Developing a data infrastructure for a learning health system: the PORTAL network

Definition of a Practical Taxonomy for Referencing Data Quality Problems in Health Care Databases

INTRODUCTION

 Health care information systems can generate and/or record huge volumes of data, some of which may be reused for research, clinical trials, or teaching. However, these databases can be affected by data quality problems; hence, an important step in the data reuse process consists in detecting and rectifying these issues. With a view to facilitating the assessment of data quality, we developed a taxonomy of data quality problems in operational databases.

MATERIAL

 We searched the literature for publications that mentioned "data quality problems," "data quality taxonomy," "data quality assessment," or "dirty data." The publications were then reviewed, compared, summarized, and structured using a bottom-up approach, to provide an operational taxonomy of data quality problems. The latter were illustrated with fictional examples (though based on reality) from clinical databases.

RESULTS

 Twelve publications were selected, and 286 instances of data quality problems were identified and were classified according to six distinct levels of granularity. We used the classification defined by Oliveira et al to structure our taxonomy. The extracted items were grouped into 53 data quality problems.

DISCUSSION

 This taxonomy facilitated the systematic assessment of data quality in databases by presenting the data's quality according to their granularity. The definition of this taxonomy is the first step in the data cleaning process. The subsequent steps include the definition of associated quality assessment methods and data cleaning methods.

CONCLUSION

 Our new taxonomy enabled the classification and illustration of 53 data quality problems found in hospital databases.

A scoping review of semantic integration of health data and information

OBJECTIVE

We summarized a decade of new research focusing on semantic data integration (SDI) since 2009, and we aim to: (1) summarize the state-of-art approaches on integrating health data and information; and (2) identify the main gaps and challenges of integrating health data and information from multiple levels and domains.

MATERIALS AND METHODS

We used PubMed as our focus is applications of SDI in biomedical domains and followed the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) to search and report for relevant studies published between January 1, 2009 and December 31, 2021. We used Covidence-a systematic review management system-to carry out this scoping review.

RESULTS

The initial search from PubMed resulted in 5,326 articles using the two sets of keywords. We then removed 44 duplicates and 5,282 articles were retained for abstract screening. After abstract screening, we included 246 articles for full-text screening, among which 87 articles were deemed eligible for full-text extraction. We summarized the 87 articles from four aspects: (1) methods for the global schema; (2) data integration strategies (i.e., federated system vs. data warehousing); (3) the sources of the data; and (4) downstream applications.

CONCLUSION

SDI approach can effectively resolve the semantic heterogeneities across different data sources. We identified two key gaps and challenges in existing SDI studies that (1) many of the existing SDI studies used data from only single-level data sources (e.g., integrating individual-level patient records from different hospital systems), and (2) documentation of the data integration processes is sparse, threatening the reproducibility of SDI studies.

The SHOnet learning health system: Infrastructure for continuous learning in pediatric rehabilitation

INTRODUCTION

To describe the development and implementation of learning health system (LHS) infrastructure for a pediatric specialty care health system to support LHS research in pediatric rehabilitation settings.

METHODS

An existing pediatric common data model (eg, PEDSnet) of standardized medical terminologies for research was expanded and leveraged for this stud, and applied to SHOnet, a clinical research data resource consisting of deidentified data extracted from the electronic health record (EHR) from the Shriners Hospitals for Children speacialty pediatric health care system. We mapped EHR data for laboratory, procedures, drugs, and conditions to standardized vocabularies including ICD-10, CPT, RxNorm, and LOINC to the common data model using an established extraction-transformation-loading process. Rigorous quality checks were conducted to ensure a high degree of data conformance, completeness, and plausibility. SHOnet data elements from all sources are de-identified and the server is managed by the SHC Information Systems Department. SHOnet data are refreshed monthly and data elements are continually expanded based on new research endeavors.

INTERVENTIONS

Not applicable.

RESULTS

The Shriners Health Outcomes Network (SHOnet) includes data for over 10 000 distinct observational data elements based on over two million patient encounters between 2011 and present.

CONCLUSION

The systematic process to develop SHOnet is replicable and flexible for other pediatric rehabilitation research settings interested in building out their LHS capabilities. Challenges and facilitators may arise for building such LHS infrastructure for rehabilitation in areas of (a) data capture, curation, query, and governance, (b) generating knowledge from data, and (c) dissemination and implementation of new institutional knowledge. Further research studies are needed to evaluate these data resources for scalable system-learning endeavors.SHOnet is an exemplar of an LHS for rehabilitation and specialty care settings. The success of an LHS is dependent on engagement of multiple stakeholders, shared governance, effective knowledge translation, and deep commitment to long-term strategies for engaging clinicians, administration, and families in leveraging knowledge to improve clinical outcomes.

The Norwegian PraksisNett: a nationwide practice-based research network with a novel IT infrastructure

PURPOSE

Clinical research in primary care is relatively scarce. Practice-based research networks (PBRNs) are research infrastructures to overcome hurdles associated with conducting studies in primary care. In Norway, almost all 5.4 million inhabitants have access to a general practitioner (GP) through a patient-list system. This gives opportunity for a PBRN with reliable information about the general population. The aim of the current paper is to describe the establishment, organization and function of PraksisNett (the Norwegian Primary Care Research Network).

MATERIALS AND METHODS

We describe the development, funding and logistics of PraksisNett as a nationwide PBRN.

RESULTS

PraksisNett received funding from the Research Council of Norway for an establishment period of five years (2018-2022). It is comprised of two parts; a human infrastructure (employees, including academic GPs) organized as four regional nodes and a coordinating node and an IT infrastructure comprised by the Snow system in conjunction with the Medrave M4 system. The core of the infrastructure is the 92 general practices that are contractually linked to PraksisNett. These include 492 GPs, serving almost 520,000 patients. Practices were recruited during 2019-2020 and comprise a representative mix of rural and urban settings spread throughout all regions of Norway.

CONCLUSION

Norway has established a nationwide PBRN to reduce hurdles for conducting clinical studies in primary care. Improved infrastructure for clinical studies in primary care is expected to increase the attractiveness for studies on the management of disorders and diseases in primary care and facilitate international research collaboration. This will benefit both patients, GPs and society in terms of improved quality of care.Key pointsPractice-based research networks (PBRNs) are research infrastructures to overcome hurdles associated with conducting studies in primary careImproved infrastructure for clinical studies in primary care is expected to increase the attractiveness for studies on the management of disorders and diseases in primary care and facilitate international research collaborationWe describe PraksisNett, a Norwegian PBRN consisting of 92 general practices including 492 GPs, serving almost 520,000 patientsAn advanced and secure IT infrastructure connects the general practices to PraksisNett and makes it possible to identify and recruit patients in a novel way, as well as reuse clinical dataPraksisNett will benefit both patients, GPs and society in terms of improved quality of careThis paper may inform and inspire initiatives to establish PBRNs elsewhere.

Evaluating Completeness of Discrete Data on Physical Functioning for Children With Cerebral Palsy in a Pediatric Rehabilitation Learning Health System

OBJECTIVE

The purpose of this study was to determine the extent that physical function discrete data elements (DDE) documented in electronic health records (EHR) are complete within pediatric rehabilitation settings.

METHODS

A descriptive analysis on completeness of EHR-based DDEs detailing physical functioning for children with cerebral palsy was conducted. Data from an existing pediatric rehabilitation research learning health system data network, consisting of EHR data from 20 care sites in a pediatric specialty health care system, were leveraged. Completeness was calculated for unique data elements, unique outpatient visits, and unique outpatient records.

RESULTS

Completeness of physical function DDEs was low across 5766 outpatient records (10.5%, approximately 2 DDEs documented). The DDE for Gross Motor Function Classification System level was available for 21% (n = 3746) outpatient visits and 38% of patient records. Ambulation level was the most frequently documented DDE. Intercept only mixed effects models demonstrated that 21.4% and 45% of the variance in completeness for DDEs and the Gross Motor Function Classification System, respectively, across unique patient records could be attributed to factors at the individual care site level.

CONCLUSION

Values of physical function DDEs are missing in designated fields of the EHR infrastructure for pediatric rehabilitation providers. Although completeness appears limited for these DDEs, our observations indicate that data are not missing at random and may be influenced by system-level standards in clinical documentation practices between providers and factors specific to individual care sites. The extent of missing data has significant implications for pediatric rehabilitation quality measurement. More research is needed to understand why discrete data are missing in EHRs and to further elucidate the professional and system-level factors that influence completeness and missingness.

