Participatory methods to support team science development for predictive analytics in health.

Predictive analytics in health is a complex, transdisciplinary field requiring collaboration across diverse scientific and stakeholder groups. Pilot implementation of participatory research to foster team science in predictive analytics through a partnered-symposium and funding competition. In total, 85 stakeholders were engaged across diverse translational domains, with a significant increase in perceived importance of early inclusion of patients and communities in research. Participatory research approaches may be an effective model for engaging broad stakeholders in predictive analytics.

Comparative Effectiveness of a Technology-Facilitated Depression Care Management Model in Safety-Net Primary Care Patients With Type 2 Diabetes: 6-Month Outcomes of a Large Clinical Trial.

BACKGROUND: Comorbid depression is a significant challenge for safety-net primary care systems. Team-based collaborative depression care is effective, but complex system factors in safety-net organizations impede adoption and result in persistent disparities in outcomes. Diabetes-Depression Care-management Adoption Trial (DCAT) evaluated whether depression care could be significantly improved by harnessing information and communication technologies to automate routine screening and monitoring of patient symptoms and treatment adherence and allow timely communication with providers. OBJECTIVE: The aim of this study was to compare 6-month outcomes of a technology-facilitated care model with a usual care model and a supported care model that involved team-based collaborative depression care for safety-net primary care adult patients with type 2 diabetes. METHODS: DCAT is a translational study in collaboration with Los Angeles County Department of Health Services, the second largest safety-net care system in the United States. A comparative effectiveness study with quasi-experimental design was conducted in three groups of adult patients with type 2 diabetes to compare three delivery models: usual care, supported care, and technology-facilitated care. Six-month outcomes included depression and diabetes care measures and patient-reported outcomes. Comparative treatment effects were estimated by linear or logistic regression models that used generalized propensity scores to adjust for sampling bias inherent in the nonrandomized design. RESULTS: DCAT enrolled 1406 patients (484 in usual care, 480 in supported care, and 442 in technology-facilitated care), most of whom were Hispanic or Latino and female. Compared with usual care, both the supported care and technology-facilitated care groups were associated with significant reduction in depressive symptoms measured by scores on the 9-item Patient Health Questionnaire (least squares estimate, LSE: usual care=6.35, supported care=5.05, technology-facilitated care=5.16; P value: supported care vs usual care=.02, technology-facilitated care vs usual care=.02); decreased prevalence of major depression (odds ratio, OR: supported care vs usual care=0.45, technology-facilitated care vs usual care=0.33; P value: supported care vs usual care=.02, technology-facilitated care vs usual care=.007); and reduced functional disability as measured by Sheehan Disability Scale scores (LSE: usual care=3.21, supported care=2.61, technology-facilitated care=2.59; P value: supported care vs usual care=.04, technology-facilitated care vs usual care=.03). Technology-facilitated care was significantly associated with depression remission (technology-facilitated care vs usual care: OR=2.98, P=.04); increased satisfaction with care for emotional problems among depressed patients (LSE: usual care=3.20, technology-facilitated care=3.70; P=.05); reduced total cholesterol level (LSE: usual care=176.40, technology-facilitated care=160.46; P=.01); improved satisfaction with diabetes care (LSE: usual care=4.01, technology-facilitated care=4.20; P=.05); and increased odds of taking an glycated hemoglobin test (technology-facilitated care vs usual care: OR=3.40, P<.001). CONCLUSIONS: Both the technology-facilitated care and supported care delivery models showed potential to improve 6-month depression and functional disability outcomes. The technology-facilitated care model has a greater likelihood to improve depression remission, patient satisfaction, and diabetes care quality.

The effect of implementing a care coordination program on team dynamics and the patient experience.

OBJECTIVES: Care coordination programs are frequently implemented in the redesign of primary care systems, focused on improving patient outcomes and reducing utilization. However, redesign can be disruptive, affect patient experiences, and undermine elements in the patient-centered medical home, such as team-based care. STUDY DESIGN: Case-controlled study with difference-in-differences (DID) and cross-sectional analyses. METHODS: The phased implementation of a care coordination program permitted evaluation of a natural experiment to compare measures of patient experience and teamwork in practices with and without care coordinators. Patient experience scores were compared before and after the introduction of care coordinators, using DID analyses. Cross-sectional data were used to compare teamwork, based on the relational coordination survey, and physician-perceived barriers to coordinated care between clinics with and without care coordinators. RESULTS: We evaluated survey responses from 459 staff and physicians and 13,441 patients in 26 primary care practices. Practices with care coordinators did not have significantly different relational coordination scores compared with practices without care coordinators, and physicians in these practices did not report reduced barriers to coordinated care. After implementation of the program, patients in practices with care coordinators reported a more positive experience with staff over time (DID, 2.6 percentage points; P = .0009). CONCLUSIONS: A flexible program that incorporates care coordinators into the existing care team was minimally disruptive to existing team dynamics, and the embedded care coordinators were associated with a small increase in patient ratings that reflected a more positive experience with staff.

