Quality improvement in health care: is it mustard or ketchup?

Gladwell M. The ketchup conundrum. The New Yorker, 6 September, 2004. Reprinted in What the Dog Saw: And Other Adventures. London: Penguin, 2010.

Introduction

In this article, Malcolm Gladwell explains how there came to be many kinds of mustard in supermarkets but only one ketchup. He tells how an almost unheard of Dijon mustard called Grey Poupon stole market share from the dominant mild yellow variety. Importantly, its flavour meant an incredible proportion of people (in food marketing terms) would change brands after a single taste. Combined with a sophisticated ad campaign its success showed that tastes could change. Mustard didn’t have to be yellow. Now supermarket shelves are filled with different mustards.

Henry J Heinz entered the ketchup market in 1906. In the process of developing a ketchup without benzoate preservative, he made a condiment that was bitter and salty, sweet (lots of extra sugar) and sour (lots more vinegar), and by using ripe tomatoes increased the ‘umami’ or ‘body’ (the same effect mono-sodium glutamate provides). By 1907, Heinz was producing 12 m bottles of ketchup a year and exporting them around the world ¹

But, not everyone likes it. And there are other ketchups. Jim Wigon, inspired by Grey Poupon’s success, put great effort into creating the ‘World’s Best’ ketchup. It has more tomato, hand-chopped basil leaves and maple syrup instead of corn syrup. But he sold only 90 jars a day and didn’t draw a salary in five years. In the consumer mind ketchup isn’t ‘ketchup’ unless it’s Heinz.

What can this have to do with health services research and policy?

So, is health care quality improvement like mustard or ketchup?

In health care there are numerous quality improvement tools and techniques that can be used to support service improvements – stakeholder mapping, plan-do-study-act cycles, measurement for improvement, system/process ownership, project initiation documents, knowledge brokers, statistics, etc. My question is: are these different mustards where you choose the right one (or quality improvement tool) for the task, effectively a pick-n-mix? Or are they separate ingredients that need to be put together to make a quality improvement ketchup recipe, where you use the whole thing?

The case for mustard

If quality improvement tools are mustard, I would look at my situation, think about the tools and simply pick out the one I want. If I want to improve a process I use process mapping. If I want to increase stakeholder engagement I use stakeholder analysis. If I want to increase referrals I use measurement for improvement and chart the referral count.

I can mix my mustards if I want. There’s nothing to stop me using process mapping, stakeholder analysis and measurement for improvement in the same improvement project. I can spread them thick or thin. I can spend weeks or months on one technique. If it’s mustard, then the selection, approach and combinations are entirely my choice. It seems easy enough. Except that generally it doesn’t happen.

People don’t mix mustards. Frequently they don’t use mustard at all. Stereotypically, clinicians and managers have an idea to implement and they tell their teams ‘Let it be so. Add this to your busy day please. It’s good for patients and/or the organization’. An audit may be completed and reported at a Grand Round. Rarely will they map processes or stakeholders. Even good audits may not be prospective or use continuous measurement. They’re busy getting on with it. They may or may not be engaging with the right people or putting in efficient effort for a sustainable reward.

The case for ketchup

Now let’s say quality improvement tools are ingredients. To make ‘QI ketchup’ someone chooses which ingredients to put together in combination to make a single quality improvement toolkit. If it’s ‘ketchup’ then this toolkit has to be used in its entirety. The justification would be that this is the ‘right’ approach to maximize the chances of long-term success. There are already several different QI ‘ketchups’, including PRINCE2, Six Sigma, Lean Thinking, Lean Sigma, the Action Research Cycle. Each preaches (in its pure form) that all elements should be completed as prescribed by that school of improvement. All of them have been used to improve health care and reported in peer-reviewed publications.

