Hamish Dibley

Home » Improvement

Category Archives: Improvement

Advertisements

Why Conventional Change has to Change: A Better Way to Improve… Part 2

Albert EinsteinYesterday I outlined the problems inherent within conventional approaches and thinking about change management. All too often conventional approaches result in no or little positive change for either service users or taxpayers. Albert Einstein was right: to solve problems and improve our thinking has to change. Indeed, real improvement comes through learning by doing and reflecting on these results. It must also be rooted in research that provides data-driven evidence.

Essential to the acquisition of hard-data is knowledge of customer demand that illustrates patterns of need set against the deployment of resources. Good service (as experienced by the customer/patient/service user) always pays, it never costs. The inability to deliver service is paradoxically expensive and the real ‘gold-plating’: it means in the public sector patients or service users receive either too little and then either ‘represent’ to the service or (in the private sector) customers leave altogether.

Conversely, reviewing services from the customer/patient/service user perspective reveals the folly of conventional change management with its preoccupations on activity assumptions, arbitrary plans and make-believe monetary savings which rarely materialise. Improvement begins with adopting a customer mind-set, studying services and putting pre-configured assumptions of what works aside. Use innovative methods to understand how and why services perform the way they do and what it takes to get them better for those that need to use them.

A Better Way to Conceive Improvement

Intelligent change starts with turning convention on its side, researching performance of local NHS and care systems through the external lens of the patient. The way to realise better service and less cost begins with taking a different, horizontal view. Patients flow through healthcare systems.

Adopting a ‘Front-to-Back Thinking’™ mind-set is about recognising that activity and costs are a consequence of what you do not why you do something. Performance improvement comes from studying patient-level demand; analysing the nature of activity that this demand generates and achieving sustainable cost savings through the elimination of activity waste. How well an organisational system meets patient needs should inform both strategy and costs.

Studying patient demand allows for intelligent system and service redesign solutions around cohorts of patients, not pathways. The principal work is two-fold: research and redesign. The former is evidence-informed problem identification through data analysis.

Research is undertaken to understand patient-demand by looking at the ‘what, where, who, why and when’ of their healthcare usage. It is important to recognise here that the requirement is to utilise both quantitative and qualitative research techniques. A common error with many analytical models is to draw linkages between correlation and causation or indeed assert causality as a consequence of data analysis. Data analysis asks ‘what’ questions which need to be linked to demand analysis’s pursuit of ‘why’ to authenticate findings.

This phase entails analysis of time-series consumption and case-mix data; encounter data to understand customer activity and patterns of usage as well as the type, frequency, volumes and predictability of customer demand together with an understanding as the balance in value versus non-value activity. The work provides patient segmentation into meaningful typology groups which forms the foundation for subsequent improvement work.

Redesign implies a set of sequential steps to undertake intelligent improvement activity. Such Improvement must focus on redesigning the care models and systems around patient cohorts, beginning with the ‘vital few’ patients who, although small in number, consume disproportionately large amounts and activity and costs.

It involves using knowledge gleaned in the research phase to undertake prototype service redesign activity. Clarity of the patient purpose is paramount; as is patient-centred performance metrics; redesigned systems, processes and roles; experimentation with operating models and continuous feedback loops to ‘learn to improve and improve to learn’.

Teams of interdisciplinary professionals are assembled and given autonomy to work together first understand and then meet the holistic (medical and non-medical) needs of these patients, who themselves set the boundary. Flexible services and systems are harmonised according to patient need. The role of leaders also changes, from mandatory monitoring in board-level meetings to problem-solving issues that beyond the control and scope of local interdisciplinary teams.

Key skills that are required include medical and technical but extend more importantly to interpersonal, organisational and problem-solving abilities. The focus is on decreasing end-to-end service times, ensuring more work is done right-first-time; reducing or removing activity that adds no value to the patient thereby preventing such patients from boomeranging around the system.