IMPACT

Completeness of DDEs reported in this study is limited and presents a significant opportunity to improve documentation and standards to optimize EHR data for learning health system research and quality measurement in pediatric rehabilitation settings.

Infrastructuring an organizational node for a federated research and data network: A case study from a sociotechnical perspective

Background:

Local nodes on federated research and data networks (FR&DNs) provide enabling infrastructure for collaborative clinical and translational research. Studies in other fields note that infrastructuring, that is, work to identify and negotiate relationships among people, technologies, and organizations, is invisible, unplanned, and undervalued. This may explain the limited literature on nodes in FR&DNs in health care.

Methods:

A retrospective case study of one PCORnet® node explored 3 questions: (1) how were components of infrastructure assembled; (2) what specific work was required; and (3) what theoretically grounded, pragmatic questions should be considered when infrastructuring a node for sustainability. Artifacts, work efforts, and interviews generated during node development and implementation were reviewed. A sociotechnical lens was applied to the analysis. Validity was established with internal and external partners.

Results:

Resources, services, and expertise needed to establish the node existed within the organization, but were scattered across work units. Aligning, mediating, and institutionalizing for sustainability among network and organizational teams, governance, and priorities consumed more work efforts than deploying technical aspects of the node. A theoretically based set of questions relevant to infrastructuring a node was developed and organized within a framework of infrastructuring emphasizing enacting technology, organizing work, and institutionalizing; validity was established with internal and external partners.

Conclusions:

FR&DNs are expanding; we provide a sociotechnical perspective on infrastructuring a node. Future research should evaluate the applicability of the framework and questions to other node and network configurations, and more broadly the infrastructuring required to enable and support federated clinical and translational science.

Identifying strategies to overcome roadblocks to utilising near real-time healthcare and administrative data to create a Scotland-wide learning health system

Creating a learning health system could help reduce variations in quality of care. Success is dependent on timely access to health data. To explore the barriers and facilitators to timely access to patients' data, we conducted in-depth semi-structured interviews with 37 purposively sampled participants from government, the NHS and academia across Scotland. Interviews were analysed using the framework approach. Participants were of the view that Scotland could play a leading role in the exploitation of routine data to drive forward service improvements, but highlighted major impediments: (i) persistence of paper-based records and a variety of information systems; (ii) the need for a proportionate approach to managing information governance; and (iii) the need for support structures to facilitate accrual, processing, linking, analysis and timely use and reuse of data for patient benefit. There is a pressing need to digitise and integrate existing health information infrastructures, guided by a nationwide proportionate information governance approach and the need to enhance technological and human capabilities to support these efforts.

The Indiana Learning Health System Initiative: Early experience developing a collaborative, regional learning health system

INTRODUCTION

Learning health systems (LHSs) are usually created and maintained by single institutions or healthcare systems. The Indiana Learning Health System Initiative (ILHSI) is a new multi-institutional, collaborative regional LHS initiative led by the Regenstrief Institute (RI) and developed in partnership with five additional organizations: two Indiana-based health systems, two schools at Indiana University, and our state-wide health information exchange. We report our experiences and lessons learned during the initial 2-year phase of developing and implementing the ILHSI.

METHODS

The initial goals of the ILHSI were to instantiate the concept, establish partnerships, and perform LHS pilot projects to inform expansion. We established shared governance and technical capabilities, conducted a literature review-based and regional environmental scan, and convened key stakeholders to iteratively identify focus areas, and select and implement six initial joint projects.

RESULTS

The ILHSI successfully collaborated with its partner organizations to establish a foundational governance structure, set goals and strategies, and prioritize projects and training activities. We developed and deployed strategies to effectively use health system and regional HIE infrastructure and minimize information silos, a frequent challenge for multi-organizational LHSs. Successful projects were diverse and included deploying a Fast Healthcare Interoperability Standards (FHIR)-based tool across emergency departments state-wide, analyzing free-text elements of cross-hospital surveys, and developing models to provide clinical decision support based on clinical and social determinants of health. We also experienced organizational challenges, including changes in key leadership personnel and varying levels of engagement with health system partners, which impacted initial ILHSI efforts and structures. Reflecting on these early experiences, we identified lessons learned and next steps.

CONCLUSIONS

Multi-organizational LHSs can be challenging to develop but present the opportunity to leverage learning across multiple organizations and systems to benefit the general population. Attention to governance decisions, shared goal setting and monitoring, and careful selection of projects are important for early success.

Fake It Till You Make It: Guidelines for Effective Synthetic Data Generation

Synthetic data provides a privacy protecting mechanism for the broad usage and sharing of healthcare data for secondary purposes. It is considered a safe approach for the sharing of sensitive data as it generates an artificial dataset that contains no identifiable information. Synthetic data is increasing in popularity with multiple synthetic data generators developed in the past decade, yet its utility is still a subject of research. This paper is concerned with evaluating the effect of various synthetic data generation and usage settings on the utility of the generated synthetic data and its derived models. Specifically, we investigate (i) the effect of data pre-processing on the utility of the synthetic data generated, (ii) whether tuning should be applied to the synthetic datasets when generating supervised machine learning models, and (iii) whether sharing preliminary machine learning results can improve the synthetic data models. Lastly, (iv) we investigate whether one utility measure (Propensity score) can predict the accuracy of the machine learning models generated from the synthetic data when employed in real life. We use two popular measures of synthetic data utility, propensity score and classification accuracy, to compare the different settings. We adopt a recent mechanism for the calculation of propensity, which looks carefully into the choice of model for the propensity score calculation. Accordingly, this paper takes a new direction with investigating the effect of various data generation and usage settings on the quality of the generated data and its ensuing models. The goal is to inform on the best strategies to follow when generating and using synthetic data.

Does an interactive trust-enhanced electronic consent improve patient experiences when asked to share their health records for research? A randomized trial

OBJECTIVE

In the context of patient broad consent for future research uses of their identifiable health record data, we compare the effectiveness of interactive trust-enhanced e-consent, interactive-only e-consent, and standard e-consent (no interactivity, no trust enhancement).

MATERIALS AND METHODS

A randomized trial was conducted involving adult participants making a scheduled primary care visit. Participants were randomized into 1 of the 3 e-consent conditions. Primary outcomes were patient-reported satisfaction with and subjective understanding of the e-consent. Secondary outcomes were objective knowledge, perceived voluntariness, trust in medical researchers, consent decision, and time spent using the application. Outcomes were assessed immediately after use of the e-consent and at 1-week follow-up.

RESULTS

Across all conditions, participants (N = 734) reported moderate-to-high satisfaction with consent (mean 4.3 of 5) and subjective understanding (79.1 of 100). Over 94% agreed to share their health record data. No statistically significant differences in outcomes were observed between conditions. Irrespective of condition, black participants and those with lower education reported lower satisfaction, subjective understanding, knowledge, perceived voluntariness, and trust in medical researchers, as well as spent more time consenting.

CONCLUSIONS

A large majority of patients were willing to share their identifiable health records for research, and they reported positive consent experiences. However, incorporating optional additional information and messages designed to enhance trust in the research process did not improve consent experiences. To improve poorer consent experiences of racial and ethnic minority participants and those with lower education, other novel consent technologies and processes may be valuable. (An Interactive Patient-Centered Consent for Research Using Medical Records; NCT03063268).

Variation in Colorectal Cancer Stage and Mortality across Large Community-Based Populations: PORTAL Colorectal Cancer Cohort

INTRODUCTION

Colorectal cancer (CRC) incidence and mortality can be reduced by effective screening and/or treatment. However, the influence of health care systems on disparities among insured patients is largely unexplored.

METHODS

To evaluate insured patients with CRC diagnosed between 2010 and 2014 across 6 diverse US health care systems in the Patient-Centered Outcomes Research Institute (PCORI) Patient Outcomes Research To Advance Learning (PORTAL) CRC cohort, we contrasted CRC stage; CRC mortality; all-cause mortality; and influences of demographics, stage, comorbidities, and treatment between health systems.

