Shedding Light on De-implementation
By Dr. Julia Moore, Executive Director
7-min read
Many people seem to struggle with implementing and scaling new things globally. Everyone, including staff and professionals, feels overworked, exhausted, burnt out, and not keen on “doing more.” As a result of this collective experience, there have been a lot of questions about de-implementation lately. People are exploring whether de-implementation could be used to move implementation initiatives forward, excited that it might not require “doing more.”
Our experience of de-implementation is based on working with people who are implementing things on the ground. We’ve noticed that many under-explored areas in de-implementation are not being discussed in the literature. Moreover, there’s a general feeling that de-implementation is qualitatively different from implementation, which hasn't been our experience.
When people talk about de-implementation, they could be describing 1 of 3 things.
De-implementation Category 1: Stopping ineffective or harmful practices
In literature, the term "de-implementation" refers to the process of discontinuing the use of practices or programs that are ineffective or harmful (here are two great papers to ground yourself in the de-implementation literature – one about defining de-implementation and another exploring de-implementation frameworks and models). These practices or programs may have been in use for years by professionals or clinicians, but new research has shown that they’re ineffective or can even cause harm. The primary objective of de-implementation is to stop using these ineffective practices. There are numerous examples of de-implementation in the literature, including the widely known Choosing Wisely campaign in healthcare and the push to discontinue D.A.R.E., a substance use prevention program in schools.
De-implementation Category 2: De-implementation as part of an implementation initiative
As part of larger implementation initiatives, it’s common for implementation teams to recognize that de-implementation is embedded in the initiative. The overall purpose of these initiatives is not to stop doing something but rather to improve current practices by doing something new. People may need to stop doing certain things or switch from doing something they once did. In our experience, this is the most common type of de-implementation happening. Though we don't always call this de-implementation, almost every implementation initiative has elements of it. We sometimes fail to acknowledge that asking people to do something new often requires them to stop doing something they previously did.
Here are two examples:
Years ago, we worked with a public health agency that wanted to scale a program that reduced over-prescription of antibiotics for urinary tract infections (UTIs) in long-term care homes. The initiative involved three new practices and two that people were asked to stop.
For a school-based example, we’ve seen people implement evidence-based programs focused on more positively reinforcing good behavior. While implementing these programs, teachers are asked to stop (or de-implement) punitive approaches when responding to undesirable student behavior.
De-implementation Category 3: Reflections on what people could stop doing
This is an aspect of de-implementation that we have virtually never seen in the implementation science literature. However, it’s becoming increasingly common in more general literature. For instance, the book "Subtract" by Leidy Klotz emphasizes the need to remove unnecessary elements to improve outcomes instead of adding more. This isn’t necessarily about reducing harm (typically the primary objective of de-implementation category 1), but more so about removing or reducing unnecessary things. Given the current state of the workforce, this seems like a type of de-implementation we should use more widely. Unfortunately, we haven’t seen examples of this type of de-implementation in implementation science literature. A few months ago, internally on our team, we asked our team members to share suggestions on things that we could stop doing, do less frequently, or do in a way that required fewer resources. Each team member was asked to come up with at least 2 suggestions, but everyone came up with more than 5. By implementing these suggestions, we immediately stopped doing almost 15 things and identified 8 things that we’re working on streamlining.
The parallels between implementation and de-implementation
From our perspective, the planning process for implementation and de-implementation is essentially the same. We first define the WHY, select the THING, define WHO is being asked to change, WHAT they are being asked to do differently, unpack their barriers and facilitators, and select appropriate change strategies. Then we build implementation teams and supports, assess the context and readiness, and plan for implementation, adaptations, sustainability, and implementation quality and outcomes.
However, we can improve our approach by more explicitly considering de-implementation. We can flag whether people are being asked to do something new, replace or change what they're currently doing, or stop doing something altogether. Each category may have unique barriers and facilitators — for example, barriers to stopping something may differ from those for doing something new. Understanding these differences can be especially helpful for people like champions who are often responsible for discussing the changes with professionals.
Considering de-implementation could enhance how we think about implementation and supporting change. We believe there are huge opportunities for the field of implementation science to more deeply unpack de-implementation and for implementation support practitioners to embed de-implementation into existing initiatives in various ways.
This article was featured in our monthly Implementation in Action bulletin! Want to receive our next issue? Subscribe here.