1. Introduction
Shared decision-making (SDM) and the use of patient decision aids are frequently highlighted as important tools for fostering patient engagement across conditions ranging from early stage breast cancer1 to musculoskeletal conditions,2and are associated with positive impacts upon patient-centered outcomes including decision satisfaction and quality in multiple studies.3-5 Research suggests that decision aids used in the context of SDM are associated with reduced anxiety, less regret and better health outcomes compared with usual care. 6Elsewhere, decision aids have been shown to support patients in making treatment choices consistent with their values - especially when there is clinical equipoise between such choices, as is the case for preference sensitive conditions. 7
Hip and knee osteoarthritis are two such preference sensitive musculoskeletal conditions for which current clinical evidence does not necessarily demonstrate superiority of surgery (arthroplasty) over non-surgical medical management (physical therapy, use of non-steroidal anti-inflammatory medications), 8 and the choice between these treatment options is likely to vary according to patient characteristics and personal values. 9 Additionally, these treatment options come with differing costs, side effects, and risks, 10 further underscoring the importance of patients being provided with information that empowers them to make informed, value-aligned treatment choices.11 As the costs of care for patients living with musculoskeletal conditions continues to rise (in 2014, direct costs for these conditions was estimated to be $980.1 billion in the United States alone) and as the prevalence of patients with hip and knee osteoarthritis continues to grow,12 so does the imperative for health systems to engage patients in conversations about treatment option benefits and tradeoffs such as through the use of decision aids intended to support SDM.
Recent scholarship in the field of SDM has underscored that the alignment between patients’ expressed treatment choices and treatments received represents a critical component of decision quality,13 14 which is defined as “the extent to which patients are informed, meaningfully involved in the decision-making process, and receive tests and treatments that reflect their goals and concerns.”15 Although such alignment has been recognized as an important outcome of SDM,16 some research shows that alignment between patient treatment choice and treatment receipt is not always achieved; in one large multi-year research trial of SDM among patients with low back pain, nearly one quarter of those patients electing for medical management received surgical interventions instead. 7Sullivan et al examined treatment choices of patients diagnosed with stage 1 lung cancer and found a lack of congruence between patient-expressed treatment choices and treatments received among 49% (n = 114) of study
participants.17 Scherr et al note that treatment choice may be dictated by other factors (such as physician rather than patient preferences) even where SDM has occurred.18 In addition to the clear gaps that exist in achieving alignment with patient expressed treatment choices, much of this research has focused upon patient populations drawn from single institutions or health systems.18,19
This paper explores the extent of alignment between patient-expressed treatment choices after exposure to decision aids and treatments received among hip and knee osteoarthritis patients within High Value Healthcare Collaborative (HVHC) systems while the collaborative sought to integrate SDM into routine clinical practice. Specifically, this work investigates which patient-level characteristics are associated with alignment between treatment choice and treatment received among patients diagnosed with hip or knee osteoarthritis within HVHC exposed to decision aids for these conditions. These analyses importantly examine such alignment in the context of patients drawn from diverse health systems in a learning collaborative,20 and in the context of routine implementation of decision aids to support SDM in clinical practice as opposed to a controlled trial setting. The aim of this study is to understand the associations between patient-level characteristics and alignment in expressed choices and treatments received for hip or knee osteoarthritis after a decision aid intervention.