RESULTS

Among 16,211 patients with CRC, there were significant differences between health care systems in CRC stage at diagnosis, CRC-specific mortality, and all-cause mortality. The unadjusted risk of CRC mortality varied from 27% lower to 21% higher than the reference system (hazard ratio [HR] = 0.73, 95% confidence interval = 0.66-0.80 to HR = 1.21, 95% confidence interval = 1.05-1.40; p < 0.01 across systems). Significant differences persisted after adjustment for demographics and comorbidities (p < 0.01); however, adjustment for stage eliminated significant differences (p = 0.24). All-cause mortality among patients with CRC differed approximately 30% between health care systems (HR = 0.89-1.17; p < 0.01). Adjustment for age eliminated significant differences (p = 0.48).

DISCUSSION

Differences in CRC survival between health care systems were largely explained by stage at diagnosis, not demographics, comorbidity, or treatment. Given that stage is strongly related to early detection, these results suggest that variation in CRC screening systems represents a modifiable systems-level factor for reducing disparities in CRC survival.

An Analysis of the Learning Health System in Its First Decade in Practice: Scoping Review

Background

In the past decade, Lynn Etheredge presented a vision for the Learning Health System (LHS) as an opportunity for increasing the value of health care via rapid learning from data and immediate translation to practice and policy. An LHS is defined in the literature as a system that seeks to continuously generate and apply evidence, innovation, quality, and value in health care.

Objective

This review aimed to examine themes in the literature and rhetoric on the LHS in the past decade to understand efforts to realize the LHS in practice and to identify gaps and opportunities to continue to take the LHS forward.

Methods

We conducted a thematic analysis in 2018 to analyze progress and opportunities over time as compared with the initial Knowledge Gaps and Uncertainties proposed in 2007.

Results

We found that the literature on the LHS has increased over the past decade, with most articles focused on theory and implementation; articles have been increasingly concerned with policy.

Conclusions

There is a need for attention to understanding the ethical and social implications of the LHS and for exploring opportunities to ensure that these implications are salient in implementation, practice, and policy efforts.

Physical Activity Assessment and Counseling in Pediatric Clinical Settings

Physical activity plays an important role in children's cardiovascular health, musculoskeletal health, mental and behavioral health, and physical, social, and cognitive development. Despite the importance in children's lives, pediatricians are unfamiliar with assessment and guidance regarding physical activity in children. With the release of the 2018 Physical Activity Guidelines by the US Department of Health and Human Services, pediatricians play a critical role in encouraging physical activity in children through assessing physical activity and physical literacy; providing guidance toward meeting recommendations by children and their families; advocating for opportunities for physical activity for all children in schools, communities, and hospitals; setting an example and remaining physically active personally; advocating for the use of assessment tools and insurance coverage of physical activity and physical literacy screening; and incorporating physical activity assessment and prescription in medical school curricula.

Antiobesity Medication Use in 2.2 Million Adults Across Eight Large Health Care Organizations: 2009-2015

OBJECTIVE

The aim of this study was to examine the prescribing patterns and use of antiobesity medications in a large cohort of patients using data from electronic health records.

METHODS

Pharmacy- and patient-level electronic health record data were obtained on 2,248,407 adults eligible for weight-loss medications from eight geographically dispersed health care organizations.

RESULTS

A total of 29,964 patients (1.3% of total cohort) filled at least one weight-loss medication prescription. This cohort was 82.3% female, with median age 44.9 years and median BMI 37.2 kg/m2 . Phentermine accounted for 76.6% of all prescriptions, with 51.7% of prescriptions being filled for ≥ 120 days and 33.8% filled for ≥ 360 days. There was an increase of 32.9% in medication days for all medications in 2015 compared with 2009. Higher prescription rates were observed in women, black patients, and patients in higher BMI classes. Of 3,919 providers who wrote at least one filled prescription, 23.8% (n = 863) were "frequent prescribers" who wrote 89.6% of all filled prescriptions.

CONCLUSIONS

Weight-loss medications are rarely prescribed to eligible patients. Phentermine accounted for > 75% of all medication days, with a majority of patients filling it for more than 4 months. Less than one-quarter of prescribing providers accounted for approximately 90% of all prescriptions.

Racial/Ethnic Disparities in the Prevalence of Diabetes and Prediabetes by BMI: Patient Outcomes Research To Advance Learning (PORTAL) Multisite Cohort of Adults in the U.S

OBJECTIVE

To examine racial/ethnic disparities in the prevalence of diabetes and prediabetes by BMI category.

RESEARCH DESIGN AND METHODS

In a consortium of three U.S. integrated health care systems, 4,906,238 individuals aged ≥20 years during 2012-2013 were included. Diabetes and prediabetes were ascertained by diagnosis and laboratory results; antihyperglycemic medications were also included for diabetes ascertainment.

RESULTS

The age-standardized diabetes and prediabetes prevalence estimates were 15.9% and 33.4%, respectively. Diabetes but not prediabetes prevalence increased across BMI categories among all racial/ethnic groups (P for trend < 0.001). Racial/ethnic minorities reached a given diabetes prevalence at lower BMIs than whites; Hawaiians/Pacific Islanders and Asians had a diabetes prevalence of 24.6% (95% CI 24.1-25.2%) in overweight and 26.5% (26.3-26.8%) in obese class 1, whereas whites had a prevalence of 23.7% (23.5-23.8%) in obese class 2. The age-standardized prediabetes prevalence estimates in overweight among Hispanics (35.6% [35.4-35.7%]), Asians (38.1% [38.0-38.3%]), and Hawaiians/Pacific Islanders (37.5% [36.9-38.2%]) were similar to those in obese class 4 among whites (35.3% [34.5-36.0%]), blacks (36.8% [35.5-38.2%]), and American Indians/Alaskan Natives (34.2% [29.6-38.8%]). In adjusted models, the strength of association between BMI and diabetes was highest among whites (relative risk comparing obese class 4 with normal weight 7.64 [95% CI 7.50-7.79]) and lowest among blacks (3.16 [3.05-3.27]). The association between BMI and prediabetes was less pronounced.

CONCLUSIONS

Racial/ethnic minorities had a higher burden of diabetes and prediabetes at lower BMIs than whites, suggesting the role of factors other than obesity in racial/ethnic disparities in diabetes and prediabetes risk and highlighting the need for tailored screening and prevention strategies.

A framework for value-creating learning health systems

BACKGROUND

Interest in value-based healthcare, generally defined as providing better care at lower cost, has grown worldwide, and learning health systems (LHSs) have been proposed as a key strategy for improving value in healthcare. LHSs are emerging around the world and aim to leverage advancements in science, technology and practice to improve health system performance at lower cost. However, there remains much uncertainty around the implementation of LHSs and the distinctive features of these systems. This paper presents a conceptual framework that has been developed in Canada to support the implementation of value-creating LHSs.

METHODS

The framework was developed by an interdisciplinary team at the Institut national d'excellence en santé et en services sociaux (INESSS). It was informed by a scoping review of the scientific and grey literature on LHSs, regular team discussions over a 14-month period, and consultations with Canadian and international experts.

RESULTS

The framework describes four elements that characterise LHSs, namely (1) core values, (2) pillars and accelerators, (3) processes and (4) outcomes. LHSs embody certain core values, including an emphasis on participatory leadership, inclusiveness, scientific rigour and person-centredness. In addition, values such as equity and solidarity should also guide LHSs and are particularly relevant in countries like Canada. LHS pillars are the infrastructure and resources supporting the LHS, whereas accelerators are those specific structures that enable more rapid learning and improvement. For LHSs to create value, such infrastructures must not only exist within the ecosystem but also be connected and aligned with the LHSs' strategic goals. These pillars support the execution, routinisation and acceleration of learning cycles, which are the fundamental processes of LHSs. The main outcome sought by executing learning cycles is the creation of value, which we define as the striking of a more optimal balance of impacts on patient and provider experience, population health and health system costs.

CONCLUSIONS

Our framework illustrates how the distinctive structures, processes and outcomes of LHSs tie together with the aim of optimising health system performance and delivering greater value in health systems.

Treatment patterns and survival differ between early-onset and late-onset colorectal cancer patients: the patient outcomes to advance learning network

PURPOSE

Our objective was to describe differences in treatment patterns and survival between early-onset (< 50 years old) and late-onset colorectal cancer (CRC) patients in community-based health systems.