Here the work involves alterations to budgets, roles, measures and, where necessary, technology. Importantly, the purpose is to develop the redesign and determine its anticipated economies through internal base-lining of current performance via patient-level not service-line reporting.

Following the prototype period, leaders make an informed choice about the benefits from adopting the new (systems) design including further roll-out opportunities. This approach is encapsulated in the Consumption Demand Method™ which enables the emergence of better services that act as testing grounds for continuous improvement set against a better understanding of patient demand.

Advertisements

Why Conventional Change has to Change: A Better Way to Improve… Part 1

The No Change ParadoxAchieving great service is straightforward if unconventional: give customers/patients/service users what they need. However, convention dictates that to adopt this approach will lead to expensive ‘gold-plating’ of services (quality service but at higher costs).

Instead leaders and organisations seek to follow convention and manage their activities by protecting budgets, imposing access restrictions via criteria or eligibility introducing service level agreements and focusing on efficiency through reducing transactional or unit costs.

Yet, the real paradox is that an explicit focus on managing activities paradoxically increases the very thing organisations seek to reduce – cost. Convention decrees that the answers to problems are already known and pre-prescribed solutions can be delivered through meticulous plans and reports. As a consequence these change approaches fail to deliver in practice.

Take conventional healthcare commissioning in the NHS – a person as a health and/or social care need; their need is assesses and then sooner or (most often) later a service is first commissioned then provided by different professionals.  Service level agreement met, project milestone ‘green-lighted’ and ticked.

What happens is because services are not designed around the need(s) of the person/patient/service user, they represent to the service in the vain hope that their need(s) maybe better met. Professionally, the response to this problem is to repeat the process of assess, commission and provide. The outcome experience of the person/patient/service user is that they continue to represent and costs rise. Why does this happen? It is because the mind-set is misguided.

Conventional business change needs to change

Conventional change or improvement relies on a wrong-headed ‘back-to-front’ perspective. What is meant by this phrase is a rear-guard focus on removing cost by reducing activity and expecting patients to change their behaviour. It results in an obsession with activity volumes and ‘bottom-line’ costs. The one problem with this approach is that it doesn’t work. Back-to-front thinking always leads to distortion of organisational performance and higher costs.

The mechanics of conventional change follow a typical path. Invalidated hunches, opinions and/or data consisting of worthless aggregated activity, arbitrary benchmarking and/or cost volumes are used to identify problems.

Understanding the patient or service user (as opposed to non-user public) perspective in this process is rarely, if ever sought. Agreeing appropriate governance arrangements typically loom large at this point and take-up not inconsiderable internal discussion, effort and time.

Much time is consumed completing a litany of project management induced paper-chasing reports such as project initiation documents or PIDs. Once signed-off this document helps formulate a project plan which is established to solve the preconceived problem. A business case is then written which outlines time, costs and predetermined outcomes.

These outcomes are then ‘monitored’ through office-based completion of reporting documents such as highlight reports full with activity and cost volumes with ‘traffic light’ systems – green for good, amber for somewhere in between and red for bad. None of this involves spending time in the work and empirically understanding ‘why’ things are the way they are.

Improvement activity is often relegated to time-limited projects and people who sit outside of the actually work. Disproportionate time and effort is then spent on conducting public consultations (‘the blind leading the blind’) which replace the opportunity to generate empirical knowledge of what is actually happening and causing problems at the sharp-end, in the work.

Performance metrics derived at the business case stage tend only to measure activity on whether a project is completed ‘on time’ and ‘to cost’. Little regard is made of how much operational improvement is achieved in resolving the real problem(s).

‘Off-the-shelf’, standardised solutions are mandated that usually involve automation or greater use of technology (sometimes referred to as ‘channel shift’ or ‘digital-by-default’); sharing or outsourcing services; restructuring to establish new ‘target operating models’; rationing buildings and service provision; charging and trading services and/or reducing staff numbers. Here abstract cost-benefit equations are the order of the day.

