Abstract
In this paper, we discuss de-implementation as an implicit part of implementation and organizational change, and consider its underlying processes of unlearning to discontinue or deviate from ineffective practice and learning to applying newer, more effective practices. We describe a typology of de-implementation that represents four types of change: partial reduction, complete reversal, substitution with related replacement and substitution with unrelated replacement of existing practice. We also explicate how learning and unlearning needed for effective change vary in these four types of de-implementation. Last, we propose coupling de-implementation and implementation efforts, which serve conceptual and logistical goals of organizational change.
Keywords: De-implementation, Organizational change, Learning, Unlearning
1. Introduction
De-implementation, or the abandonment of an outmoded or disproven clinical practice, has been likened to the mirror image of implementation of new, evidence-based clinical practices.1 In this view, the de-implementation process can be understood as the implementation process in reverse. We propose that the processes labeled de-implementation actually represents four types of change, each entailing different challenges and dynamics. As result, we contend that the mirror analogy is not necessarily true, as de-implementation and implementation processes may require different strategies. However, it may be useful to jointly consider implementation and de-implementation because of the unlearning that inherently occurs when individuals or organizations learn new practices, discontinue existing but ineffective practices, or both. A promising focus for research is coupling de-implementation with implementation strategies and considering de-implementation effects of implementation strategies.
1.1. Four types of change involving de-implementation
There are four different types of change in which providers cease outmoded or disproven clinical practices. The first type of change is reduction or partial reversal in the frequency, breadth or scale of an existing outmoded intervention, so that it is provided to only a subgroup of patients who have been demonstrated to realize the greatest benefit. Examples include initiating breast cancer screening at age 50 rather than age 40 for women at average risk,2 lengthening the interval of cervical cancer screening with cytology (Pap smear) from one to three years,3 or de-intensifying diabetic medications in older adults with well-controlled diabetes.4 Reduction initiatives may decrease providers’ clinical effort by requiring continued effort on only a subset of patients and less effort in the subset of patients who were not benefitting from the discontinued treatment. It is possible that there will be new clinical effort in risk stratification to identify and appropriately treat still-eligible patients.
The second type of de-implementation is discontinuation or complete reversal of an existing practice without replacement. Examples of this type of de-implementation include the discontinued use of routine episiotomy during childbirth,5 hormone replacement therapy for prevention of chronic conditions in postmenopausal women,6 and the Choosing Wisely campaign's recommendation against routine pre-operative or pre-procedural electrocardiography screening for asymptomatic patients with low perioperative risk of death or myocardial infarction.7,8 Discontinuation without replacement is appropriate if there is no evidence of value for any subgroup of patients. The focus in this type of change is entirely on de-implementation and there is no implementation of new practice involved.
The third type of de-implementation involves substitution/reversal with a related replacement that is a closely related and more effective practice. For example, inhaled corticosteroids, which increase risk of pneumonia among patients with mild chronic obstructive pulmonary disease, can be replaced by long-acting beta-agonists or muscarinicantagonists.9,10 Coronary procedures can be completed via the radial artery in the wrist rather than the common route via the femoral artery in the groin, with fewer bleeding complications,11 and a particular class of diabetes medications (thiazolidinediones) with unacceptable cardiovascular risk appears to be on its way to being replaced by more effective alternatives.12 In bariatric surgery for weight loss, the vertical banded gastroplasty procedure that was commonly performed in the 1970s has been completely replaced by more modern procedures (e.g., Roux-en-Y gastric bypass, vertical sleeve gastrectomy). Substitution of related practices should have minimal impact on clinical effort and practice patterns in the long-term because such a substitution largely fits existing expectations and practices.
The fourth type of de-implementation is substitution/reversal with unrelated replacement. Examples include acute lower-back pain can be treated with physical therapy rather than lower back imaging and potentially surgery13,14 and chronic stable angina can often be treated medically instead of with coronary bypass surgery or percutaneous coronary interventions.15,16 There may be practical challenges to unrelated de-implementation because it requires engagement and buy-in from two different stakeholder groups to effect the designed unrelated changes and substantial change in clinical practices.