METHODS

We used tumor registry and electronic health record data to identify and characterize patients diagnosed with adenocarcinoma of the colon or rectum from 2010 to 2014 at six US health systems in the patient outcomes to advance learning (PORTAL) network. We used logistic regression to estimate adjusted odds ratios (ORs) and 95% confidence intervals (CIs) comparing the distribution of tumor characteristics and treatment patterns in early-onset versus late-onset CRC. Cox regression models were used to estimate adjusted hazard ratios (HRs) and CIs comparing survival between early- and late-onset CRC patients.

RESULTS

There were 1,424 early-onset and 10,810 late-onset CRC cases in our analyses. Compared to late-onset CRC, early-onset CRC was significantly associated with advanced-stage disease, high-grade histology, signet ring histology, and rectal or left colon location. After adjusting for differences in tumor and patient characteristics, early-onset patients were more likely than late-onset patients to have > 12 lymph nodes examined (OR 1.60, CI 1.37-1.87), to receive systemic therapy (chemotherapy or immunotherapy) within 6 months of diagnosis (OR 2.84, CI 2.40-3.37), and to have a reduced risk of CRC-specific death (HR 0.66, CI 0.56-0.79).

CONCLUSIONS

Early-onset CRC is associated with aggressive tumor characteristics, distal location, and systemic therapy use. Despite some adverse risk factors, these patients tend to have better survival than older onset patients.

A Data Element-Function Conceptual Model for Data Quality Checks

Introduction:

In aggregate, existing data quality (DQ) checks are currently represented in heterogeneous formats, making it difficult to compare, categorize, and index checks. This study contributes a data element-function conceptual model to facilitate the categorization and indexing of DQ checks and explores the feasibility of leveraging natural language processing (NLP) for scalable acquisition of knowledge of common data elements and functions from DQ checks narratives.

Methods:

The model defines a “data element”, the primary focus of the check, and a “function”, the qualitative or quantitative measure over a data element. We applied NLP techniques to extract both from 172 checks for Observational Health Data Sciences and Informatics (OHDSI) and 3,434 checks for Kaiser Permanente’s Center for Effectiveness and Safety Research (CESR).

Results:

The model was able to classify all checks. A total of 751 unique data elements and 24 unique functions were extracted. The top five frequent data element-function pairings for OHDSI were Person-Count (55 checks), Insurance-Distribution (17), Medication-Count (16), Condition-Count (14), and Observations-Count (13); for CESR, they were Medication-Variable Type (175), Medication-Missing (172), Medication-Existence (152), Medication-Count (127), and Socioeconomic Factors-Variable Type (114).

Conclusions:

This study shows the efficacy of the data element-function conceptual model for classifying DQ checks, demonstrates early promise of NLP-assisted knowledge acquisition, and reveals the great heterogeneity in the focus in DQ checks, confirming variation in intrinsic checks and use-case specific “fitness-for-use” checks.

Safety and Effectiveness of Longer-Term Phentermine Use: Clinical Outcomes from an Electronic Health Record Cohort

OBJECTIVE

The aim of this work was to study weight loss and risk of cardiovascular disease (CVD) or death associated with longer-term phentermine use.

METHODS

Using electronic health record data, 13,972 adults were identified with a first phentermine fill in 2010 to 2015, creating exposure categories according to a patient's duration of use (referent: ≤ 3 months). Multivariable linear models were used to compare percent weight loss across categories at 6, 12, and 24 months, and Cox proportional hazards models were used to compare risk of composite CVD or death, up to 3 years after starting phentermine.

RESULTS

The cohort was 84% female and 45% white, with a mean (SD) baseline age 43.5 (10.7) years and BMI of 37.8 (7.2) kg/m² . In multivariable models, longer-term users of phentermine experienced more weight loss; patients using continuously for > 12 months lost 7.4% more than the referent group at 24 months (P < 0.001). The composite CVD or death outcome was rare (0.3%, 41 events), with no significant difference in hazard ratios between groups.

CONCLUSIONS

Greater weight loss without increased risk of incident CVD or death was observed in patients using phentermine monotherapy for longer than 3 months. Despite the limitations of the observational design, this study supports the effectiveness and safety of longer-term phentermine use for low-risk individuals.

An environmental scan of shared access to patient portals

We sought to understand the comprehensiveness of consumer-oriented information describing the availability of shared access to adult patient portals from publicly reported information on institutional websites of 20 large and geographically diverse health systems. All 20 health systems reported that they offer patients the ability to share access to their patient portal account with a family member or friend; however, the comprehensiveness of information regarding registration procedures, features, and terminology varied widely. Half of the systems (n = 10) reported having shared access available on their patient portal registration webpage. Few systems (n = 2) reported affording patients the ability to differentiate specific role-based privileges. No systems reported uptake of shared access among adult patients, which was variably described as "proxy," "caregiver," "parental," or "delegate" access. Findings suggest that engaging families through health information technology will require greater efforts to promote awareness and differentiate privileges that respect patients' choice and control in information-sharing preferences.

Exercise is Medicine as a Vital Sign: Challenges and Opportunities

Exercise is widely recognized as a critical component of a healthy lifestyle by multiple scientific and regulatory bodies. Yet only 35% of adults in the United States even report engaging in regular physical activity. In response to this challenge, the Exercise is Medicine (EIM) initiative encourages clinicians in all settings to regularly assess and counsel patients on physical activity in an attempt to reduce levels of inactivity. The purpose of this article is to highlight key challenges to the implementation of the EIM initiative to promote physical activity and to identify potential opportunities to overcome these challenges. EIM is a global health initiative that endorses "exercise" to be a "vital sign" in all clinical settings. Specifically to assess and compare the patient's physical activity level to national guidelines for physical activity and then promote discussions or referrals based on results. Since its 2007 launch, advocates of the EIM initiative have increased global awareness to promote the EIM mission of integrating and implementing physical activity into healthcare systems, local communities, and individuals' daily lives. Physical activity is fundamental to the overall health of all people across all geographic locations and throughout the lifespan. Promoting physical activity as a healthy lifestyle change has specific challenges but necessary to decrease health disparities related to inactivity.

Health care improvement and survivorship priorities of colorectal cancer survivors: findings from the PORTAL colorectal cancer cohort survey

PURPOSE

Few population-level surveys have explored patient-centered priorities for improving colorectal cancer survivors' care. Working with patients, we designed a survey to identify care improvement and survivorship priorities.

METHODS

We surveyed a random sample of 4000 patients from a retrospective, population-based cohort of colorectal cancer survivors diagnosed during 2010-2014. The survey included two multiple response questions: "What would you have changed about your cancer diagnosis and treatment experience?" and "What are your biggest health or lifestyle concerns (other than having cancer) since being diagnosed?" Multivariable regression identified characteristics associated with endorsement of health care experience and survivorship concerns.

RESULTS

Survey response rate was 50.2% (2000/3986). Fifty-three percent reported at least one unmet need, most commonly for more information about life after treatment (26.7%). Survivors of rectal cancer reported more needs than respondents with colon cancer; persons of color reported more needs than non-Hispanic whites; individuals without high school diplomas reported more needs than individuals with more education. Fear of recurrence was the most common health/lifestyle concern (58.9%). Respondents under age 65 reported nearly all health/lifestyle concerns more often than respondents over age 74. Rectal cancer survivors reported more concerns about activity limitation, changes, and body function and appearance than colon cancer survivors. Persons of color were more likely to report financial concerns than non-Hispanic whites.

CONCLUSIONS

The greatest needs for intervention are among survivors of rectal cancer, survivors of minority racial/ethnic background, and survivors of younger age. Survivors with low educational attainment and those with higher stage disease could also benefit.

