EGEMS (Washington, DC)最新文献

Examples of fully integrated machine learning models that drive clinical care are rare. Despite major advances in the development of methodologies that outperform clinical experts and growing prominence of machine learning in mainstream medical literature, major challenges remain. At Duke Health, we are in our fourth year developing, piloting, and implementing machine learning technologies in clinical care. To advance the translation of machine learning into clinical care, health system leaders must address barriers to progress and make strategic investments necessary to bring health care into a new digital age. Machine learning can improve clinical workflows in subtle ways that are distinct from how statistics has shaped medicine. However, most machine learning research occurs in siloes, and there are important, unresolved questions about how to retrain and validate models post-deployment. Academic medical centers that cultivate and value transdisciplinary collaboration are ideally suited to integrate machine learning in clinical care. Along with fostering collaborative environments, health system leaders must invest in developing new capabilities within the workforce and technology infrastructure beyond standard electronic health records. Now is the opportunity to break down barriers and achieve scalable growth in the number of high-impact collaborations between clinical researchers and machine learning experts to transform clinical care.

Current approaches to addressing the problems families face when navigating complex service systems on behalf of their children rely largely on state or nationally driven efforts around the development of systems of care (SOCs). However, operationalizing meaningful family involvement within SOCs remains a challenge, with little attention paid to the role of personal social support networks (PSSNs). Specifically, risk factors related to the variations in the social connectedness of family social support networks are difficult to identify, assess, and track over time. This paper summarizes families' descriptions of their PSSNs and describes the development of a social network analysis tool, the Person-Centered Network App (PCNA), used to measure and monitor the social connectedness of families of children with special health care and developmental needs. Twenty-nine families participated in the project and completed social network surveys, identifying a total of 38 unique types of support partners and 230 partnerships (dyadic relationships). Families identified a range of formal and informal members including primary care providers, medical specialists, family, friends, faith-based organizations, insurance providers, nurses, community organizations, early interventionists, school resources, other families, online support groups, and public resources, rating 61 percent of them as "very important." Informal network members (e.g., family, friends) provided emotional and day-to-day support. Primary care providers, medical specialists, and public resources provided health care services while early intervention and medical specialists provided therapies. PSSNs were characterized by high levels of trust but low levels of coordination. These findings inform providers and case workers that families can readily describe their social connectedness in ways that may affect health care access and utilization. Understanding how PSSNs function in the lives of families of children with complex health care needs provides opportunities for improving systems of care (e.g., medical homes) and ultimately, enhancing health and developmental outcomes.

Objectives: This research tackles a critical issue in modern health care systems-namely, to determine if creating a user-centered health information system that is easy to utilize would lead to consumers who are more satisfied and more likely to accept the system.

Materials and methods: The health information system is a consumer service center that receives inquiries from consumers on how to find and pay for care. To understand if a system designed to decrease effort results in satisfaction, we redesigned the system, deployed it for 3 months, and then compared consumer satisfaction results to a control group. Satisfaction and Net Promoter surveys were provided to consumers who used the control system and consumers using the redesigned system.

Results: This study was completed over a 6 month continual time period where over 100,000 consumer interactions took place. Using 11 different metrics and data from over 5,000 random system users, it was shown that consumers were more satisfied with an information system designed to reduce their administrative effort.

Discussion: While not all consumer survey results were statistically significant, they all showed a shift towards improved satisfaction with the health care system. Statistically, it was shown that there was a dependency between the design of the system to provide information and many needs of the consumers.

Conclusion: A health care system designed to reduce effort in accessing care results in improved consumer satisfaction. Consumers are also more likely to trust the assistance provided by the organization.

Introduction: Hospital falls are a continuing clinical concern, with over one million falls occurring each year in the United States. Annually, hospital-acquired falls result in an estimated $34 billion in direct medical costs. Falls are considered largely preventable and, as a result, the Centers for Medicare and Medicaid Services have announced that fall-related injuries are no longer a reimbursable hospital cost. While policies and practices have been implemented to reduce falls, little sustained reduction has been achieved. Little empirical evidence supports the validity of published fall risk factors. While chart abstraction has been used to operationalize risk factors, few studies have examined registered nurses' (RNs') narrative notes as a source of actionable data. Therefore, the purpose of our study was to explore whether there is meaningful fall risk and prevention information in RNs' electronic narrative notes.