Patient-Provider Communication and Health Outcomes Among Individuals With Atherosclerotic Cardiovascular Disease in the United States: Medical Expenditure Panel Survey 2010 to 2013.

BACKGROUND: Consumer-reported patient-provider communication (PPC) assessed by Consumer Assessment of Health Plans Survey in ambulatory settings is incorporated as a complementary value metric for patient-centered care of chronic conditions in pay-for-performance programs. In this study, we examine the relationship of PPC with select indicators of patient-centered care in a nationally representative US adult population with established atherosclerotic cardiovascular disease. METHODS AND RESULTS: The study population consisted of a nationally representative sample of 6810 individuals (aged ≥18 years), representing 18.3 million adults with established atherosclerotic cardiovascular disease (self-reported or International Classification of Diseases, Ninth Edition diagnosis) reporting a usual source of care in the 2010 to 2013 pooled Medical Expenditure Panel Survey cohort. Participants responded to questions from Consumer Assessment of Health Plans Survey that assessed PPC, and we developed a weighted PPC composite score using their responses, categorized as 1 (poor), 2 (average), and 3 (optimal). Outcomes of interest were (1) patient-reported outcomes: 12-item Short Form physical/mental health status, (2) quality of care measures: statin and ASA use, (3) healthcare resource utilization: emergency room visits and hospital stays, and (4) total annual and out-of-pocket healthcare expenditures. Atherosclerotic cardiovascular disease patients reporting poor versus optimal were over 2-fold more likely to report poor outcomes; 52% and 26% more likely to report that they are not on statin and aspirin, respectively, had a significantly greater utilization of health resources (odds ratio≥2 emergency room visit, 1.41 [95% confidence interval, 1.09-1.81]; odds ratio≥2 hospitalization, 1.36 [95% confidence interval, 1.04-1.79]), as well as an estimated $1243 ($127-$2359) higher annual healthcare expenditure. CONCLUSIONS: This study reveals a strong relationship between PPC and patient-reported outcomes, utilization of evidence-based therapies, healthcare resource utilization, and expenditures among those with established atherosclerotic cardiovascular disease.

Complexity in Redesigning Depression Care: Comparing Intention Versus Implementation of an Automated Depression Screening and Monitoring Program.

Health care technology promises to improve quality and decrease costs while improving the patient experience. A recent trial tested a novel health technology that sought to reduce barriers to guideline-recommended depression care by coupling automated telephonic patient assessment with provider notification and decision support. This study compares the intended effects and the end users' experience with using this technology. Responses to semi-structured interviews with the leadership and design team (the intention group) were compared with responses from health professionals in clinics using the technology (the implementation group). Qualitative methods were applied to compare perspectives on the role of technology in care delivery and on the technology's impact on barriers to providing guideline-recommended care. Four members of the intention group and 17 members of the implementation group were interviewed. The 2 groups had similar notions of technology's role in care delivery. The technology increased provider awareness, empowered midlevel providers, and facilitated collaboration among care team members. However, the implementation group identified gaps in care processes and in the informatics system that the technology did not address, and for which providers remained responsible. The 2 groups had comparable perspectives on the value and limits of technology in improving adherence to guideline-recommended care. However, the intention group did not discuss many of the shortcomings identified by the implementation group. Also, the interviews suggest that although some barriers persisted and a few new ones arose, the technology succeeded in reducing barriers to guideline-recommended care.

Innovative approach to patient-centered care coordination in primary care practices.

OBJECTIVES: Although care coordination is an essential component of the patient-centered medical home structure, current case manager models have limited usefulness to population health because they typically serve a small group of patients defined based on disease or utilization. Our objective was to support our health system's population health by implementing and evaluating a program that embedded nonlicensed coordinators within our primary care practices to support physicians in executing care plans and communicating with patients. STUDY DESIGN: Matched case-control differences-in-differences. METHODS: Comprehensive care coordinators (CCC) were introduced into 14 of the system's 28 practice sites in 2 waves. After a structured training program, CCCs identified, engaged, and intervened among patients within the practice in conjunction with practice primary care providers. We counted and broadly coded CCC activities that were documented in the intervention database. We examined the impact of CCC intervention on emergency department (ED) utilization at the practice level using a negative binomial multivariate regression model controlling for age, gender, and medical complexity. RESULTS: CCCs touched 10,500 unique patients over a 1-year period. CCC interventions included execution of care (38%), coordination of transitions (32%), self-management support/link to community resources (15%), monitor and follow-up (10%), and patient assessment (1%). The CCC intervention group had a 20% greater reduction in its prepost ED visit rate compared with the control group (P < .0001). CONCLUSIONS: Our CCC intervention demonstrated a significant reduction in ED visits by focusing on the centrality of the primary care provider and practice. Our model may serve as a cost-effective and scalable alternative for care coordination in primary care.

Development of a Clinical Forecasting Model to Predict Comorbid Depression Among Diabetes Patients and an Application in Depression Screening Policy Making.

