Group Health’s Orthopedic Division’s Implementation Findings
In 2009 Group Health leadership implemented video and written decision aids across multiple specialty clinics as part of a regional demonstration project in Washington State, distributing over 12,000 aids over two years. The decision support tools were provided free of cost from the Informed Medical Decisions Foundation and Health Dialog for the first two years. Twelve tools were implemented in total; example topics included knee and hip osteoarthritis, benign prostatic hyperplasia, and early-stage breast cancer. A patient satisfaction survey accompanied the decision aid, however only 18% of patients returned the survey (King & Moulton, 2013). However, over 95% of responding patients positively rated the decision aid as helping them understand their options, preferences, and preparing them to meet with their provider.
In their orthopedic division, providers were expected to order decision aids through the EHR for any patient seen for knee or hip osteoarthritis. In 2012 they published the implementation results from their first year (Arterburn et al., 2012). They compared rates of surgery and associated costs after implementation of decision tools to averages in the year prior. While implementing decision tools, Group Health was also involved in a quality improvement program providing audit and feedback data to surgeons on their rates of surgery and volume of decision aid delivery.
Overall rates of hip and knee replacement surgery declined in the six months following implementation of decision aids compared to the historical control period (absolute difference of 12 fewer hip surgeries and 7 fewer knee per 100 patients). However, only one-third of eligible patients received a decision aid. In a sub-analysis, those who received a decision aid experienced increased surgery rates (44% higher for hip replacement, 103% higher for knee replacement). The authors proposed that the providers were screening patients and using a referral process to select patients with more advanced disease instead of using the aid widely across all patients.
Cost estimates included patient level health care costs, both inpatient and outpatient, in the year following implementation compared to the year prior. Unadjusted total costs trended down following implementation of decision aids from a mean of $16,557 to $13,489 for the hip osteoarthritis cohort, $10,040 to $8,041 in the knee cohort. Estimates are limited as they do not include implementation costs and Group Health providers are salaried, thus their income is not dependent on productivity limiting the generalizability of their findings implementation costs and Group Health providers are salaried, thus their income is not dependent on productivity limiting the generalizability of their findings.
Decision Aids for Patients with Type 2 Diabetes Mellitus
Dr. Victor Montori and colleagues at the Mayo Clinic investigated the impact of decision aids on patients with type 2 diabetes mellitus receiving care in nonacademic and rural community practices. Decision aids in the trial addressed medication options for diabetes or statins for cardiovascular disease prevention. The study design was a cluster-randomization model, with a clinic implementing one decision aid and following usual care for the other topic (e.g. use of a diabetes decision aid with usual care for statin initiation or vice versa).
The majority of clinicians (80%) reported the decision aids were easy to use and integrate into clinic workflow. Patients reported increases in knowledge and greater involvement in decision making conversations. In video-recorded encounters, the use of a decision aid prompted providers to engage patients in decision making to a greater degree.
In addition to the statin and diabetes medications decision aids, the Mayo Shared Decision Making National Resource Center also has tools related to depression medication choices, osteoporosis, smoking around the time of surgery, rheumatoid arthritis medication options, and advanced chest pain testing in the emergency room. All tools are publicly available for download with several aids available in Spanish.
Branda, M. E., LeBlanc, A., Shah, N. D., Tiedje, K., Ruud, K., Van Houten, H., ... & Montori, V. M. (2013). Shared decision making for patients with type 2 diabetes: a randomized trial in primary care. BMC health services research, 13(1), 1.
Developed by researchers at Oregon Health & Science University, Mammopad, is a targeted decision aid for women in their 40s considering breast cancer screening. The interactive web-based tool provides information on breast cancer, screening approaches, and risk assessment. The tool helps women identify their values and preferences in regards to breast cancer screening. Ultimately, women receive a decision summary document to bring in to their provider to guide further conversations. In evaluation studies, Mammopad reduced decision conflict for women and helped them feel more confident in their ultimate decision.
Eden, K. B., Scariati, P., Klein, K., Watson, L., Remiker, M., Hribar, M., ... & Nelson, H. D. (2015). Mammography decision aid reduces decisional conflict for women in their forties considering screening. Journal of Women's Health, 24(12), 1013-1020.