Associations of overweight/obesity and socioeconomic status with hypertension prevalence across racial and ethnic groups

Racial/ethnic disparities in the prevalence of diagnosed hypertension are persistent but may be partially explained by racial/ethnic differences in weight category and neighborhood socioeconomic status. The authors compared hypertension prevalence rates among 4 060 585 adults with overweight or obesity across 10 healthcare systems by weight category and neighborhood education level in geographically and racially diverse individuals. Data were obtained from electronic health records. Hypertension was defined as at least two outpatient visits or one inpatient hospitalization with a coded diagnosis. Logistic regression, adjusted for age, sex, and site, with two-way interactions between race/ethnicity and weight category or neighborhood education, was used to examine the association between hypertension and race/ethnicity, with whites as the reference. Results documented that odds ratios for hypertension prevalence were greater for blacks, American Indians/Alaskan Natives, Asians, and Native Hawaiians/other Pacific Islanders compared with whites and lower for Hispanics in similar weight categories and neighborhood education levels. Although two-way interactions were statistically significant, the magnitude of the odds of hypertension compared with whites did not substantially vary across weight or neighborhood education. Hypertension odds were almost double relative to whites for blacks and Native Hawaiians/other Pacific Islanders across most weight categories and all neighborhood education levels. Odds of hypertension were about 50% greater for Asians relative to whites across weight categories. Results suggest that other factors might be associated with racial/ethnic disparities in hypertension. More research is needed to understand the many factors that may contribute to variation in diagnosed hypertension across racial/ethnic groups with overweight or obesity.

Monitoring Depression Rates in an Urban Community: Use of Electronic Health Records

Objectives:

Depression is the most common mental health disorder and mediates outcomes for many chronic diseases. Ability to accurately identify and monitor this condition, at the local level, is often limited to estimates from national surveys. This study sought to compare and validate electronic health record (EHR)-based depression surveillance with multiple data sources for more granular demographic subgroup and subcounty measurements.

Design/Setting:

A survey compared data sources for the ability to provide subcounty (eg, census tract [CT]) depression prevalence estimates. Using 2011-2012 EHR data from 2 large health care providers, and American Community Survey data, depression rates were estimated by CT for Denver County, Colorado. Sociodemographic and geographic (residence) attributes were analyzed and described. Spatial analysis assessed for clusters of higher or lower depression prevalence.

Main Outcome Measure(s):

Depression prevalence estimates by CT.

Results:

National and local survey-based depression prevalence estimates ranged from 7% to 17% but were limited to county level. Electronic health record data provided subcounty depression prevalence estimates by sociodemographic and geographic groups (CT range: 5%-20%). Overall depression prevalence was 13%; rates were higher for women (16% vs men 9%), whites (16%), and increased with age and homeless patients (18%). Areas of higher and lower EHR-based, depression prevalence were identified.

Conclusions:

Electronic health record–based depression prevalence varied by CT, gender, race/ethnicity, age, and living status. Electronic health record–based surveillance complements traditional methods with greater timeliness and granularity. Validation through subcounty-level qualitative or survey approaches should assess accuracy and address concerns about EHR selection bias. Public health agencies should consider the opportunity and evaluate EHR system data as a surveillance tool to estimate subcounty chronic disease prevalence.

Patient preferences toward an interactive e-consent application for research using electronic health records

OBJECTIVE

The purpose of this study was to assess patient perceptions of using an interactive electronic consent (e-consent) application when deciding whether or not to grant broad consent for research use of their identifiable electronic health record (EHR) information.

MATERIALS AND METHODS

For this qualitative study, we conducted a series of 42 think-aloud interviews with 32 adults. Interview transcripts were coded and analyzed using a modified grounded theory approach.

RESULTS

We identified themes related to patient preferences, reservations, and mixed attitudes toward consenting electronically; low- and high-information-seeking behavior; and an emphasis on reassuring information, such as data protections and prohibitions against sharing data with pharmaceutical companies. Participants expressed interest in the types of information contained in their EHRs, safeguards protecting EHR data, and specifics on studies that might use their EHR data.

DISCUSSION

This study supports the potential value of interactive e-consent applications that allow patients to customize their consent experience. This study also highlights that some people have concerns about e-consent platforms and desire more detailed information about administrative processes and safeguards that protect EHR data used in research.

CONCLUSION

This study contributes new insights on how e-consent applications could be designed to ensure that patients' information needs are met when seeking consent for research use of health record information. Also, this study offers a potential electronic approach to meeting the new Common Rule requirement that consent documents contain a "concise and focused" presentation of key information followed by more details.

Quality Requirements for Electronic Health Record Systems*. A Japanese-German Information Management Perspective

BACKGROUND

For more than 30 years, there has been close cooperation between Japanese and German scientists with regard to information systems in health care. Collaboration has been formalized by an agreement between the respective scientific associations. Following this agreement, two joint workshops took place to explore the similarities and differences of electronic health record systems (EHRS) against the background of the two national healthcare systems that share many commonalities.

OBJECTIVES

To establish a framework and requirements for the quality of EHRS that may also serve as a basis for comparing different EHRS.

METHODS

Donabedian's three dimensions of quality of medical care were adapted to the outcome, process, and structural quality of EHRS and their management. These quality dimensions were proposed before the first workshop of EHRS experts and enriched during the discussions.

RESULTS

The Quality Requirements Framework of EHRS (QRF-EHRS) was defined and complemented by requirements for high quality EHRS. The framework integrates three quality dimensions (outcome, process, and structural quality), three layers of information systems (processes and data, applications, and physical tools) and three dimensions of information management (strategic, tactical, and operational information management).

CONCLUSIONS

Describing and comparing the quality of EHRS is in fact a multidimensional problem as given by the QRF-EHRS framework. This framework will be utilized to compare Japanese and German EHRS, notably those that were presented at the second workshop.

Toward an Information Infrastructure for Global Health Improvement

Profound global challenges to individual and population health, alongside the opportunities to benefit from digital technology, have spawned the concept of the Learning Health System. Learning Health Systems (LHSs)--which can function at organizational, network, regional, and national levels of scale--have the capability of continuous data-driven self-study that promotes change and improvement. The LHS concept, which originated in the U.S. in 2007, is rapidly gaining attention around the world. LHSs require, but also transcend, the secondary use of health data. This paper describes the key features of LHSs, argues that effective and sustainable LHSs must be supported by infrastructures that allow them to function with economies of scale and scope, and describes the services that such infrastructures must provide. While it is relatively straightforward to describe LHSs, achieving them at the high level of capability necessary to promote significant health benefits will require advancements in science and engineering, engaging the field of informatics among a wider range of disciplines. It also follows from this vision that LHSs cannot be built from an imposed blueprint; LHSs will more likely evolve from efforts at smaller scales that compose into larger systems.

The Colorado Collaborative for Nursing Research: Providing Trended Nurse-Sensitive Patient Outcome Data Across Entities

Nurse leaders lack timely access to trended electronic health record (EHR) data to drive decision making. Robust nurse-sensitive patient outcome data are difficult to locate in EHRs and largely absent across entities. The Colorado Collaborative for Nursing Research is currently testing a federated data system to get nurse leaders the information they need, when they need it.

Mental illness in bariatric surgery: A cohort study from the PORTAL network

OBJECTIVE

To compare bariatric surgery outcomes according to preoperative mental illness category.

METHODS

Electronic health record data from several US healthcare systems were used to compare outcomes of four groups of patients who underwent bariatric surgery in 2012 and 2013. These included the following: people with (1) no mental illness, (2) mild-to-moderate depression or anxiety, (3) severe depression or anxiety, and (4) bipolar, psychosis, or schizophrenia spectrum disorders. Groups were compared on weight loss trajectory using generalized estimating equations using B-spline bases and on all-cause emergency department visits and hospital days using zero-inflated Poisson and negative binomial regression up to 2 years after surgery. Models were adjusted for demographic and health covariates, including baseline healthcare use.

RESULTS

Among 8,192 patients, mean age was 44.3 (10.7) years, 79.9% were female, and 45.6% were white. Fifty-seven percent had preoperative mental illness. There were no differences between groups for weight loss, but patients with preoperative severe depression or anxiety or bipolar, psychosis, or schizophrenia spectrum disorders had higher follow-up levels of emergency department visits and hospital days compared to those with no mental illness.

CONCLUSIONS

In this multicenter study, mental illness was not associated with differential weight loss after bariatric surgery, but additional research could focus on reducing acute care use among these patients.

Cardiometabolic Risk Factors Among 1.3 Million Adults With Overweight or Obesity, but Not Diabetes, in 10 Geographically Diverse Regions of the United States, 2012-2013

INTRODUCTION

Various phenotypes of overweight and obesity pose various health risks. The objective of this study was to determine the prevalence of 4 commonly measured cardiometabolic risk factors (CRFs) among adults with overweight or obesity, but not diabetes, at the time of the study.

