2.1 Study Design and Patient Population
This retrospective study draws upon data from patients diagnosed with hip or knee osteoarthritis who received care within one of 10 High Value Healthcare Collaborative (HVHC) systems. HVHC is a provider learning collaborative of health systems from across the United States that share data pertaining to multiple quality improvement projects with the aim of elevating the provision of high value, evidence-based patient care.21 The 10 member systems of HVHC include: Baylor Health System, Beth Israel Deaconess Medical Center, Denver Health, Eastern Main Healthcare Systems, Intermountain Healthcare, MaineHealth, Mayo Clinic, Scott and White Health, UCLA Health System, and Virginia Mason Medical Center. Data from each HVHC system was analyzed during the three-year period (from July 1, 2012 to June 30, 2015) when HVHC received a grant entitled “The High Value Healthcare Collaborative: Engaging Patients to Meet the Triple Aim” from the Center for Medicare and Medicaid Innovation (CMMI) to facilitate the routine use of decision aids to support SDM among patients with hip or knee osteoarthritis.22
Adult (age 18-86) patients with diagnoses of hip (International Classification of Diseases, Clinical Modification codes (ICD-9 CM) diagnostic codes 715.09, 715.15, 715.25, 715.35 or 715.95) or knee (ICD-9 CM diagnostic codes 715.09, 715.16, 715.26, 715.36 or 715.96) osteoarthritis who were exposed to decision aids within HVHC systems and for whom complete post-decision aid surveys assessing patient choices were completed prior to the end of the CMMI grant period (June 30, 2015) were included. Among 1,791 hip and knee osteoarthritis patients who were exposed to decision aids and completed post-decision aid patient surveys, a total of 1,351 expressed a post-decision aid treatment choice of either surgical or non-surgical intervention (n = 974 knee patients and n = 377 hip patients). Since the outcome of interest was alignment between post-decision aid treatment choice and treatment received, patients who expressed uncertainty after receiving decision aids (n = 440) were excluded.
HVHC’s grant-funded project made use of decision aids (specific for hip or knee osteoarthritis) meant to support SDM conversations between clinicians and patients and pre- and post-decision aid web-based questionnaires. For both hip and knee osteoarthritis patients, HVHC provider systems utilized Health Dialog-produced decision aids. Health Dialog decision aids meet standards set by the International Patient Decision Aid Standards Collaborative 23 and provide balanced information about treatment options for conditions including hip and knee osteoarthritis. Patients were invited to view decision aids online, on DVD, or on a tablet in the orthopedic surgeon’s office either prior to or following their consultations. Before and after viewing the decision aids, patients completed surveys that assessed their treatment choices, decision-making stage, and gathered information including patients’ education, insurance status, and self-assessed pain (as measured by the pain sub-scale of the Knee Injury and Osteoarthritis Outcome Score (KOOS) and Hip Injury and Osteoarthritis Outcome Score (HOOS) for knee and hip osteoarthritis patients, respectively).
For patients who viewed decision aids in the medical office prior to their appointment, health coaches - who in most cases were registered nurses – were available to answer questions regarding the decision aids and patient surveys. Health coaches received decision support training sponsored by the collaborative through a 4-week distance learning course that incorporated “active learning assignments” building upon the Ottawa Personal Decision Guide tools and framework.24During the patient’s orthopedic appointment, the orthopedic surgeon could review survey responses, clarify questions, and address patient concerns that may not have been answered by the decision aid alone. The adoption of decision aids among HVHC systems was motivated by the collaborative’s focus upon improving overall functioning of patients considering hip or knee surgery and ensuring that rates of such surgeries reflected patient treatment preferences.