Implementing Shared Decision-Making for Cancer Screening in Primary Care

点击翻译按钮获取中文摘要

Even more than other screening services, it is difficult to know which patients are most likely to benefit from lung cancer screening (LCS). More difficult yet is explaining to patients the trade-offs of screening for them personally, particularly when time for discussing preventive services is limited to less than 5 minutes during a busy visit. Decision support can help, but must contain 3 key features to be maximally effective and sustainable: 1) communication of accurate estimates of the benefit of LCS for an individual patient; 2) design and content that incorporates the values and concerns of informed stakeholders (clinicians, VA leaders, and Veterans) and a practical approach to routinely personalizing screening during busy clinic visits; and 3) design that optimizes the tool’s fit with VA clinical workflows. In collaboration with my mentors and steering committee, I propose to carry out work that will leverage and substantially build on a project I am co-leading – Ann Arbor VA’s funded PROVE QUERI – which studies expeditious implementation of a more traditional shared decision- making (SDM) tool equipped with only the first feature (individualized estimates of LCS benefit). To achieve these goals the CDA has 3 aims. Aim 1 – Assess informed stakeholder views (3 projects): In this aim I will use democratic deliberation to extensively inform and engage key LCS stakeholders to understand their opinions/recommendations about offering and discussing LCS. In project 1a (“Clinical Forum”) I will recruit a group of 12-15 clinical stakeholders (PCPs, screening coordinators, and clinical leaders) from VA nationwide to inform them of the heterogeneity of LCS benefit and engage them in deliberation, obtaining their recommendations for “simple boundary rules” for when screening should be encouraged, discouraged, or left wholly to l patient choice. In project 1b (2 “Veteran Forums”) I will recruit a random sample of up to 64 VISN 10 Veterans who are LCS-eligible and conduct 2 separate daylong forums. The core objective will be to utilize 2 content experts with differing viewpoints to inform Veterans about the challenges of LCS implementation and the constraints of clinical practice, and then to obtain individual and group recommendations for “simple how-to rules” clinicians can use to involve patients in LCS discussions given the competing demands PCPs face. In project 1c I will use risk communication theory to develop innovative new tool features that incorporate the VA Stakeholder guidance and can help guide clinicians in how to efficiently personalize LCS. Aim 2 – Optimize decision support ‘fit’ with the VA clinical context (2 projects): In project 2a I will conduct a synthesis analysis of multiple streams of PROVE QUERI data to understand how LCS decision support can be redesigned to overcome implementation challenges and enhance motivators for tool use. Then, in project 2b I explore wiki surveys as a new method of engaging clinical stakeholders to: 1) evaluate the new decision support content and design strategies generated in projects 1c/2a; 2) contribute new ideas; and 3) help prioritize which strategies and content for inclusion in the upgraded tool. Aim 3 – Pilot (3 projects): Informed by Aims 1 and 2, I will develop a paper-based prototype in project 3a. Then, in project 3b I will assess usability by conducting 5-8 monitored mock clinical encounters (each mock encounter will include a patient, a PCP, and a PACT staff member). In project 3c, I will pilot the tool with 30 PCPs who are purposively sampled from 6 diverse outpatient VA settings. Each PCP will first receive a 1-on-1 educational intervention. I will then test the feasibility of using my paper-based decision support with 2-3 LCS- eligible Veterans per participating PCP. Encounters will be audio recorded to assess the content and length of LCS discussions. PCPs and patients will complete surveys and interviews to assess satisfaction and comfort with the LCS decision-making process. The core goal is to assess whether the tool can support brief and high- quality personalized LCS discussions while being acceptable to a diverse sample of PCPs and Veterans.