I was inspired to write his blog by a tweet from @mikelombard earlier this week who was trying to respond to a question someone asked him, and he was trying to explain the difference between doing ‘tests of change’ and Toyota Kata (Rother, 2009). See his response in the image below. In addition, I led new a Kata learners/coaches session this week as did my friend Ann, and similar sense-making questions were asked.
I’m no expert on Toyota Kata, I’m still learning about it, with all its hidden depths and insights (see my 1st sketchnote below, sorry, I’m not an artist, just learning everyday!). Nor am I an expert on tests of change for that matter. But these questions made me curious enough to wonder how would I answer the question and suddenly, I really felt that I wanted to know the answer to the question ‘how does Kata differ from tests of change and doing experiments?’ It made me think ‘where is my threshold of knowledge on this, and can I shift that in an PDSA by writing this blog?’ So, here goes.
Toyota Kata* is made up of two main routines: the coaching kata (CK) and the improvement kata (IK), that are interdependent (IK/CK). The CK involves 5 primary questions that a coach asked a learner in a structured way to support the mutual achievement of a Challenge, moving from the actual condition through a series of target conditions. The learner uses the IK to set up experiments using PDSA to help remove or get around obstacles on the way to each target condition en route to achieving the challenge. The IK therefore uses the Plan Do Study Act (PDSA) approach within its routine.
Whereas, a tests of change approach, usually combined with the Model for Improvement indicates: first set an Aim by asking ‘what are we trying to achieve?’ (arguably akin to setting a Kata Challenge). Second, ask ‘how will we know?’, (ie the measures trying to be achieved) and then third asking ‘what changes can we make that will result in improvement?’ The changes are then chosen through many ways from the brainstormed list (often using a driver diagram or fishbone) to be developed for action, initially through small tests of change using PDSA.
Sounds really similar so far doesn’t it? Like maybe the IK and tests of change are much the same? Both are just PDSA with some dressing around them? I thought so too, about 6 months ago, when I first learnt a bit about Kata. I also thought, oh my, it’s only 5 questions what’s the fuss all about, this is really not hard. But now, I really don’t think so. For starters, it is decided deceptive and much harder that it seems at first glance. Knowledge is most definately not understanding. And in terms of similarity, rather than difference, I am beginning to consider that Kata is the next revision of our PDSA knowledge and understanding – an upgrade if you like, on our current theories and practices of how to do PDSA. In a way, PDSA v2.0 … or even Tests of change v2.0. It is already sometimes called lean 2.0 or lean’s missing link, bringing the crucial cultural and human side of change together with a systemised improvement approach.
Why am I beginning to think that?
Over the last few years, there have been repeated efforts to try to determine the evidence base for the nascent ‘improvement science’ field in healthcare (or however you might call it). This research has highlighted a lot of problems with PDSA. Dr Julie Reed and Dr Alan Card’s excellent BMJ Q&S ‘the problem with PDSA cycles’ article springs immediately to mind. This article, together with a number of others, show the weaknesses within the use of PDSA in healthcare. Such as, this study reviewing the evidence indicating that less than 20% of studies actually using all 4 PDSA steps with fidelity.
Other critiques suggest that few people attempt more than 1 test or cycle of change and there is a tendency towards ‘project-itis’. That only using PDSA or tests of change doesn’t build long lasting improvement capability. That, the P – ‘planning’ takes too long. That there are many soapbox uses of PDSA not really linked to big organisational needs and goals, wasting effort. That few studies show that the learning from the tests is then used. That often tests start small and even if successful, remain small never scaling into mainstream practice. That there is often no fact based way used to prioritise and select the changes to be tested. That PDSA leads to long lists of changes and arguably improvement overburden with so much going on that nothing is achieved. That gaps between test cycles are too long. That there is little to no follow up by leaders after tests (they have moved on to the next fire). And that PDSA is applied in an ad-hoc tool like and superficial way, isolated, without being part of mainstream leadership and team practices.
For me, this is where Kata really seems to be an improvement, for improvement practice. It seems to address some of those critiques of PDSA, in explicit and also some subtle ways. For example, the use of the CK with the IK, ‘tethering’ the coach and learner together, seems to me to make Kata a stronger leadership behaviour and habit. A kind of shared accountability and risk-taking yet risk-managing process. Supporting the embedding of daily improvement habits, behaviours and routines, that use scientific thinking in daily [responsive] social processes. And to develop those habits, behaviours and routines, kind of ‘meta-skills’ if you like; daily practice, leadership commitment and involvement is required, rather than long gaps. These routines help leaders in mainstreaming their role and behaviours in improvement practice – leadership standard work.
The daily practice also helps to reduce the improvement overburden by using the rapid, frequent experiments to get through/past/by the obstacles one by one, rather than 20 tests of change and action plans to report and evaluate some 30 days or so later. These Kata routines, at scale, to my mind has the potential to support development of organisational improvement capability and then the modification of organisational culture. The daily practice of the IK/CK and the 5th CK question [how quickly can we go and see what have learned from taking that step?] mean that the gap between experiments has to be short, and the Kata routines themselves, help to build good scientific thinking habits, likely to improve the fidelity in the use of PDSA and thus help us to make better and more sustainable improvements to patient care.
However, I am beginning to wonder if, as well as the radically different leadership role and involvement required, if the main difference, or rather enhancement to the test of change approach, by the Kata, is that of the target condition being only a couple of weeks or so away. Unlike the ‘Aim’, ‘Challenge’ or ‘knowing measures’, the ‘target condition’ changes every couple of weeks as the learner experiments and learns how to meet it on the way to the challenge. As the target condition does change over time, the pattern of work and even the measures can change, as obstacles are resolved and learning generated.
In that sense, for me, the Kata accepts the uncertainty of achieving a challenge and recognises that we may not ‘know’ at the beginning, even if we have ideas. Thus, the second Model for Improvement question – ‘how would we know if a change is an improvement?’ is quite different and possibly implies a differing view of uncertainty. This question is still about the Aim [in a Kata sense, this still relates to the Challenge] (I think) yet there is no smaller target condition, to help break down steps to get there, whilst learning as you go. Perhaps therefore, the substantial difference is that the 2nd Model for Improvement question assumes that you already know something, whereas Kata does not.
So, now I have thought about and written this, what have I learnt? Well perhaps the question ‘how does Kata differ from tests of change and doing experiments?’ has some preconceived biases, mindset and assumptions in it. I wonder is it really hiding subconscious questions and thoughts such as ‘is Kata is better?’, ‘don’t I already know this?’ and ‘I don’t need to go through the pain of unlearning about tests of change, I don’t have time’. It is hard to unlearn and let go.
I suppose for me, I have learnt maybe change my mindset, instead of really wanting to know the differences, I am now thinking, does it matter? Wouldn’t it be cool if I just want to learn how can Kata help me be a better improvement practitioner? Suspending that judgement. Isn’t that really my personal Challenge? I can use Kata to help me become better, building on what I already know about tests of change, the good and the bad. Then, I can experiment to keep moving my threshold of knowledge about improvement practice, enhancing it further everyday by building better habits for learning. So, that’s what I’m doing, and Kata routines are already helping me in my work to help healthcare fail quicker and learn faster, everyday, to make things better for staff and patients.
What’s my next step? Well, publishing this, and learning from the feedback…
*If Toyota Kata is new to you, I suggest you take a look at this video.