Methods: This study utilized a natural language processing design. Data for this study were extracted from the publicly available Medical Information Mart for Intensive Care (MIMIC-III) database. The date comprises deidentified EHR data associated with patients who stayed in critical care units between 2001 and 2012. Text mining procedures were performed on RN's narrative notes following the traditional steps of knowledge discovery.

Results: The corpus of data extracted from MIMIC-III database was comprised of 1,046,053 RNs' notes from 36,583 unique patients. We identified 3,972 notes (0.4 percent) representing 1,789 (5 percent) patients with explicit documentation related to fall risk/prevention. Around 10 percent of the notes (103,685) from 23,025 patients mentioned intrinsic (patient-related) factors that have been theoretically associated with risk of falling. An additional 1,322 notes (0.1 percent) from 692 patients (2 percent) mentioned extrinsic risk factors, related to organizational design and environment. Moreover, 7672 notes (0.7 percent) from 2,571 patients (7 percent) included information on interventions that could theoretically impact patient falls.

Conclusions: This exploratory study using a NLP approach revealed that meaningful information related to fall risk and prevention may be found in RNs' narrative notes. In particular, RNs' notes can contain information about clinical as well as environmental and organizational factors that could affect fall risk but are not explicitly recorded by the provider as a fall risk factors. In our study, potential fall risk factors were documented for more than half of the sample. Further research is needed to determine the predictive value of these factors.

Implications for policy or practice: This study highlights a potentially rich but understudied source of actionable fall risk data. Furthermore, the application of novel methods to identify quality and safety meas

Introduction: Although researchers recognize that sharing disparate data can improve population health, barriers (technical, motivational, economic, political, legal, and ethical) limit progress. In this paper, we aim to enhance the van Panhuis et al. framework of barriers to data sharing; we present a complementary solutions-based data-sharing process in order to encourage both emerging and established researchers, whether or not in academia, to engage in data-sharing partnerships.

Brief description of major components: We enhance the van Panhuis et al. framework in three ways. First, we identify the appropriate stakeholder(s) within an organization (e.g., criminal justice agency) with whom to engage in addressing each category of barriers. Second, we provide a representative sample of specific challenges that we have faced in our data-sharing partnerships with criminal justice agencies, local clinical systems, and public health. Third, and most importantly, we suggest solutions we have found successful for each category of barriers. We grouped our solutions into five core areas that cut across the barriers as well as stakeholder groups: Preparation, Clear Communication, Funding/Support, Non-Monetary Benefits, and Regulatory Assurances.Our solutions-based process model is complementary to the enhanced framework. An important feature of the process model is the cyclical, iterative process that undergirds it. Usually, interactions with new data-sharing partner organizations begin with the leadership team and progress to both the data management and legal teams; however, the process is not always linear.

Conclusions and next steps: Data sharing is a powerful tool in population health research, but significant barriers hinder such partnerships. Nevertheless, by aspiring to community-based participatory research principles, including partnership engagement, development, and maintenance, we have overcome barriers identified in the van Panhuis et al. framework and have achieved success with various data-sharing partnerships.In the future, systematically studying data-sharing partnerships to clarify which elements of a solutions-based approach are essential for successful partnerships may be helpful to academic and non-academic researchers. The organizational climate is certainly a factor worth studying also because it relates both to barriers and to the potential workability of solutions.

Introduction: A sizeable number of youth are currently struggling with anxiety, depression, and suicidal thoughts, yet many will not receive treatment. We sought to better understand if immediate response technology (IRT) could be used to gather mental health care data and educate youth on telemental health (TMH) resources.

Methods: Using an IRT imbedded within an interactive, media-rich school-based presentation, we gathered mental health history and preferences for TMH resources from 2,789 adolescents with a wide range of demographic and psychological characteristics.

Results: More than 80 percent of adolescents satisfied inclusion criteria for survey completion, and responses were statistically comparable across four diverse high school settings. Using Chi-squared analyses, we found that less than 10 percent of adolescents, especially girls and those with high depression/anxiety scores, had previously used TMH resources. After interacting with the IRT, many more (29 percent to 43 percent) expressed willingness to use these resources.