Unless consultants are engaged, ‘delivery’, ‘execution’ or ‘implementation’ is then solely contracted out to frontline staff who are left to try and make the problem fit the predetermined standardised solution(s). Benefits realisation plans are written remotely and focus on the completion of activity tasks not performance improvement.

Consequently, ‘change’ seeks to solve symptoms not address root-causes. The predictable result of this way of approaching improvement are failing projects, higher costs and poorer service as experienced by patients.

Tomorrow I will outline the more intelligent way to conceive of how to change for better and undertake meaningful performance improvement.

Delaying the inevitable: a better way to think about delayed transfers of care

DTOCDelayed transfers of care, or ‘DTOC’ as it is colloquially referred to in NHS circles, is an entrenched national problem. It is often claimed that a combination of ‘cultural and structural factors’, an over-reliance in bed-based care and increasing numbers of people with complex conditions moving between providers is causing undue strain on the system. It is one of the principal issues said to affect the performance of A&E units up and down the country.

DTOC has a big impact across the whole healthcare economy in terms of activity, costs and reputation. It is generally accepted to centre on acute services for older people. But the question I pose is: by focusing on the acute are we concentrating efforts on the symptom or the cause? Every problem presents an opportunity. Every challenge requires an intention to signal a desire to improve through thinking and acting differently. DTOC is a symptom of a dysfunctional health care economy.

Concerted effort and numerous attempts have been made, over several years, to successfully address DTOC. These initiatives include the usual list of initiatives including ‘Appropriate Care for Everyone Programmes’ (without ever defining what appropriate in this context means), QIPP projects, engagement and public education packages and Older People’s Joint Commissioning Strategies.

Yet, successive schemes fail to impact on DTOC as they compartmentalise the problem through focusing on urgent care or complex care rather than understanding the root cause of the problems. The impact of these actions on improving DTOC performance remains stubbornly low. Initiatives tend to adopt a reactive and reductionist approach to resolving the problem.

The principal reasons for failure requires detailed study but can be summarised by adopting a (silo) service-first not patient-perspective and focusing on activity reduction and discharge pathways. The causes of waste or ‘system limitations’ affecting DTOC include:

  • Outsourcing of adult social care provision
  • Multiple patient assessments
  • Eligibility criteria in community and social care
  • Different organisational budgets, performance metrics and ways of working
  • Counter-productive financial incentives across the system
  • Service fragmentation such as a ‘medically fit’ focus
  • Alleged risk-averse attitudes amongst some clinical staff
  • An efficiency rather than effectiveness focus on reducing length of stay
  • A target mentality which limits attempts to understand in order to improve
  • Understanding the problem from the organisational perspective

‘Pooled budgets’ for services used by older people offers hope to reduce rates but only if it is accompanied by a change in commissioning and operational thinking and ways of working that extends beyond joint ‘commissioning of care’.

‘Current DTOC performance’ is monitored on a regular basis through weekly returns taken on one day each week from local acute trusts. The whole thing resembles the game-show Play Your Cards Right. Snapshot analysis shows the weekly rates of DTOC can vary with some weeks being ‘higher’ and others ‘lower’.

All of which leaves me with more questions than answers, and those questions are – what does this tell us? How does this information help us understand to improve? Does this ‘activity counting’ help develop knowledge and understanding around the patients who experience delays and the type and nature of activity this creates which leads to costs? What has operationally changed as a result of monitoring? How do we know if this has been effective? Or like the game show – is it simply a game of chance? Such ‘activity monitoring’ is as counter-productive as it is distorting. Improvement work must take place in the work not in meeting rooms.

I have recently completed a piece of work looking at the problem. I hope the study will provide the platform to overcome successive false starts, identify root causes and design a healthcare system to reduce DTOC levels through better understanding patient demand. But that, of course, is dependent on the willingness of local NHS leaders to adopt a different mindset.

It begins with adopting a different patient-centred perspective and intelligently analysing the problem. This work seeks to take a holistic approach and understand it from the patient perspective. It asks how many patients cause what type, patterns, predictability and volume of activity which results in costs. This way of thinking and acting is the means to achieve intelligent change leading to sustainable performance improvement.