2. Learning and unlearning to de-implement
De-implementation has been described as the mirror of implementation,1 because they have similar cognitive challenges and processes. We argue that this is not always the case, and that de-implementation is a graduated continuum of individual, team, and organizational change that require different strategies in terms of learning and unlearning. Learning refers to the process of acquiring new skills or knowledge. Unlearning is a process of discarding outdated mental models to make room for alternative models.17,18 In each of the four types of change described above, change requires effort to learn new knowledge and to unlearn what was thought to be effective. Moreover, the relative efforts required for learning and unlearning vary by the type of change.
Table 1 illustrates hypothetical differences in the organizational efforts associated with learning and unlearning in each of the four types of change. The circles denote different levels of effort potentially associated with learning tasks. For example, partial reduction/reversal of practice may require minimal learning effort because new skills are not required for an established practice but some degree of education is needed on the new evidence and updated and restricted patient eligibility criteria. The associated unlearning effort to de-implement in the subgroup of patients where a practice is least effective requires relatively more effort to identify target populations for discontinued practice and to develop systems that implement practice discontinuation (e.g., active monitoring, clinical flags).
Table 1.
Partial Reversal | Complete Reversal | Replacement | ||
---|---|---|---|---|
Related | Unrelated | |||
Description | Reducing the frequency, breadth or scale of an existing outmoded practice, so that it is provided to a narrower subgroup of patients. | Universal discontinuation of ineffective practice. | Substitution of existing practice, replaced by a closely related and more effective intervention. | Substitution of existing practice, replaced by a more effective intervention that is unrelated to usual care. |
Learning | ||||
|
|
|
|
|
Unlearning | ||||
|
|
|
|
Complete reversal or discontinuation of an existing practice without replacement may require significant effort to overcome confirmation bias or loss aversion that might slow or prevent discontinuation, as well as effort to facilitate divestiture of practice (e.g., overcome behavioral inertia). For de-implementation efforts involving replacement, the learning component of organizational change may require a focus on adoption of evidence and engineering new processes to promote new routines. The unlearning objectives may require educating and technical supports to steer practitioners away from outmoded practice.
Strategies known for promoting implementation – educational outreach, reminders, multifaceted interventions19 – also apply to the learning and unlearning processes that are essential for effective de-implementation. However, strategies for de-implementation will vary in their suitability and impact by the type and context of de-implementation activities. Specifically, persuasive strategies such as provider education, academic detailing, local opinion leaders, and audit and feedback appear to be effective in learning and unlearning to reduce antibiotic prescribing for hospital inpatients.20 However, persuasive strategies do not appear to be effective in discontinuing advanced imaging for lower back pain.21
Policy intervention, such as financial incentives and payment reform may be a potent approach in unlearning efforts for partial reduction or complete discontinuation. For example, Medicare rolled its separate billable payments for medication and ancillary services into a new bundled payment for outpatient dialysis treatment in its efforts to curb unnecessary utilization of erythropoiesis stimulating agents (ESAs) that increase the risk of stroke or death in patients with kidney failure.22 After the bundled payment eliminated the opportunity to bill for these ancillary services, ESA use declined markedly.23,24 For other de-implementation involving partial reduction, financial incentives may be an inadequate lever to overcome behavioral inertia or logistical challenges of curtailing care delivery practices to only a subset of patients.
3. Strategies for change: coupling implementation and de-implementation
It would be useful for the field and for practitioners to identify strategies that simultaneously achieve both implementation and de-implementation goals. For example, an effort initially focused on implementation of a new practice may be more favorably received by a health care organization if it is coupled with a de-implementation effort by implementing a new practice that partially or fully replaces an existing practice that is not viewed favorably. Adding de-implementation to the implementation effort could result in an effort neutral change (in the long term) that would be more favorably received by over-burdened clinicians than (yet another) implementation effort without taking something away. Similarly, a more expansive de-implementation effort could include implementation of a more effective replacement. Depending on the nature of the de-implementation activity, a coupled strategy might involve a multicomponent intervention, with some components targeting de-implementation goals and others targeting implementation objectives. In the context of de-implementation involving substitution, this requires practitioners to identify strategies to (1) eliminate existing intervention/practice and (2) replace it with newer, more effective intervention that also address the barriers and facilitators to substitution. An example would be implementing low-value care measures alongside de-implementation of existing performance measures. It should be noted that coupling may not be necessary in all contexts. There may be ineffective practices that are not deeply entrenched that can be removed without introducing a new practice.