Discussion: The IRT system was effective in gathering mental electronic health data, delivering targeted mental health education, and promoting positive attitudes towards TMH among adolescents.

Conclusions: IRTs and other non-formalized technologies should be explored as cost-effective, easy-to-implement resources for electronic health data gathering and health care education.

Improving quality measurement while reducing costs helps public health programs identify and better support critical aspects of the care and services delivered to the patients they serve. This is true for state-based early hearing detection and intervention (EHDI) programs as they strive to develop robust clinical quality measures to help track the quality of hearing health services provided during the EHDI processes. Leveraging today's electronic health records and public health surveillance system functionalities, state reporting requirements facilitate and yield efficient collection and analysis of data for quality measurement. In this study, we tested three EHDI quality measures endorsed by the National Quality Forum using a retrospective sample of more than 1,100,000 newborns from 3 states using electronic health data available in the state EHDI Information Systems (EHDI-IS). The results of the analysis reported herein from a large multi-state cohort provide a "real life" benchmark for future quality improvement projects and of where EHDI stands today. Reflecting on these findings, suggestions are posed for enhancing the EHDI quality measures in future updates.

Objective: To understand the impact of varying measurement period on the calculation of electronic Clinical Quality Measures (eCQMs).

Background: eCQMs have increased in importance in value-based programs, but accurate and timely measurement has been slow. This has required flexibility in key measure characteristics, including measurement period, the timeframe the measurement covers. The effects of variable measurement periods on accuracy and variability are not clear.

Methods: 209 practices were asked to extract and submit four eCQMs from their Electronic Health Records on a quarterly basis using a 12-month measurement period. Quarterly submissions were collected via REDCap. The measurement periods of the survey data were categorized into non-standard (3, 6, 9 months and other) and standard periods (12 months). For comparison, patient-level data from three clinics were collected and calculated in an eCQM registry to measure the impact of varying measurement periods. We assessed the central tendency, shape of the distributions, and variability across the four measures. Analysis of variance (ANOVA) was conducted to analyze the differences among standard and non-standard measurement period means, and variation among these groups.

Results: Of 209 practices, 191 (91 percent) submitted data over three quarters. Of the 546 total submissions, 173 had non-standard measurement periods. Differences between measures with standard versus non-standard periods ranged from -3.3 percent to 14.2 percent between clinics (p < .05 for 3 of 4), using the patient-level data yielded deltas of -1.6 percent to 0.6 percent when comparing non-standard and standard periods.

Conclusion: Variations in measurement periods were associated with variation in performance between clinics for 3 of the 4 eCQMs, but did not have significant differences when calculated within clinics. Variations from standard measurement periods may reflect poor data quality and accuracy.

Physicians need nearly a decade of training to understand complex patient data such as laboratory tests and genomic data. How can these data possibly be delivered to patients in ways that they can understand and use?

Objectives: Patients have unique information needs to help them interpret and make decisions about laboratory test results they receive on web-based portals. However, current portals are not designed in a patient-centered way and little is known on how best to harness patients' information needs to inform user-centered interface design of portals. We designed a patient-facing laboratory test result interface prototype based on requirement elicitation research and used a mixed-methods approach to evaluate this interface.

Methods: After designing an initial test result display prototype, we used multiple evaluation methods, including focus group review sessions, expert consultation, and user testing, to make iterative design changes. For the user testing component, we recruited 14 patient-users to collect and analyze three types of data: comments made during testing sessions, responses to post-session questionnaires, and system usability scores.

Results: Our initial patient-centered interface design included visual ranges of laboratory values, nontechnical descriptions of the test and result, and access to features to help patients interpret and make decisions about their results. Findings from our evaluation resulted in 6 design iterations of the interface. Results from user testing indicate that the later versions of the interface fulfilled patient's information needs, were perceived as usable, and provided access to information and techniques that facilitated patient's ability to derive meaning from each test result.

Conclusions: Requirement elicitation studies can inform the design of a patient-facing test result interface, but considerable user-centered design efforts are necessary to create an interface that patients find useful. To promote patient engagement, health information technology designers and developers can use similar approaches to enhance user-centered software design in patient portals.