However, are any of these ‘QI ketchups’ the ‘Heinz’ – the ketchup to beat all other ketchups? According to Gladwell, Heinz ketchup not only hits all five primal tastes (sweet, sour, salt, bitter and umami) but the blend and balance is good too. It has ‘high amplitude’. Heinz sells 650 million bottles of ketchup in over 140 countries each year. ²

I couldn’t tell you which QI ketchup is being used most in health care. None predominates to my knowledge. The lasting impact and spread of these approaches in their pure form in health care is hard to see. PRINCE2 is considered too cumbersome (except perhaps for IT innovations); better for construction. Lean works well in manufacturing but has a sustainability issue in health care organizations which are, anecdotally, not very lean. ³ Even conceptually Six Sigma seems an unlikely fit for the human complexities of health care. The development of Lean Six Sigma hardly seems a glowing endorsement for either system. As for the evidence, a recent systematic review of several QI ketchups (Six Sigma, Lean/Toyota Production System and Studer's Hardwiring Excellence) found only nine studies could be included and the majority had methodological limitations. ⁴

Lots of people in health care have heard of them and even been trained. They might have been involved in Kaizen events or in value stream mapping. But staff don’t routinely build these approaches into their daily work. They’re complicated and difficult. Integrating these approaches is just not the obvious thing to do. Health care staff are busy. They care but they’re really busy with lots of important and urgent priorities to deliver before something else can be added. And making change is difficult. Attempting change using an approach that is complicated and difficult is, in short, just never going to happen.

In Gladwell’s terms, the current QI ketchups seem to be like ‘World’s Best’ ketchup. The aficionados identified all the things that made ‘World’s Best’ different from Heinz: the extra tomato, the different sweetness, the lower salt and little vinegar. But they also identified that ‘World’s Best’ had really low amplitude. That cooked-tomatoey flavour formed an aftertaste. The blend and balance were not right. It was judged to be more of a sauce. I imagine that’s a pretty harsh criticism of a ketchup.

This seems to relate to what was suggested by Justin Keen: that high modernist ideologies such as evidence-based medicine and implementation science are attempting to impose scientific standardization onto automaton clinicians. ⁵ He proposes that the reason these things work at all is not their scientific evidence base but the common sense of clinicians who understand their complex, human working lives. My suggestion is that the improvement ‘ketchup’ approaches listed above could be perceived as being high modernist ideologies. If these well-known toolkits are not ‘Heinz’, there’s perhaps a gap in the market for the ultimate common-sensical health care quality improvement ketchup.

Could there be a QI Heinz?

To be a QI Heinz ketchup the approach has to be so intuitive and straightforward, that once you’ve tried it you never want your health care improvement fridge to be without it. That way, every time you reach for an improvement, your QI ketchup is right there waiting for you. It’s an approach where you have just the perfect minimum amount of aim agreement and goal setting, an intuitive process mapping approach, the optimal amount of stakeholder analysis and engagement, in collaboration with patients and the public.

As with any condiment, there’s nothing to stop extras being added in your local context. If you need more salty process mapping, add some. If you need to spend more time on measurement or re-defining your aim you have a whole spice rack at your disposal; optional extras, to locally enhance your initial basic mix.

What you don’t need is a cheap imitation in a Heinz bottle. It’ll be likely to leave a slightly bitter aftertaste and taste tests would show it has low amplitude. It just won’t grab you. It won’t achieve sustainability in your QI fridge. Equally, anything too expensive might be perfect for the occasional treat but again, won’t suit the mass market. And health care improvement needs to be a mass market.

Policymakers don’t currently seem to be considering systematic and sustainable approaches to improvement. Policy seems to be focussing on solutions – improvements that must be made quickly – and they’d better save money.

Some people might say there are different correct approaches for different settings. Gladwell talks about Moskowitz identifying the ‘plural nature of perfection’ and how there are now 36 varieties of Ragu spaghetti sauce. But really, who wants to have to work out which subtly different version of spaghetti sauce you should use when you want to improve your health service. I prefer ketchup: Gladwell’s universal option (citing Elizabeth Rozin) – part of the collective unconscious – with an “unprecedented ability to provide something for everyone.” But then again, maybe it’s all just mustard.