Problem-framing series. Blog 3/3: PDSA Cycles

This blog series is written as additional resource for an academic program where participants undertake an intervention project that falls within the scope of control, does not require additional resources and where cause and effect can be determined (i.e. ordered side of the Cynefin framework). 

Design your first iteration

This is the third blog in the series, and now we are moving to the nitty gritty of intervention. This content should feel familiar to those in healthcare, as I will be referring to the PDSA cycles. There are various other similar methods, and other names for PDSA cycles e.g. Deming or the Stewhardt cycle for Continuous learning, etc. The name is not as important as the principle of using iterative cycles of testing and experimentation, studying the impact, and adjusting if needed. 

I am often asked whether one can run more than on PDSA cycle at a time. In my experience, it depends. I know that is not a great answer, and even though it depends on the interventions, more importantly, it depends on the team’s capacity to remain on top of various experiments and still do their other work. This can be viewed as a learning cycle over and above the PDSA – testing what the team is comfortable with and capable of. Start with simple and small experiments that one can cycle through rapidly. Something else to experiment with is the mechanism of capturing the learnings. In other words, the team needs to not only wrap their heads around the PDSA, but they also need to figure out what is possible, how to capture learnings, etc. 

From a leadership perspective, PDSA cycles are not a desktop exercise, it is an operational exercise – it happens on the shop floor! If you are the owner of the cycle, you must be present and involved in the cycles. Often improvement is managed from afar, with someone checking the measures, and especially in a STEM world, quantitative measures might feel objective. I’m not saying that objective measures are not important, but if you are the owner of the PDSA cycle you want to observe what is happening, you want to know how the change or cycle is experienced and you need subjective information. 

The true value of PDSA’s is not the problem-solving, the intervention, or implementation. Rather the value lies in learning and understanding the current situation, the boundaries and the information about the system. From a curiosity and learning perspective, it thus makes sense to be as involved in the actual ‘experiments or testing’ as possible. 

This also enforces a mindset of viewing failure as an opportunity to learn, and it reveals boundaries in the system, the current capability and perhaps our thinking. Failures show us a way forward. The trick is to have a small enough intervention to be able to see whether it would have an impact, but also to be able to stop or abandon it if it is not working.

The PDSA cycle:

It might feel tempting to only pay attention to the Do part of the PDSA, however each aspect requires attention, and learning. PDSA cycles are a process.

PLAN

Plan the first experiment or action and develop the measurements.

A few questions to consider before intervention:

• Who will be responsible for the intervention?
• What will be done/what is the intervention? 
• Does everyone involved share a view of what the intervention is?
• What is in the scope of the project and what is out of scope?
• How will it be measured it? And when – be very specific. 
• What does ‘success’ look like?
• What does ‘failure’ look like? 
• If it goes well, can it be amplified? 
• If it doesn’t go well, can it be slowed it down? 

DO

Implement the experiment or action and measure as agreed upon in the planning phase.

Keep notes on aspects such as how easy/tricky it is to introduce the change. In other words, supplement the objective measures with subjective experience/views/observations.

During the huddle or team meet – chat about the implementation, the challenges and celebrate the small successes. 

STUDY

This is the analyses phase, interpret the information and the measurements. 

Also listen to what others say about the change. This is not a desktop and evaluation role. 

Decide what the measurements mean, and whether it is an improvement. 

Analyse any unintended impact or effects of the change. 

This is about observing, reflecting and learning.  Often we continue to act without learning or we take the next steps without pausing and just thinking about what the assumptions are, why this did or didn’t work. Test assumptions, ask naive questions. It doesn’t need to be protracted or take a massive analysis, just jot down some observations.

Also, consider – what should be happening, what is actually happening? 

Remember that the interventions are dependent on how the situation was framed, so if the team feels stuck, consider whether some assumptions needs to be updated and whether the problem or interventions needs to be reframed.

ACT

I like calling the next actions the 3A’s:

ADAPT: Either the next PLAN will be ADAPTED or changed according to the STUDY component.

AMPLIFY: Or it will be AMPLIFIED which means maybe it is made larger, or more people are pulled in. Thus same intervention but at larger scale in the next PDSA cycle.

ABANDON: Or it didn’t work, and there is no reason to continue with this intervention. The next PDSA will be a new experiment.

Conclusion

This concludes the 3 blogs on how to frame and get started with your intervention. I hope this was helpful.