What you discover is that the DTOC problem is NOT a general acute service for older people problem; hospitals deal with the consequences of the issue. When you compare A&E waiting time against DTOC rates it reveals different sets of challenges. One lot are external system problems where the acute provider is trying to deal with the consequences of fragmented care. But there are also internal system issues such as the impact of the 4-hour waiting time target which distorts performance and paradoxically helps make the problem worse (see my previous blog on the deficiencies with the A&E waiting time target).

As ever with healthcare, DTOC is a ‘vital few’ challenge – small numbers of patients consuming disproportionately high levels of activity, capacity and resources. For example, 1,700 patients admitted suffered with a DTOC over a 12 month period. The length of stay from their emergency admissions contributed to a third of the overall bed capacity of the hospital. Moreover, the real costs of DTOC are much higher than simply a blinkered focus on ‘excess bed days’ would allow for (an abstract and unhelpful financial ‘tariff’ distinction). Less than 2,000 people cost this particular secondary care system over £11 million every year. How much of this money is well spent? We simply don’t know.

But the small numbers show that the problem is predictable and therefore manageable.  It also represents a big opportunity providing healthcare leaders have the courage to make profound changes in the way we commission and provide healthcare services. To improve (and thereby reduce) DTOC there is a need to better understand patients and what services they already consume to design more effective medical and non-medical services around their care needs.

The issues affecting DTOC are complex. The mistake often made with improvement efforts is to try and cut through or standardise complexity. Wrong. As W Ross Ashby’s Law of Requisite Variety teaches us the most intelligent way to deal with variety is design systems and services against that variety. In healthcare the way to undertake this is to understand and then design against patient-level demand. This will be effective and if you think about it, if you are effective at something; by definition you will be efficient.

Lop-sided logic: A&E and the 4-hour waiting time target

A&E Performance and the 4 Hour Target

Broadcast and newspaper headlines in the past couple of months have all been about pressures on A&E services. As ever with whirlwind media storms they typically tend to blow out before reasoned conclusions can be drawn.

Which leaves the question, what really does bedevil A&E? Is the problem facing the front-end of acute hospitals down to underfunding of emergency medicine; the ‘downgrading’ or closure of A&E units; fragmented and unresponsive primary care services; NHS 111 (more on this in a future blog); the payment or ‘tariff’ system; unprecedented patient demand; blockages including static or declining bed capacity and delays in transferring patients out of hospital or all of the above? 

The direction of NHS England is to ‘centralise’ or ‘standardise’ emergency provision around fewer locations and downgrading others into Minor Injury Units (MIU) or Urgent Care Centres (UCC). To critics this move is one which creates a ‘two-tier’ system. The logic is an economy of scale one and follows where earlier efforts in stroke and heart care have lead; namely centralise medical expertise and services in order to improve outcomes.

Two things are for sure – we shouldn’t expect ‘healthcare professionals’, code in this case for paramedics, doctors and nurses to work ‘harder’, they already are. Nor should we condemn people for using A&E services inappropriately in whatever numbers. Clifford Mann, the President of the Royal College of Emergency Medicine was absolutely right when he said last month: “I don’t think we should blame people for going to the emergency department when we (the system) told them to go there. It’s absurd.”

It is true that over time, we have both incrementally and intentionally designed a reactive, fire-fighting, hospital focused, medical health system, with care very much an after-thought. Hospitals declaring they can’t cope through issuing ‘major incident’ alerts is the predictable consequence of some pretty foolhardy thinking. One such example is the 4-hour waiting time target in A&E. Ludicrously this activity measure is regarded by many as both a good indicator of A&E performance and quality. It is nothing of the sort.

