A foundational barrier to process improvement, whatever the approach, is that it generally offers a moral wound up front, with a promise of suturing, healing, and better things for the distant future. Appreciative Inquiry is an approach to qualitative research that offers good news up front, followed by the promise of even better in the future. This may increase adoption by administrators and clinical staff alike who may otherwise be reluctant to risk the initial shock of bad findings.


Process improvement initiatives typically start by finding (almost) all your faults, categorizing them, underlining them, and specifying in great and gory detail how they might imperil safety, timeliness, effectiveness, and efficiency. As a bonus, they typically also detail in what ways you fail to be equitable and patient-centered, and how badly you fail to be accessible. This is all good though, because it then moves on to outlining what may fix these gremlins, and how to test that they are fixed.

So, first, a poke in the eye and a moral wound left open; then, after a delay, a process to suture the wound, and finally a promise of full healing.

Most administrators, and not a few clinical department heads, are more than a little put- off by the typical process improvement experience. They (somewhat rightly) fear that detailing all the shortcomings may be a very painful experience and may easily get into press and become a public embarrassment (at best)—and maybe even result in costly lawsuits. The promise of a future state in which things are much better is cold comfort to those more oriented to be in the moment, so to speak. A here-and-now detail of flaws looms like a certainty over the vague and hazy outline of a better future. As a result, many process improvement initiatives fail to get the needed administration and leadership support and wither after an initial buzz of best intent. Additionally, organizations tend to go toward what they focus on. If the focus is on problems, that usually creates even more problems.


Appreciative Inquiry

Appreciative Inquiry (AI) comes at process improvement like a big warm hug from a slightly batty but beloved relative. Lots of cooing and endearment, highly fragranced embraces, and all soft edges and no prickly bits.

AI was pioneered in the 1980s by David Cooperrider and Suresh Srivastva, two professors at the Weatherhead School of Management at Case Western Reserve University. The general idea is not new and resurrects a past “positive outlier” or “positive deviance” focus. The use of positive deviance has always been a solid evidence-based improvement on traditional quality improvement methodologies—which, to be frank, often became very mystical and shouty about goals. Setting improvement goals was often more about sucking a “stretch goal” out of a fundamental orifice than any real insight. Mean length of stay of 4 days? Halve it! Door to Doc time of 2hrs? Do it in 20 min! 30d readmission rate of 15%? Make it 5%! If anyone questioned the targets, they might be stared down and marked as “difficult,” but there was a lingering doubt that the targets were based on reality.

Positive outliers, however, are the best you had ever done for a specific metric, and the question became what needed to be changed to shift that from being a rare outlier to becoming the norm. Still not a perfect approach, but at least based on the local reality; if you had achieved it at least once before, it wasn’t all that unreasonable to think that with some confluence of factors, the feat might be repeated.

AI takes that approach and shifts the focus from strictly metrics to the people actually doing the work, and in a warm embrace kind of way, it solicits feedback from the front line on best instances. It often expands that to seeking what was good for the person and the best that person could bring forward (Figure 1).


Appreciative Inquiry & COVID-19

For a study we are just starting on COVID-19, we are asking the clinicians involved the following questions regarding their experiences of the COVID-19 response at their facility:

  • What energized you or gave you a sense of satisfaction about your work or your COVID-19 response?
  • What did you do that resulted in a highly successful or unexpectedly positive outcome?
  • What team strengths, capabilities, processes, or tools contributed to your efficiency or effectiveness?
  • What particular strengths can you focus on more as a group, for similar positive results?

This positions and contextualizes the interviews and data collection in a specific framework—not just “Hey, what’s great today?” At the same time, the analysis is not so tightly confined that we miss essential parts of what makes things work well. In this construct, we are looking specifically for what energizes people and brings a sense of satisfaction, as well as what was going on when they hit the metrics or experienced something positive. The approach also importantly looks at the team dynamics and context.  Few things in medicine are the result of a single actor, and we want to know what one part of the system positively influenced another area. For example, if something the bed-cleaning crew did resulted in the clinical team achieving something great, that would be an area of focus.

In this example, the approach will be used to learn from the COVID-19 response, and to help improve later and future outbreak responses. The first phase will recruit people involved in the COVID-19 response to record a daily diary for the four questions. In the second phase, the study team will interview those people to get greater depth and context. In the final phase, the study team will analyze the data, provide a report on the themes, and generate a response typology of the things that worked best. This might build on the standard root cause typology, such as in Figure 2, or build an entirely new one using a grounded theory approach that lets the data dictate the categories.

There are three anticipated benefits of this approach:

  1. Being asked about good things is a FAR more positive experience than what it typically feels like to be interviewed about quality and safety deficits. Typically, the person being interviewed about root causes of missed metrics has a highly negative experience and leaves it feeling drained and exhausted. The typical experience of the AI method is a sense of upliftment, teamwork, purpose.
  2. Instead of just finding causes of deficits between performance and metrics, the shift in focus to positive outliers typically results in generative solutions that do better than the norm rather than just meeting it.
  3. The constructivist “good news” approach makes it far less of a moral wound and generates many more memorable positives than risks, issues, and gaps that might have been viewed as threats or embarrassments. The net take-away for chiefs and administrators is that this is safer and more beneficial, while interviewees find it gentler, more uplifting, and actually enjoyable.

The downsides to this approach are real enough, and have more to do with trust and group dynamics than the methodology itself:

  1. Focus on the positive may be seen as “ignoring the negative” or “astroturfing” where there is a low-trust environment, such as when staff have come to be suspicious of management motives.
  2. There may be reluctance to focus on the positive out of a fear that this ignores real risks. This is a true hurdle, and when clear gaps or deficits are encountered, they should be reviewed against the positive themes that emerge. If no positive theme addresses the deficit, then regular root cause analysis into the deficit should be carried out.
  3. There is a risk of perpetuating siloed thinking. What is “best ever” for one department or team may in fact be highly deleterious for another. To avoid this, positives should be reviewed in terms of the end-to-end value chain in which they occur, and their relationship to the end goal.


Using AI as an approach to improving the response to COVID-19 brings more of an evidence-based rigor to goal selection and does so in a manner that is kinder to the study participants, reduces the sense of risk for administration and potential sponsors, and potentially develops solutions that are generative rather than focused on deficits.