METHODS

We analyzed data for 1,294,174 adults (aged ≥20 y) who were members of one of 4 integrated health systems. Each cohort member had a body mass index in 2012 or 2013 that indicated overweight or obesity. We determined the presence of 4 CRFs within 1 year of the first BMI measurement: elevated blood pressure (systolic ≥130 mm Hg or diastolic >85 mm Hg or ICD-9-CM [International Classification of Diseases, Ninth Revision, Clinical Modification] diagnosis code 401.0-405.9); elevated triglycerides (≥150 mg/dL or ICD-9-CM 272.1); low high-density lipoprotein cholesterol (<40 mg/dL for men or <50 mg/dL for women or ICD-9-CM 272.5); and prediabetes (fasting glucose 100-125 mg/dL or HbA_1c 5.7%-6.4% or ICD-9-CM 790.2x). We tested the risk of having 1 or more CRFs after adjusting for obesity class and demographic characteristics with multivariable logistic regression.

RESULTS

Among participants with overweight (52.5% of the sample), 18.6% had none of the 4 CRFs. Among the 47.5% of participants with obesity, 9.6% had none; among participants with morbid obesity, 5.8% had none. Age was strongly associated with CRFs in multivariable analysis.

CONCLUSION

Almost 10% of participants with obesity had no CRFs. Overweight or obesity increases cardiometabolic risk, but the number and type of CRFs varied substantially by age, even among participants with morbid obesity.

Clinical Research Informatics for Big Data and Precision Medicine

OBJECTIVES

To reflect on the notable events and significant developments in Clinical Research Informatics (CRI) in the year of 2015 and discuss near-term trends impacting CRI.

METHODS

We selected key publications that highlight not only important recent advances in CRI but also notable events likely to have significant impact on CRI activities over the next few years or longer, and consulted the discussions in relevant scientific communities and an online living textbook for modern clinical trials. We also related the new concepts with old problems to improve the continuity of CRI research.

RESULTS

The highlights in CRI in 2015 include the growing adoption of electronic health records (EHR), the rapid development of regional, national, and global clinical data research networks for using EHR data to integrate scalable clinical research with clinical care and generate robust medical evidence. Data quality, integration, and fusion, data access by researchers, study transparency, results reproducibility, and infrastructure sustainability are persistent challenges.

CONCLUSION

The advances in Big Data Analytics and Internet technologies together with the engagement of citizens in sciences are shaping the global clinical research enterprise, which is getting more open and increasingly stakeholder-centered, where stakeholders include patients, clinicians, researchers, and sponsors.

Interacting effects of obesity, race, ethnicity and sex on the incidence and control of adult-onset asthma

Background

To improve care and control for patients with adult-onset asthma, a better understanding of determinants of their risk and outcomes is important. We investigated how associations between asthma, asthma control and obesity may be modified by patient demographic characteristics.

Methods

This retrospective study of adults enrolled in several health plans across the U.S. (n = 2,860,305) examined the interacting effects of obesity, age, race, and sex on adult-onset asthma and asthma control. Multivariable adjusted Cox and logistic regression models estimated hazard ratios (HR), and 95 % confidence intervals (CI) for the associations between body mass index (BMI) and study outcomes, and interactions of BMI with demographic characteristics.

Results

Compared with individuals who had a BMI <25 kg/m², the hazard of adult-onset asthma progressively increased with increasing BMI, from a 12 % increase among persons with a BMI of 25.0–29.9 kg/m² (HR 1.12, 95 % CI 1.10, 1.14) to an almost 250 % increase among persons with a BMI ≥50 kg/m² (HR 2.49, 95 % CI 2.38, 2.60). The magnitude of the association between obesity and asthma risk was greater for women (compared with men) and lower for Blacks (compared with non-Hispanic Whites). Among individuals with asthma, obesity was associated with poorly controlled and high-risk asthma.

Conclusions

The present study demonstrates that the magnitude of the associations between obesity and adult-onset asthma incidence and control are modified by race, age, and sex. Understanding the role of obesity in the development of adult-onset asthma will help to improve asthma treatment algorithms and to develop targeted interventions.

The Learning Healthcare System: Where are we now? A systematic review

BACKGROUND AND OBJECTIVE

The Learning Healthcare System paradigm has attracted the attention of researchers worldwide. The great potential originating from high-scale health data reuse and the inclusion of patient perspectives into care models promises personalized care, lower costs of health services and minimized consumption of resources. The aim of this review is to summarize the attempts to adopt the novel paradigm, putting emphasis on implementations and evaluating the impact on current medical practices.

METHOD

PRISMA methodology was followed for structuring the review process. Three major research databases (PubMed, IEEE Xplore and ACM DL) were queried with the predefined search terms "learning healthcare" and "learning health". Publications containing specific theoretical or empirical results were considered.

RESULTS

Three hundred and fifty-eight publications were identified; however, only 32 met the inclusion criteria. Nineteen papers were characterized as theoretical contributions, while the rest presented empirical achievements. Only one paper described the initial estimates of impact and economy.

DISCUSSION

Individualistic communication of studies ignoring popular frameworks for assessing and reporting research achievements prevents the systematic generation of knowledge. Evaluating the impact of the Learning Healthcare System instances where it is implemented could work as a catalyst in reaching higher acceptance and adoption of the proposed ideas by healthcare worldwide; however, it mostly remains described in theory.

CONCLUSIONS

The review demonstrated the interest of researchers in exploring the Learning Healthcare System ideas. However, it also revealed minimal focus on evaluating the impact of the novel paradigm on both healthcare service delivery and patient outcome.

Health at hand: A systematic review of smart watch uses for health and wellness

INTRODUCTION

Smart watches have the potential to support health in everyday living by: enabling self-monitoring of personal activity; obtaining feedback based on activity measures; allowing for in-situ surveys to identify patterns of behavior; and supporting bi-directional communication with health care providers and family members. However, smart watches are an emerging technology and research with these devices is at a nascent stage.

METHODS

We conducted a systematic review of smart watch studies that engaged people in their use by searching PubMed, Embase, IEEE XPlore and ACM Digital libraries. Participant demographics, device features, watch applications and methods, and technical challenges were abstracted from included studies.

RESULTS

Seventy-three studies were returned in the search. Seventeen studies published were included. Included studies were published from 2014 to 2016, with the exception of one published in 2011. Most studies employed the use of consumer-grade smart watches (14/17, 82%). Patient-related studies focused on activity monitoring, heart rate monitoring, speech therapy adherence, diabetes self-management, and detection of seizures, tremors, scratching, eating, and medication-taking behaviors. Most patient-related studies enrolled participants with few exclusion criteria to validate smart watch function (10/17, 58%). Only studies that focused on Parkinson's disease, epilepsy, and diabetes management enrolled persons living with targeted conditions. One study focused on nursing work in the ICU and one focused on CPR training for laypeople.

CONCLUSION

Consumer-grade smart watches have penetrated the health research space rapidly since 2014. Smart watch technical function, acceptability, and effectiveness in supporting health must be validated in larger field studies that enroll actual participants living with the conditions these devices target.

Report of the National Heart, Lung, and Blood Institute Working Group: An Integrated Network for Congenital Heart Disease Research

The National Heart, Lung, and Blood Institute convened a working group in January 2015 to explore issues related to an integrated data network for congenital heart disease research. The overall goal was to develop a common vision for how the rapidly increasing volumes of data captured across numerous sources can be managed, integrated, and analyzed to improve care and outcomes. This report summarizes the current landscape of congenital heart disease data, data integration methodologies used across other fields, key considerations for data integration models in congenital heart disease, and the short- and long-term vision and recommendations made by the working group.

Racial and ethnic minority patients report different weight-related care experiences than non-Hispanic Whites

Our objective was to compare patients' health care experiences, related to their weight, across racial and ethnic groups. In Summer 2015, we distributed a written survey with telephone follow-up to a random sample of 5400 racially/ethnically and geographically diverse U.S. adult health plan members with overweight or obesity. The survey assessed members' perceptions of their weight-related healthcare experiences, including their perception of their primary care provider, and the type of weight management services they had been offered, or were interested in. We used multivariable multinomial logistic regression to examine the relationship between race/ethnicity and responses to questions about care experience. Overall, 2811 members (53%) responded to the survey and we included 2725 with complete data in the analysis. Mean age was 52.7 years (SD 15.0), with 61.7% female and 48.3% from minority racial/ethnic groups. Mean BMI was 37.1 kg/m² (SD 8.0). Most (68.2%) respondents reported having previous discussions of weight with their provider, but interest in such counseling varied by race/ethnicity. Non-Hispanic blacks were significantly less likely to frequently avoid care (for fear of discussing weight/being weighed) than whites (OR 0.49, 95% CI 0.26–0.90). Relative to whites, respondents of other race/ethnicities were more likely to want weight-related discussions with their providers. Race/ethnicity correlates with patients' perception of discussions of weight in healthcare encounters. Clinicians should capitalize on opportunities to discuss weight loss with high-risk minority patients who may desire these conversations.

