We recently convened an expert roundtable to tackle how health systems, payers, and providers can spur the ‘de-adoption’ of medical practices and technologies no longer considered valuable. This got us thinking – while the process by which ineffective practices or technologies are abandoned is neither simple nor automatic, even the language used to describe it is not clear. And language matters. It often reflects an unstated focus on one mechanism or one level of decision-making. Here we review, and potentially clarify, the terminology.

About 35 years ago, systems theorist and LDI Senior Fellow John Kimberly coined the term exnovation: the removal of an innovation from an organization. Exnovation occurs when an organization divests itself of an innovation in which it had previously invested. Although the term wasn’t widely adopted, it gave us three insights that remain: first, that the innovation adoption cycle has a natural end, when previously adopted practices are abandoned to leave room for others; second, that the organizational processes and practices that support the use of the innovation are also critical factors in its abandonment; and third, that the concept differs from simply discontinuing use, in that it implies an active rejection of an established practice.

More recently, the editors of Implementation Science urged that more attention be paid to the science of de-implementation, which they define broadly as “stopping practices that are not evidence-based.” Prasad and Ioannidis call for applying the principles of evidence-based medicine to the task of de-implementation, while recognizing that “entrenched practices and other biases” can make evidence alone insufficient to accomplish the task.  Montini & Graham, echoing Kimberly, suggest that de-implementation involves intervening in the broader context in which clinicians work—the social, political, and economic realms—to change systems rather than trying to change individuals.

The term disinvestment has gained traction internationally, but its focus on a centralized budgeting process makes it less applicable to the U.S. health care system. Elshaug et al. define the concept as the partial or complete withdrawal of health resources from existing health care practices, procedures, technologies or pharmaceuticals that are deemed to deliver little or no health gain for their cost. Although it is often applied to decisions about pharmaceuticals, the concept of disinvestment—simply put, ceasing to pay for something—is equally relevant to other aspects of clinical practice.

But MacKean et al. note that disinvestment focuses attention on the cost component of a technology, and implies that the end-result is a foregone conclusion. They state that the language used will set the tone for engagement around this issue, and recommend the more neutral term health technology reassessment (HTR), which can have many outcomes, including no change in use, narrowing the scope of use, increasing scope of use, or stopping use all together. They define HTR as “a structured, evidence-based assessment of the clinical, social, ethical, and economic effects of a technology currently used in the healthcare system, to inform optimal use of that technology in comparison to its alternatives.”

MacKean et al. focus on the process by which policymakers can achieve the greatest benefit for the health care dollar, which grounds it firmly in the broader US conversation of reducing low-value care, with disinvestment of resources being one strategy to do so.  In a recent review, Niven et al. identified 43 different terms used in 109 articles to refer to the process of reducing or eliminating low-value care. The majority of citations (65%) referred to the process using more than one term, with a median of 3 terms. Disinvest* was the most frequently cited term (39%). Other commonly cited terms included decrease use (24 %), discontinu* (16%), abandon* (16%), reassess* (14 %), obsole* (12%), medical reversal (11%), and contradict* (10%). Terms such as de-implement* and de-adopt* were infrequently cited (4% and 3 %, respectively). They recommended that de-adoption be used to standardize the literature on low-value clinical care, despite its infrequent use to date.  

Gnjidic and Elshaug illustrated the 43 different terms in a word cloud:

As shown, disinvestment remains the most common term, especially as applied to the pharmaceutical reimbursement decision. Parkinson et al. recently reviewed the mechanisms by which policymakers disinvest in low-value drugs. Although they focus on drugs, their framework of active and passive disinvestment is helpful when applied more broadly to other aspects of health care that might be de-adopted. Active disinvestment methods include de-listing, restriction of treatments, price reductions, and the encouragement of generic prescribing. They note that active disinvestment is rare and difficult to know when to implement. Passive disinvestment, on the other hand, does not rely upon direct intervention from reimbursement policymakers, which might make it the most relevant term for the U.S. context. It occurs through market forces – for example, by shifting the price of a drug down after patent expiry. Other methods include changes in physician prescribing behavior, manufacturer withdrawal because of financial reasons, withdrawal of license by manufacturer or regulator because of safety concerns, and removal of a drug from professional treatment guidelines.

What to do with all this terminology? For the purposes of our roundtable, we chose to use the term de-adoption as the one least laden with implicit assumptions about the target, level, and methods to use in addressing the issue. Precision in language may be a small but necessary first step in agreeing on the best approach to reducing low-value clinical care.