The 4-hour waiting time target is a nationally-derived arbitrary indicator. The NHS in England (or more appropriately individual NHS Trusts) has missed its four-hour A&E waiting time target with performance dropping to its lowest level for a decade. Figures show that from October to December 2014, 92.6% of patients were seen in four hours – below the 95% target. The performance is the worst quarterly result since the target was introduced in 2004. Viewed a different way, 9 out of 10 people who go to A&E get ‘seen and treated’, discharged or admitted within the 4-hour window. Interestingly, whether we actually solved their problem is not measured and consequently never recorded.

The waiting time target’s reductionist logic is simple. It is an arbitrary indicator which is set from outside the immediate organisation (hospital) or business unit (A&E). The hospital and A&E is then expected to achieve the target come hell or high-water, through greater effort by its employees. 

The reason this happens is because there is a mistaken belief that it is people themselves who are the limiting constraint on performance within organisations – they need to work harder or refrain from using services for reasons that are deemed ‘inappropriate’. The reality is that people’s performance or usage of a particular service is a consequence of the influence of other parts of the system, of which they form a part – policies, processes, procedures, systems, management. 

To use an analogy, if a person enters a 10 mile race, in the knowledge that they can only run 5, the only way they can finish the race is to ‘game’ or ‘cheat’ and catch a bus after the 5 mile mark. This is what happens in organisations. If the A&E arbitrary target for transferring patients to a ward is 4 hours, but the existing capability (if measured) is only 5, then in effect, the bus is a trolley parked in the hospital corridor. Only now it is alleged, the wheels are coming off. 

The other problem with arbitrary targets is that they are self-limiting. With no knowledge of patient demand and existing capability to meet that demand, the artificial target actually could be well within the range of what is improvement is possible. Paradoxically, setting a target can actually lead to under-achievement, in both people and organisations. 

To compound the problem, along with the deficiencies inherent in setting an arbitrary target, measuring and judging performance at a single static point or over a single period of time is also counter-productive. This ignores the nature of variation that exists in almost everything we do, individually and collectively within an organisation. Here averages distort unless viewed within the context of the overall distribution of performance and the underlying trend, viewed against the demand placed on the system at any given point in time. 

Whilst we can’t remove the 4-hour waiting time (the real limiting constraint), we should treat it as unavoidable system limitation but not drive performance solely on achievement of the arbitrary number. The alternative ‘solution’ is very straightforward – establish more insightful information streams and/or make better use of data in a more operationally meaningful way. Asking useful questions will help understand and in-turn resolve problems: 

  • Before measuring anything, ensure you are measuring the right thing. Is the 4-hour target measuring the right thing? Measure what matters to patients – do we actually know what this is? Is it quick diagnosis, speedy treatment, medical or psychological reassurance or getting help?
  • When it comes to measurement we first have to understand patient demand. Do we empirically know why people choose to use A&E (patient demands) in order to understand how A&E can be better designed to deal with these demands (capability)? Simply ranking demands by their perceived inappropriateness without understanding the patient context doesn’t solve any problem.
  • Measure your existing capability over time and understand the statistical variation that exists, against an understanding of demand.
  • Express what you measure statistically, based on the nature of the distribution of the thing being measured.
  • Unless you intend to change something and if you must set a target, understand what is within your current organisational or business unit capability.
  • If you want to set a target outside your current capability, then identify what you are going to change to achieve this objective.
  • Do not let the setting of a target act as a system constraint in itself.
  • Do not make a business out of it – the aim should be continuous improvement not ranking.

This alternative approach would require us to regularly understand and measure local demand placed on the system (find out from their perspective why people actually come to A&E, don’t assume to know the answers) and the local capability to respond to this demand. Is demand predictable over time? What is the current system capability (staff mix, capability to meet the nature of patient needs and resourcing) to successfully address this demand?

A sophisticated understanding of patient demand and the capability to meet it will provide providers and commissioners with the ‘business intelligence’ they need to have a more effective A&E service. It informs us as to the level and nature of professional expertise required in A&E and when; availability of appropriate test facilities and beds; person-centred processes to effectively meet needs and even how best to approach the design and layout of A&E.

One thing we can forget though: the growing trend to rebrand A&E Emergency Department (ED) – that has zero impact.