•

Minority patients may desire more information about weight management.

•

Non-Hispanic whites may be more likely to avoid care due to fear of weight bias.

•

Clinicians could conduct more sensitive weight loss discussions with all patients.

Toward a Learning Health-care System - Knowledge Delivery at the Point of Care Empowered by Big Data and NLP

The concept of optimizing health care by understanding and generating knowledge from previous evidence, ie, the Learning Health-care System (LHS), has gained momentum and now has national prominence. Meanwhile, the rapid adoption of electronic health records (EHRs) enables the data collection required to form the basis for facilitating LHS. A prerequisite for using EHR data within the LHS is an infrastructure that enables access to EHR data longitudinally for health-care analytics and real time for knowledge delivery. Additionally, significant clinical information is embedded in the free text, making natural language processing (NLP) an essential component in implementing an LHS. Herein, we share our institutional implementation of a big data-empowered clinical NLP infrastructure, which not only enables health-care analytics but also has real-time NLP processing capability. The infrastructure has been utilized for multiple institutional projects including the MayoExpertAdvisor, an individualized care recommendation solution for clinical care. We compared the advantages of big data over two other environments. Big data infrastructure significantly outperformed other infrastructure in terms of computing speed, demonstrating its value in making the LHS a possibility in the near future.

The Patient Outcomes Research To Advance Learning (PORTAL) Network Adult Overweight and Obesity Cohort: Development and Description

Background

The Patient-Centered Outcomes Research Institute (PCORI) created a new national network infrastructure to enable large-scale observational comparative effectiveness research across diverse clinical care settings. As part of testing the feasibility of this effort, each clinical data research network (CDRN) was required to construct cohorts of patients, including one of patients with overweight and obesity.

Objective

The aim of this paper is to report on the development of the Patient Outcomes Research to Advance Learning (PORTAL) overweight and obese cohort, which includes patients from 10 health plans located across the United States.

Methods

Information was gathered from each plan’s electronic health records (EHR). Eligibility included 18 years of age or older, a valid height and weight in 2012 or 2013, and body mass index (BMI) greater than 22.9 kg/m². Pre-diabetes and diabetes status was defined using the American Diabetes Association (ADA) criteria, using lab values of glycated hemoglobin (HbA1c) or fasting glucose available in the EHR. Hypertension was identified from the International Classification of Diseases (ICD) diagnosis codes. Individuals were classified into BMI categories: healthy weight (23.0-24.9 kg/m²), overweight (25.0-29.9 kg/m²), obese class 1 (30.0-34.9 kg/m²), obese class 2 (35.0-39.9 kg/m²), obese class 3 (40.0-49.0 kg/m²), and obese class 4 (>50.0 kg/m²).

Results

A cohort of 5,293,458 non-pregnant adults was created. Weight status was 20.39% (1,079,289/5,293,458) healthy weight, 40.40% (2,138,520/5,293,458) overweight, 22.78% (1,205,866/5,293,458) obese class 1, 9.86% (521,872/5,293,458) obese class 2, 5.59% (295,786/5,293,458) obese class 3, and 0.98% (52,125/5,293,458) obese class 4. Race/ethnicity was 49.02% (2,594,776/5,293,458) non-Hispanic white, 22.89% (1,211,677/5,293,458) Hispanic, 10.40% (550,608/5,293,458) Asian, 10.83% (573,506/5,293,458) black, and 6.59% (348,830/5,293,458) other. About 34.33% (1,817,438/5,293,458) met the definition of hypertension, 20.49% (1,660,940/5,293,458) of individuals met the criteria for pre-diabetes, and 14.98% (793,069/5,293,458) met criteria for diabetes. Prevalence of pre-diabetes and diabetes varied across health plans to a greater extent than expected based on hypertension prevalence and BMI status variability.

Conclusions

This large, race, ethnic, and geographically diverse cohort will be useful for future studies of rare exposures or outcomes and differences in health care practices.

Building a Governance Strategy for CER: The Patient Outcomes Research to Advance Learning (PORTAL) Network Experience

INTRODUCTION

The Patient Outcomes Research to Advance Learning (PORTAL) Network was established with funding from the Patient-Centered Outcomes Research Institute (PCORI) in 2014. The PORTAL team adapted governance structures and processes from past research network collaborations. We will review and outline the structures and processes of the PORTAL governance approach and describe how proactively focusing on priority areas helped us to facilitate an ambitious research agenda.

BACKGROUND

For years a variety of funders have supported large-scale infrastructure grants to promote the use of clinical datasets to answer important comparative effectiveness research (CER) questions. These awards have provided the impetus for health care systems to join forces in creating clinical data research networks. Often, these scientific networks do not develop governance processes proactively or systematically, and address issues only as problems arise. Even if network leaders and collaborators foresee the need to develop governance approaches, they may underestimate the time and effort required to develop sound processes. The resulting delays can impede research progress.

INNOVATION

Because the PORTAL sites had built trust and a foundation of collaboration by participating with one another in past research networks, essential elements of effective governance such as guiding principles, decision making processes, project governance, data governance, and stakeholders in governance were familiar to PORTAL investigators. This trust and familiarity enabled the network to rapidly prioritize areas that required sound governance approaches: responding to new research opportunities, creating a culture of trust and collaboration, conducting individual studies, within the broader network, assigning responsibility and credit to scientific investigators, sharing data while protecting privacy/security, and allocating resources. The PORTAL Governance Document, complete with a Toolkit of Appendices is included for reference and for adaptation by other networks.

CREDIBILITY

As a result of identifying project-based governance priorities (IRB approval, subcontracting, selection of new research including lead PI and participating sites, and authorship) and data governance priorities (reciprocal data use agreement, analytic plan procedures, and other tools for data governance), PORTAL established most of its governance structure by Month 6 of the 18 month project. This allowed science to progress and collaborators to experience first-hand how the structures and procedures functioned in the remaining 12 months of the project, leaving ample time to refine them and to develop new structures or processes as necessary.

DISCUSSION

The use of procedures and processes with which participating investigators and their home institutions were already familiar allowed project and regulatory requirements to be established quickly to protect patients, their data, and the health care systems that act as stewards for both. As the project progressed, PORTAL was able to test and adjust the structures it put place, and to make substantive revisions by Month 17. As a result, priority processes have been predictable, transparent and effective.

CONCLUSION/NEXT STEPS

Strong governance practices are a stewardship responsibility of research networks to justify the trust of patients, health plan members, health care delivery organizations, and other stakeholders. Well-planned governance can reduce the time necessary to initiate the scientific activities of a network, a particular concern when the time frame to complete research is short. Effective network and data governance structures protect patient and institutional data as well as the interests of investigators and their institutions, and assures that the network has built an environment to meet the goals of the research.

Achieving a High-Quality Cancer Care Delivery System for Older Adults: Innovative Models of Care

OBJECTIVES

To examine innovative models and other research-based interventions that hold potential to assure high-quality care for the growing older adult population living with cancer as one of multiple chronic conditions. Evidence from these care delivery approaches provides a roadmap for the development of future care models.

DATA SOURCES

Published peer-reviewed literature, policy analyses, and web-based resources.

CONCLUSION

Available evidence suggests the need for models that engage patients and their family caregivers, focus on patient's functional capacities, emphasize palliative care, and maximize the contributions of all team members.

IMPLICATIONS FOR NURSING PRACTICE

Nurses are uniquely positioned to lead or play a major role in the evolution and implementation of care models targeting older adults with cancer, but must increase their knowledge and skills related to both oncology and geriatrics to maximize their contributions.