INTRODUCTION: Depression is a common but often undiagnosed comorbid condition of people with diabetes. Mass screening can detect undiagnosed depression but may require significant resources and time. The objectives of this study were 1) to develop a clinical forecasting model that predicts comorbid depression among patients with diabetes and 2) to evaluate a model-based screening policy that saves resources and time by screening only patients considered as depressed by the clinical forecasting model. METHODS: We trained and validated 4 machine learning models by using data from 2 safety-net clinical trials; we chose the one with the best overall predictive ability as the ultimate model. We compared model-based policy with alternative policies, including mass screening and partial screening, on the basis of depression history or diabetes severity. RESULTS: Logistic regression had the best overall predictive ability of the 4 models evaluated and was chosen as the ultimate forecasting model. Compared with mass screening, the model-based policy can save approximately 50% to 60% of provider resources and time but will miss identifying about 30% of patients with depression. Partial-screening policy based on depression history alone found only a low rate of depression. Two other heuristic-based partial screening policies identified depression at rates similar to those of the model-based policy but cost more in resources and time. CONCLUSION: The depression prediction model developed in this study has compelling predictive ability. By adopting the model-based depression screening policy, health care providers can use their resources and time better and increase their efficiency in managing their patients with depression.

Integrating Science and Engineering to Implement Evidence-Based Practices in Health Care Settings.

Integrating two distinct and complementary paradigms, science and engineering, may produce more effective outcomes for the implementation of evidence-based practices in health care settings. Science formalizes and tests innovations, whereas engineering customizes and optimizes how the innovation is applied tailoring to accommodate local conditions. Together they may accelerate the creation of an evidence-based healthcare system that works effectively in specific health care settings. We give examples of applying engineering methods for better quality, more efficient, and safer implementation of clinical practices, medical devices, and health services systems. A specific example was applying systems engineering design that orchestrated people, process, data, decision-making, and communication through a technology application to implement evidence-based depression care among low-income patients with diabetes. We recommend that leading journals recognize the fundamental role of engineering in implementation research, to improve understanding of design elements that create a better fit between program elements and local context.

Opioid prescribing: a systematic review and critical appraisal of guidelines for chronic pain.

BACKGROUND: Deaths due to prescription opioid overdoses have increased dramatically. High-quality guidelines could help clinicians mitigate risks associated with opioid therapy. PURPOSE: To evaluate the quality and content of guidelines on the use of opioids for chronic pain. DATA SOURCES: MEDLINE, National Guideline Clearinghouse, specialty society Web sites, and international guideline clearinghouses (searched in July 2013). STUDY SELECTION: Guidelines published between January 2007 and July 2013 addressing the use of opioids for chronic pain in adults were selected. Guidelines on specific settings, populations, and conditions were excluded. DATA EXTRACTION: Guidelines and associated systematic reviews were evaluated using the Appraisal of Guidelines for Research and Evaluation II (AGREE II) instrument and A Measurement Tool to Assess Systematic Reviews (AMSTAR), respectively, and recommendations for mitigating opioid-related risks were compared. DATA SYNTHESIS: Thirteen guidelines met selection criteria. Overall AGREE II scores were 3.00 to 6.20 (on a scale of 1 to 7). The AMSTAR ratings were poor to fair for 10 guidelines. Two received high AGREE II and AMSTAR scores. Most guidelines recommend that clinicians avoid doses greater than 90 to 200 mg of morphine equivalents per day, have additional knowledge to prescribe methadone, recognize risks of fentanyl patches, titrate cautiously, and reduce doses by at least 25% to 50% when switching opioids. Guidelines also agree that opioid risk assessment tools, written treatment agreements, and urine drug testing can mitigate risks. Most recommendations are supported by observational data or expert consensus. LIMITATION: Exclusion of non-English-language guidelines and reliance on published information. CONCLUSION: Despite limited evidence and variable development methods, recent guidelines on chronic pain agree on several opioid risk mitigation strategies, including upper dosing thresholds; cautions with certain medications; attention to drug-drug and drug-disease interactions; and use of risk assessment tools, treatment agreements, and urine drug testing. Future research should directly examine the effectiveness of opioid risk mitigation strategies. PRIMARY FUNDING SOURCE: California Department of Industrial Relations and California Commission on Health and Safety and Workers' Compensation.

Repetitive transcranial magnetic stimulation of human area MT/V5 disrupts perception and storage of the motion aftereffect.

Following adaptation to a moving stimulus, the introduction of a stationary pattern creates the illusion of motion. This phenomenon, known as the motion aftereffect (MAE), can be delayed by placing a blank storage interval between the adapting and test stimuli. Human motion selective area MT/V5 has been proposed as the likely neural origin of MAEs. To examine the role of MT/V5 in perceiving and storing MAEs, we applied repetitive transcranial magnetic stimulation (rTMS) to this area during a 10s storage interval and while subjects perceived illusory motion. Our results show that rTMS disrupts perception of the MAE when it is delivered in the early parts of the storage period and when it is applied during the perceptual MAE itself. Stimulation of control regions corresponding to V1 or Cz did not affect the MAE. In addition, magnetic stimulation of dorsolateral prefrontal and posterior parietal cortices did not disrupt MAE perception. These data provide experimental support for the notion that MT/V5 subserves perception and storage of the motion aftereffect.