A coupling approach serves conceptual, practical and logistical purposes. Implementation of innovation sometimes requires not only have to learn new knowledge/skills (or at least figure out what current knowledge/skills are still relevant) but also having to unlearn what is already known (or at least figure how what knowledge/skills no longer apply). Singularly, each of these processes require significant cognitive and logistical effort. Ignoring one aspect of change may result in incomplete or ineffective efforts. By creating a dependence between the two activities, there is likely significant investment in short-term cognitive and physical effort. In the long term, coupling may generate economies of scale to minimize effort and investment for behavior change. As such, coupling may result in effort-neutral change and considerably raise the likelihood for change, as intended. Importantly, evidence is needed to support these hypotheses. As the field evolves, formal consideration of when to couple implementation and de-implementation efforts is a promising area that is ripe for testing, theory development, and refinement.
Acknowledgments
The authors thank David Simel, MD and the anonymous reviewers for their contributions to the development of this work. This study was funded in part by the Department of Veterans Affairs (VA) Quality Enhancement Research Initiative (QUE-16–170 and QUE-15–271) and a VA Research Career Scientist award (RCS 10–391), and supported by the Center of Innovation for Health Services Research in Primary Care at the Durham VA Health Care System (CIN 13–410) and the Seattle-Denver Center of Innovation for Veteran-Centered and Value-Driven Care at the VA Puget Sound Health Care System (CIN 13–402). The interpretation and reporting of these data are the responsibility of the authors and do not reflect the official policy or interpretation of the US Government, the Department of Veterans Affairs, Duke University, or University of Washington.
Footnotes
Disclosures
MLM reports ownership of Amgen stock due to his spouse's employment. All other authors (VW, CDH, and BJW) report no potential conflicts of interest.
References
- 1.van Bodegom-Vos L, Davidoff F, Marang-van de Mheen PJ. Implementation and de-implementation: two sides of the same coin? BMJ Qual Saf. 2017;26(6):495–501. doi: 10.1136/bmjqs-2016-005473. [DOI] [PubMed] [Google Scholar]
- 2.Siu AL Force USPST. Screening for breast cancer: U.S. preventive services task force recommendation statement. Ann Intern Med. 2016;164(4):279–296. doi: 10.7326/M15-2886. [DOI] [PubMed] [Google Scholar]
- 3.Moyer VA Force USPST. Screening for cervical cancer: u.s. Preventive Services Task Force recommendation statement. Ann Intern Med. 2012;156(12):880–891. doi: 10.7326/0003-4819-156-12-201206190-00424. [W312] [DOI] [PubMed] [Google Scholar]
- 4.Pogach L, Tseng CL, Soroka O, Maney M, Aron D. A proposal for an out-of-range glycemic population health safety measure for older adults with diabetes. Diabetes care. 2017;40(4):518–525. doi: 10.2337/dc16-0953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.American College of O Gynecologists' Committee on Practice B-O. Practice Bulletin No. 165: Prevention and Management of Obstetric Lacerations at Vaginal Delivery. Obstetrics and gynecology. 2016;128(1):e1–e15. doi: 10.1097/AOG.0000000000001523. [DOI] [PubMed] [Google Scholar]
- 6.Moyer VA, Force USPST. Menopausal hormone therapy for the primary prevention of chronic conditions: U. S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2013;158(1):47–54. doi: 10.7326/0003-4819-158-1-201301010-00553. [DOI] [PubMed] [Google Scholar]
- 7.American Board of Internal Medicine (ABIM) Foundation CW. American College of Cardiology: Electrocardiography pre-operative screening. 2017 〈 http://www.choosingwisely.org/clinician-lists/acc-ecg-pre-op-screening/〉.
