LAUSR

To cite: Riganti P, Kopitowski KS, McCaffery K, et al. BMJ EvidenceBased Medicine Epub ahead of print: [please include Day Month Year]. doi:10.1136/ bmjebm-2022-112201 © Author(s) (or their employer(s)) 2023. No commercial reuse. See rights and permissions. Published by BMJ. Introduction In the last decades, researchers, governments and public campaigns have increased awareness about healthcare overuse. Lowvalue care is described as care unlikely to benefit the patient given the harms, costs or available alternatives. Clinical practice guidelines with ‘donotdo’ recommendations and other deimplementation strategies were promoted to reduce it. One of these strategies is shared decision making (SDM). SDM was traditionally described as an approach to enhance patient involvement in healthcare decisions by communicating evidencebased information about options, their pros and cons, and eliciting patients' preferences to support them to deliberate about those options. Currently, some authors consider SDM as a broader concept, a method that adapts to a wider range of situations where patients and clinicians make decisions together using different approaches (not only limited to cases with a fine balance between benefits and harms, where practitioners help to weigh pros and cons with patients' values and preferences). Different studies showed that when patients are better informed about the benefits and harms of interventions (eg, surgeries, screening tests, medications), they tend to decline lowvalue care. 6 7 These findings might explain why SDM has been promoted to reduce lowvalue care. 6–9 Focusing on the conversations between patients, caregivers and clinicians during the clinical encounter, we reflect on why using SDM for deimplementation of lowvalue care can be paradoxical. When it is problematic to use SDM for low-value care de-implementation Both deimplementation of lowvalue care and SDM are informed by new evidence that contradicts current practice, that is medical reversal, but have different purposes and pursue different outcomes (see table 1). Deimplementation strategies aim at removing, replacing, reducing or restricting lowvalue care to solve problems of rising healthcare costs and harms. SDM aims to respond to patients' problems to find the care that best fits the patients' and their families' unique context, aligned with their needs and preferences. We can pragmatically classify lowvalue care in practices with highquality or lowquality evidence, with stronger or more conditional recommendations. In the first category, we can find lowvalue care practices with highquality evidence and a ‘strong recommendation against’ because the practice has harms that always outweigh benefits and needs to be discontinued (eg, do not perform vaginal ultrasound for ovarian cancer screening) or because it has proven ineffective or harmful for some situations but proven beneficial for others and needs to be limited to a specific population (do not perform a percutaneous coronary intervention in patients with stable chronic angina without risk factors that are well controlled with medication or do not prescribe antibiotics for viral upper respiratory infections without specific criteria). 12 13 In the second category, lowvalue care practices might have lowquality evidence or a tight balance between benefits and harms and a conditional recommendation (eg, do not routinely