Savitz, Lucy A.

AbstractIntroductionCOVID‐19 has created additional challenges for the analysis of non‐randomized interventions in health system settings. Our objective is to evaluate these challenges and identify lessons learned from the analysis of a medically tailored meals (MTM) intervention at Kaiser Permanente Northwest (KPNW) that began in April 2020.MethodsWe identified both a historical and concurrent comparison group. The historical comparison group included patients living in the same area as the MTM recipients prior to COVID‐19. The concurrent comparison group included patients admitted to contracted non‐KPNW hospitals or admitted to a KPNW facility and living outside the service area for the intervention but otherwise eligible. We used two alternative propensity score methods in response to the loss of sample size with exact matching to evaluate the intervention.ResultsWe identified 452 patients who received the intervention, 3873 patients in the historical comparison group, and 5333 in the concurrent comparison group. We were able to mostly achieve balance on observable characteristics for the intervention and the two comparison groups.ConclusionsLessons learned included: (a) The use of two different comparison groups helped to triangulate results; (b) the meaning of utilization measures changed pre‐ and post‐COVID‐19; and (c) that balance on observable characteristics can be achieved, especially when the comparison groups are meaningfully larger than the intervention group. These findings may inform the design for future evaluations of interventions during COVID‐19.

AbstractHow to deliver best care in various clinical settings remains a vexing problem. All pertinent healthcare-related questions have not, cannot, and will not be addressable with costly time- and resource-consuming controlled clinical trials. At present, evidence-based guidelines can address only a small fraction of the types of care that clinicians deliver. Furthermore, underserved areas rarely can access state-of-the-art evidence-based guidelines in real-time, and often lack the wherewithal to implement advanced guidelines. Care providers in such settings frequently do not have sufficient training to undertake advanced guideline implementation. Nevertheless, in advanced modern healthcare delivery environments, use of eActions (validated clinical decision support systems) could help overcome the cognitive limitations of overburdened clinicians. Widespread use of eActions will require surmounting current healthcare technical and cultural barriers and installing clinical evidence/data curation systems. The authors expect that increased numbers of evidence-based guidelines will result from future comparative effectiveness clinical research carried out during routine healthcare delivery within learning healthcare systems.

AbstractIntroductionThe learning health system (LHS) aligns science, informatics, incentives, stakeholders, and culture for continuous improvement and innovation. The Agency for Healthcare Research and Quality and the Patient‐Centered Outcomes Research Institute designed a K12 initiative to grow the number of LHS scientists. We describe approaches developed by 11 funded centers of excellence (COEs) to promote partnerships between scholars and health system leaders and to provide mentored research training.MethodsSince 2018, the COEs have enlisted faculty, secured institutional resources, partnered with health systems, developed and implemented curricula, recruited scholars, and provided mentored training. Program directors for each COE provided descriptive data on program context, scholar characteristics, stakeholder engagement, scholar experiences with health system partnerships, roles following program completion, and key training challenges.ResultsTo date, the 11 COEs have partnered with health systems to train 110 scholars. Nine (82%) programs partner with a Veterans Affairs health system and 9 (82%) partner with safety net providers. Clinically trained scholars (n = 87; 79%) include 70 physicians and 17 scholars in other clinical disciplines. Non‐clinicians (n = 29; 26%) represent diverse fields, dominated by population health sciences. Stakeholder engagement helps scholars understand health system and patient/family needs and priorities, enabling opportunities to conduct embedded research, improve outcomes, and grow skills in translating research methods and findings into practice. Challenges include supporting scholars through roadblocks that threaten to derail projects during their limited program time, ranging from delays in access to data to COVID‐19‐related impediments and shifts in organizational priorities.ConclusionsFour years into this novel training program, there is evidence of scholars' accomplishments, both in traditional academic terms and in terms of moving along career trajectories that hold the potential to lead and accelerate transformational health system change. Future LHS training efforts should focus on sustainability, including organizational support for scholar activities.

Background: The dominant healthcare payment model in the US, fee-for-service, provides incentives to deliver costlier care for heart failure and disincentives to deliver higher quality care. One proposed method to align cost and quality incentives is to impose financial penalties on providers when patients are readmitted within 30 days after a heart failure hospitalization. We hypothesized that this strategy might not lower overall costs, because lower readmission costs might be offset by higher non-readmission post-acute care costs. Methods: We studied cost to Medicare from 2008-09 for all heart failure hospitalizations (MS-DRG 291, 292, or 293) in the 15 healthcare delivery systems participating in the High Value Health Collaborative. We categorized costs for the 30, 90, and 180 days after the day of admission for first heart failure hospitalization in the study period into acute care costs (inpatient care, outpatient facility costs with 3 day look back window, part B professional fees), inpatient readmission cost, and non-readmission post-acute care costs (outpatient facility fee, sub-acute rehab, inpatient rehab, long term care, home health care, part B professional fees). Results: Using the most common MS DRG (DRG 292) as an example, across the 15 healthcare systems, we found wide variation in median total costs for the 30- ($6487-$11565), 90- ($8049-$18918), and 180- ($9844-$24752) day time frames. For the 30-day time frame, readmission costs ($700 - $2684) and non-readmission post-acute care costs ($477 - $6398) also showed wide variation. The Figure is a plot of these two costs, with bubbles sized to reflect the number of cases for the hospital system (30-day time frame); results were comparable for 90 and 180 days. Contrary to our hypothesis, we found no inverse relationship between readmission costs and non-readmission post-acute care costs. These data do not account for possible differences in case mix and in geographic differences in labor and facility costs. Conclusions: Non-readmission post-acute costs for heart failure are unrelated to readmission costs but were greater than readmission costs in a number of healthcare deliver systems, raising questions about the overall value to the healthcare systems of an exclusive focus on readmission as a strategy for reducing the costs of heart failure.

AbstractClinical decision-making is based on knowledge, expertise, and authority, with clinicians approving almost every intervention—the starting point for delivery of “All the right care, but only the right care,” an unachieved healthcare quality improvement goal. Unaided clinicians suffer from human cognitive limitations and biases when decisions are based only on their training, expertise, and experience. Electronic health records (EHRs) could improve healthcare with robust decision-support tools that reduce unwarranted variation of clinician decisions and actions. Current EHRs, focused on results review, documentation, and accounting, are awkward, time-consuming, and contribute to clinician stress and burnout. Decision-support tools could reduce clinician burden and enable replicable clinician decisions and actions that personalize patient care. Most current clinical decision-support tools or aids lack detail and neither reduce burden nor enable replicable actions. Clinicians must provide subjective interpretation and missing logic, thus introducing personal biases and mindless, unwarranted, variation from evidence-based practice. Replicability occurs when different clinicians, with the same patient information and context, come to the same decision and action. We propose a feasible subset of therapeutic decision-support tools based on credible clinical outcome evidence: computer protocols leading to replicable clinician actions (eActions). eActions enable different clinicians to make consistent decisions and actions when faced with the same patient input data. eActions embrace good everyday decision-making informed by evidence, experience, EHR data, and individual patient status. eActions can reduce unwarranted variation, increase quality of clinical care and research, reduce EHR noise, and could enable a learning healthcare system.