- 8.Fleisher LA, Fleischmann KE, Auerbach AD, et al. ACC/AHA guideline on perioperative cardiovascular evaluation and management of patients undergoing non-cardiac surgery: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines. [2014] J Am Coll Cardiol. 2014;64(22):e77–137. doi: 10.1016/j.jacc.2014.07.944. [DOI] [PubMed] [Google Scholar]
- 9.Diette GB, Orr P, McCormack MC, Gandy W, Hamar B. Is pharmacologic care of chronic obstructive pulmonary disease consistent with the guidelines? Popul Health Manag. 2010;13(1):21–26. doi: 10.1089/pop.2008.0048. [DOI] [PubMed] [Google Scholar]
- 10.Franssen FM, Spruit MA, Wouters EF. Determinants of polypharmacy and compliance with GOLD guidelines in patients with chronic obstructive pulmonary disease. Int J Chron Obstruct Pulmon Dis. 2011;6:493–501. doi: 10.2147/COPD.S24443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jolly SS, Yusuf S, Cairns J, et al. Radial versus femoral access for coronary angiography and intervention in patients with acute coronary syndromes (RIVAL): a randomised, parallel group, multicentre trial. Lancet. 2011;377(9775):1409–1420. doi: 10.1016/S0140-6736(11)60404-2. [DOI] [PubMed] [Google Scholar]
- 12.Lipska KJ, Yao X, Herrin J, et al. Trends in drug utilization, glycemic control, and rates of severe hypoglycemia, 2006–2013. Diabetes Care. 2017;40(4):468–475. doi: 10.2337/dc16-0985. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Searle A, Spink M, Ho A, Chuter V. Exercise interventions for the treatment of chronic low back pain: a systematic review and meta-analysis of randomised controlled trials. Clin Rehabil. 2015;29(12):1155–1167. doi: 10.1177/0269215515570379. [DOI] [PubMed] [Google Scholar]
- 14.van Tulder MW, Koes BW, Bouter LM. Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions. Spine. 1997;22(18):2128–2156. doi: 10.1097/00007632-199709150-00012. [DOI] [PubMed] [Google Scholar]
- 15.Desai NR, Bradley SM, Parzynski CS, et al. Appropriate use criteria for coronary revascularization and trends in utilization, patient selection, and appropriateness of percutaneous coronary intervention. JAMA. 2015;314(19):2045–2053. doi: 10.1001/jama.2015.13764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Fihn SD, Blankenship JC, Alexander KP, et al. ACC/AHA/AATS/PCNA/SCAI/STS focused update of the guideline for the diagnosis and management of patients with stable ischemic heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines, and the American Association for Thoracic Surgery, Preventive Cardiovascular Nurses Association, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. [2014] Circulation. 2014;130(19):1749–1767. doi: 10.1161/CIR.0000000000000095. [DOI] [PubMed] [Google Scholar]
- 17.Bonchek M. Why the Problem with Learning Is Unlearning. Harvard Business Review; 2016. [Google Scholar]
- 18.Tsang EW, Zahra SA. Organizational unlearning. Human Relat. 2008;61(10):1435–1462. [Google Scholar]
- 19.Bero LA, Grilli R, Grimshaw JM, Harvey E, Oxman AD, Thomson MA. Closing the gap between research and practice: an overview of systematic reviews of interventions to promote the implementation of research findings. The Cochrane Effective Practice and Organization of Care Review Group. BMJ. 1998;317(7156):465–468. doi: 10.1136/bmj.317.7156.465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Davey P, Brown E, Charani E, et al. Interventions to improve antibiotic prescribing practices for hospital inpatients. Cochrane Database Syst Rev. 2013;(4):CD003543. doi: 10.1002/14651858.CD003543.pub3. [DOI] [PubMed] [Google Scholar]
- 21.Jenkins HJ, Hancock MJ, French SD, Maher CG, Engel RM, Magnussen JS. Effectiveness of interventions designed to reduce the use of imaging for low-back pain: a systematic review. CMAJ. 2015;187(6):401–408. doi: 10.1503/cmaj.141183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Brookhart MA, Schneeweiss S, Avorn J, Bradbury BD, Liu J, Winkelmayer WC. Comparative mortality risk of anemia management practices in incident hemodialysis patients. JAMA. 2010;303(9):857–864. doi: 10.1001/jama.2010.206. [DOI] [PubMed] [Google Scholar]
- 23.Swaminathan S, Mor V, Mehrotra R, Trivedi AN. Effect of medicare dialysis payment reform on use of erythropoiesis stimulating agents. Health Serv Res. 2015;50(3):790–808. doi: 10.1111/1475-6773.12252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wang C, Kane R, Levenson M, et al. Association between changes in CMS reimbursement policy and drug labels for erythrocyte-stimulating agents with outcomes for older patients undergoing hemodialysis covered by fee-for-service medicare. JAMA Intern Med. 2016;176(12):1818–1825. doi: 10.1001/jamainternmed.2016.6520. [DOI] [PubMed] [Google Scholar]