An informatics research agenda to support precision medicine: seven key areas

The recent announcement of the Precision Medicine Initiative by President Obama has brought precision medicine (PM) to the forefront for healthcare providers, researchers, regulators, innovators, and funders alike. As technologies continue to evolve and datasets grow in magnitude, a strong computational infrastructure will be essential to realize PM's vision of improved healthcare derived from personal data. In addition, informatics research and innovation affords a tremendous opportunity to drive the science underlying PM. The informatics community must lead the development of technologies and methodologies that will increase the discovery and application of biomedical knowledge through close collaboration between researchers, clinicians, and patients. This perspective highlights seven key areas that are in need of further informatics research and innovation to support the realization of PM.

Preparing for the ICD-10-CM Transition: Automated Methods for Translating ICD Codes in Clinical Phenotype Definitions

Background:

The national mandate for health systems to transition from ICD-9-CM to ICD-10-CM in October 2015 has an impact on research activities. Clinical phenotypes defined by ICD-9-CM codes need to be converted to ICD-10-CM, which has nearly four times more codes and a very different structure than ICD-9-CM.

Methods:

We used the Centers for Medicare & Medicaid Services (CMS) General Equivalent Maps (GEMs) to translate, using four different methods, condition-specific ICD-9-CM code sets used for pragmatic trials (n=32) into ICD-10-CM. We calculated the recall, precision, and F score of each method. We also used the ICD-9-CM and ICD-10-CM value sets defined for electronic quality measure as an additional evaluation of the mapping methods.

Results:

The forward-backward mapping (FBM) method had higher precision, recall and F-score metrics than simple forward mapping (SFM). The more aggressive secondary (SM) and tertiary mapping (TM) methods resulted in higher recall but lower precision. For clinical phenotype definition, FBM was the best (F=0.67), but was close to SM (F=0.62) and TM (F=0.60), judging on the F-scores alone. The overall difference between the four methods was statistically significant (one-way ANOVA, F=5.749, p=0.001). However, pairwise comparisons between FBM, SM, and TM did not reach statistical significance. A similar trend was found for the quality measure value sets.

Discussion:

The optimal method for using the GEMs depends on the relative importance of recall versus precision for a given use case. It appears that for clinically distinct and homogenous conditions, the recall of FBM is sufficient. The performance of all mapping methods was lower for heterogeneous conditions. Since code sets used for phenotype definition and quality measurement can be very similar, there is a possibility of cross-fertilization between the two activities.

Conclusion:

Different mapping approaches yield different collections of ICD-10-CM codes. All methods require some level of human validation.

Family Caregivers and Consumer Health Information Technology

Health information technology has been embraced as a strategy to facilitate patients' access to their health information and engagement in care. However, not all patients are able to access, or are capable of using, a computer or mobile device. Although family caregivers assist individuals with some of the most challenging and costly health needs, their role in health information technology is largely undefined and poorly understood. This perspective discusses challenges and opportunities of engaging family caregivers through the use of consumer-oriented health information technology. We compile existing evidence to make the case that involving family caregivers in health information technology as desired by patients is technically feasible and consistent with the principles of patient-centered and family-centered care. We discuss how more explicit and purposeful engagement of family caregivers in health information technology could advance clinical quality and patient safety by increasing the transparency, accuracy, and comprehensiveness of patient health information across settings of care. Finally, we describe how clarifying and executing patients' desires to involve family members or friends through health information technology would provide family caregivers greater legitimacy, convenience, and timeliness in health system interactions, and facilitate stronger partnerships between patients, family caregivers, and health care professionals.

Trends in biomedical informatics: automated topic analysis of JAMIA articles

Biomedical Informatics is a growing interdisciplinary field in which research topics and citation trends have been evolving rapidly in recent years. To analyze these data in a fast, reproducible manner, automation of certain processes is needed. JAMIA is a "generalist" journal for biomedical informatics. Its articles reflect the wide range of topics in informatics. In this study, we retrieved Medical Subject Headings (MeSH) terms and citations of JAMIA articles published between 2009 and 2014. We use tensors (i.e., multidimensional arrays) to represent the interaction among topics, time and citations, and applied tensor decomposition to automate the analysis. The trends represented by tensors were then carefully interpreted and the results were compared with previous findings based on manual topic analysis. A list of most cited JAMIA articles, their topics, and publication trends over recent years is presented. The analyses confirmed previous studies and showed that, from 2012 to 2014, the number of articles related to MeSH terms Methods, Organization & Administration, and Algorithms increased significantly both in number of publications and citations. Citation trends varied widely by topic, with Natural Language Processing having a large number of citations in particular years, and Medical Record Systems, Computerized remaining a very popular topic in all years.

R-U policy frontiers for health data de-identification

OBJECTIVE

The Health Insurance Portability and Accountability Act Privacy Rule enables healthcare organizations to share de-identified data via two routes. They can either 1) show re-identification risk is small (e.g., via a formal model, such as k-anonymity) with respect to an anticipated recipient or 2) apply a rule-based policy (i.e., Safe Harbor) that enumerates attributes to be altered (e.g., dates to years). The latter is often invoked because it is interpretable, but it fails to tailor protections to the capabilities of the recipient. The paper shows rule-based policies can be mapped to a utility (U) and re-identification risk (R) space, which can be searched for a collection, or frontier, of policies that systematically trade off between these goals.

METHODS

We extend an algorithm to efficiently compose an R-U frontier using a lattice of policy options. Risk is proportional to the number of patients to which a record corresponds, while utility is proportional to similarity of the original and de-identified distribution. We allow our method to search 20 000 rule-based policies (out of 2(700)) and compare the resulting frontier with k-anonymous solutions and Safe Harbor using the demographics of 10 U.S. states.

RESULTS

The results demonstrate the rule-based frontier 1) consists, on average, of 5000 policies, 2% of which enable better utility with less risk than Safe Harbor and 2) the policies cover a broader spectrum of utility and risk than k-anonymity frontiers.

CONCLUSIONS

R-U frontiers of de-identification policies can be discovered efficiently, allowing healthcare organizations to tailor protections to anticipated needs and trustworthiness of recipients.

A system to build distributed multivariate models and manage disparate data sharing policies: implementation in the scalable national network for effectiveness research

Background Centralized and federated models for sharing data in research networks currently exist. To build multivariate data analysis for centralized networks, transfer of patient-level data to a central computation resource is necessary. The authors implemented distributed multivariate models for federated networks in which patient-level data is kept at each site and data exchange policies are managed in a study-centric manner.

Objective The objective was to implement infrastructure that supports the functionality of some existing research networks (e.g., cohort discovery, workflow management, and estimation of multivariate analytic models on centralized data) while adding additional important new features, such as algorithms for distributed iterative multivariate models, a graphical interface for multivariate model specification, synchronous and asynchronous response to network queries, investigator-initiated studies, and study-based control of staff, protocols, and data sharing policies.

Materials and Methods Based on the requirements gathered from statisticians, administrators, and investigators from multiple institutions, the authors developed infrastructure and tools to support multisite comparative effectiveness studies using web services for multivariate statistical estimation in the SCANNER federated network.

Results The authors implemented massively parallel (map-reduce) computation methods and a new policy management system to enable each study initiated by network participants to define the ways in which data may be processed, managed, queried, and shared. The authors illustrated the use of these systems among institutions with highly different policies and operating under different state laws.

Discussion and Conclusion Federated research networks need not limit distributed query functionality to count queries, cohort discovery, or independently estimated analytic models. Multivariate analyses can be efficiently and securely conducted without patient-level data transport, allowing institutions with strict local data storage requirements to participate in sophisticated analyses based on federated research networks.

Federated queries of clinical data repositories: Scaling to a national network

Federated networks of clinical research data repositories are rapidly growing in size from a handful of sites to true national networks with more than 100 hospitals. This study creates a conceptual framework for predicting how various properties of these systems will scale as they continue to expand. Starting with actual data from Harvard's four-site Shared Health Research Information Network (SHRINE), the framework is used to imagine a future 4000 site network, representing the majority of hospitals in the United States. From this it becomes clear that several common assumptions of small networks fail to scale to a national level, such as all sites being online at all times or containing data from the same date range. On the other hand, a large network enables researchers to select subsets of sites that are most appropriate for particular research questions. Developers of federated clinical data networks should be aware of how the properties of these networks change at different scales and design their software accordingly.