resolved until the question of value is resolved. Those who value smooth and fast throughput will lean one way, while those who value as complete a medical response as possible will lean another. From this case study, it is clear that there are at least two dimensions of patient-centred value, namely one based around the responsiveness of the system and another that addresses clinical priorities.


Once we move beyond the concept of value-to-a-single-customer, we encounter a bewildering array of value-concepts, reflected in a plethora of quality measures and frameworks. For instance, Ellis and Whittington34 propose nine dimensions of healthcare quality; the Dartmouth Clinical Improvement model35 appeals to a four-point Clinical Value Compass (functional status, cost, satisfaction and clinical audit); while Brown36 proposes a different set (access, effectiveness, safety and satisfaction); and Lohr and Harris-Wehling37 analyse services in terms of their ability to align knowledge with outcomes. Another community addresses the quality of life. The Barthel score, for example, may be used to assess the mobility of a patient on discharge from a hospital. It is an interesting example, since, while it is widely referred to, there is considerable debate about its validity,3840 and it is not clear how it fits alongside other measures.41 Measures related to patient experiences may even include patient-defined criteria.42

Some unification of these various concepts may be possible by considering value in terms of utility. The quality adjusted life year (QALY) is a convenient way of calibrating health states onto a single scale from 1 (perfect health) to 0 (worst imaginable health state, usually death), discounted over the appropriate period of health improvement or loss.4345 This measure can be used to compare the outcome improvements effected by quite different interventions, and critically the measure provides a means of economic evaluation. For our purposes, the fact that QALYs focus heavily on outcomes means that the experiential elements of a process may not be captured. For instance, Berwick and Weinstein46 discovered that women felt that as much as half the value of an ultrasound scan was to have the picture of their unborn baby—a benefit not traditionally reflected in QALYs.

Even this limited and rather arbitrary survey shows that there is no single customer with a simple view of value that can drive Lean Thinking in healthcare, but that the field is full of advanced views of value that have yet to be interconnected in a systematic way.

The first major attempt to link some clinical values with some other value-concepts in healthcare is reported by the Institute of Medicine,47 which has proposed 10 “rules” for the delivery of care. This endorses a multifaceted view of value, while suggesting that there is a manageable limit to the number of facets. What the rules do not do is to specify the trade-off between them when time, money, or access to other resources forces the case. Moreover, one might argue that there are some other, unspoken, desirable values, such as access or affordability that would feature highly for most people. The Institute of Medicine48 also proposed six criteria for quality assessment, namely that treatment be safe, effective, patient centred, timely, efficient and equitable. More recently, Gray48 suggests there are five important value perspectives to consider (the payer’s, the patient’s, the clinician’s, the manager’s and industry’s). Mainstream Lean literature is alert to these issues,49 but the healthcare paradigm is particularly difficult in this respect. However, there is some evidence that patients do better when they value the experience more.5051

The reason this matters is that most choices in healthcare service design include unintended outcomes. Until we have a connect-up view of value, it is impossible to know whether some aspect of waste-saving outweighs some other loss, or depth of diagnosis, for instance. It is critical, therefore, to find the minimum set of values required to proceed with Lean Thinking; to understand how this value set connects to existing value-concepts in healthcare; and to determine how value gains in one dimension equate to losses in another. It is possible that Lean is genuinely virtuous in all dimensions, but without a framework it will be impossible to know.

A proposal that is consistent with all of the above is that there are at least three critical dimensions to value:

  • Clinical: The prime clinical value is to achieve the best patient outcome (and is likely to be shared by clinicians and patients alike).

  • Operational: The prime operational value is likely to be the effectiveness of the service, measured primarily in terms of cost (including that which is lost through delay and poor quality). Clearly, this matters most to service providers and their managers, although clinicians and patients will also buy into this view of value. Our contention is that most, if not all, Lean Healthcare has been driven by an Operational view of value.

  • Experiential: Clearly patients value (or otherwise) their experiences of care, as will carers and those working in healthcare systems, either through their ability to empathise with patients, or in their own right.

Crossing from one theme to another—or attempting to trade-off benefit in one dimension with disbenefit in another—may be possible.52 The fact that the NHS53 in the UK will pay up to £30 000 per QALY may provide a boundary limit in converting clinical to operational values. Interestingly, Santry54 reports that even this limit may be too high, authorising the unaffordable. While reducing everything to money may seem rather crude, at least it enables one to envisage how the argument may be developed. Finally, we all manage to evaluate experiential elements of our lives, and, indeed, willingness to pay55 is a well-understood concept that links experiential to monetary values.

Pulling all this together under a patient-centred paradigm may, indeed, be possible, but it will require either patients to develop more, and specialised, knowledge in these fields, or careful attention to be paid to the way in which patients are protected by other stakeholders who come to represent them.


The way in which Lean Thinking is being adopted in healthcare appears to follow a trajectory consistent with the way in which other industrial methodologies have been taken up in other sectors, where uptake is mixed, and practice may be pragmatic rather than pure.

Having briefly reviewed the attempts to distil out a smaller set of common values, we propose that there are three key themes, or dimensions, to value in healthcare: the clinical, the operational, and the experiential. Moreover, since all three are subject to economic evaluation, it should, in principle, be possible to undertake a comparative analysis between benefits along these separate axes.

For Lean, however, the absence of a single customer with a compelling view of value is perhaps the most important feature of healthcare, and consideration of value within the many “customer” communities reveals a complex and fragmented scene. We contend that having demonstrated value in the sector, Lean must now engage with these many value concepts in a rigorous and, if possible, homogeneous fashion.


The authors wish to thank P Hines (Cardiff University), K Young (Warwick University) and N Proudlove (Manchester Business School) for helpful comments, and T Grocott (King’s College London), J Hobart (Derriford Hospital), R Lilford (Birmingham University) and the referees, for helpful pointers to the literature. They especially wish to thank D Bensley and the Department of Health for access to the Improvement Partnership for Hospitals Evaluation interviews. K Silvester and P Walley have also been very supportive in preparing this paper, and we thank them for their help. Finally, TY would like to thank M Feneley (University College London Hospitals) and G Sutton (Sutton Kaizen Consulting) for data and help with the haematuria example. L Steuten, with her excellent proofreading, along with J Eatock and C Weekes, with their help with the references, have provided invaluable support at Brunel University. The views expressed are entirely those of the authors.



  • Funding: This work has not been supported by any specific programme. However, much of this thinking has benefited from the involvement of TY in the MATCH Programme (EPSRC Grant GR/S29874/01) and SMcC and TY in the RIGHT proposal and programme (EPSRC Grant EP/E019900/01).

  • Competing interests: TY has received fees and research funding for work on process improvement, simulation and information technology in relation to healthcare.