OBJECTIVES/SPECIFIC AIMS: Our goal is to explore the value of learning algorithms to improve both the efficiency and accuracy of a clinician undertaking the cognitive task of selecting the best resuscitative intervention for a hemodynamically unstable patient in the ICU. Machine learning is an ideal discipline to solve this problem. The ICU is a data rich environment, however there is significant uncertainty regarding the interdependency of this data. Experts consistently struggle to develop deterministic models of the underlying forces driving hemodynamic perturbations and intervention responsiveness. Machine learning, especially deep learning, assumes no correlation between inputs. Deep architectures disentangle these high-level relationships through exposure to abundant, diverse data sets such as those used in this project, obviating the need to manually explore confounding interactions. METHODS/STUDY POPULATION: We are using the “Medical Information Mart for Intensive Care” (MIMIC-III) database for this project. MIMIC-III is a large, single-center database comprising information relating to patients admitted to critical care units at Beth Israel Deaconess Medical Center, a large tertiary care hospital, from 2001 to 2012. It contains data associated with 38,597 distinct adult patients and 53,423 distinct hospital admissions for those patients, with a mean of 4579 charted observations and 380 laboratory measurements available for each hospital admission. Classes of data in the MIMIC-III are varied and include billing, intervention, laboratory, medication, and physiologic data among others. In addition to training an RNN in the task of predicting hemodynamic states, we will also attempt to train 2 additional models on the same data—a multidimensional linear regression and a nonsequence-oriented deep neural network. For each of these models we will measure accuracy using root mean squared error (RMSE) and mean absolute error (MAE) to provide scale-dependent measurements of accuracy. RESULTS/ANTICIPATED RESULTS: Our results will be reported in 2 primary categories: numerical accuracy of the RNN model and applicability, utility, and accuracy in a live clinical setting. The use of RNNs in biomedical informatics, and in general, is a relatively new phenomenon. This means that the body of literature which could provide a basis for our expected results is limited. Because of this we have chosen staged goals in assessing our model. First, we hope to achieve a model that reliably predicts the direction of response. Being able to answer only the question of how a patient will respond—will they move toward or away from our therapeutic goal—is as good as existing prediction methods. It is well established in the literature that, by almost any metric, ~50% of hemodynamically unstable patients respond to a fluid challenge. If we are within 10% of this average (40%–60% respond), then we can be confident in the accuracy of our model in predicting direction. Upon achieving this, we will then measure accurate prediction of response magnitude. To this affect, we hope to achieve an RMSE <10% between our test data and corresponding predicted output before proceeding further. In addition to numeric accuracy, we acknowledge that a plan for practical, clinical validation is needed before utilizing this tool in a clinical environment. Such validation will require 3 separate components. First, numeric accuracy will need to be determined again as compared with prospective data on actual patients in the ICU. This step is critical to prove that no information leakage from target data back to input data occurred during training. Second, there must be a comparison to existing prediction methods, such as the passive leg raise in combination with measurement of cardiac output to predict volume responsiveness. Finally, we must measure the cost to the clinician of implementing our model in an ICU, specifically how it impacts their time to accomplish the task of selecting an intervention for the hemodynamically unstable patient. However, these tasks are beyond the scope of this project and will be left for later investigations. DISCUSSION/SIGNIFICANCE OF IMPACT: If we are successful, this study will provide the first step toward a data-driven model for predicting patient responsiveness to a given hemodynamic intervention or collection of interventions. As compared with current best practice maneuvers, this model will not require manipulation of the patient, have less rigid criteria for reliable interpretation, and not require as specific of a technical skillset to interpret. Furthermore, it will include many common categories of resuscitative therapies (eg, vasopressors, inotropes, fluids) and will allow effects of a combination of interventions to be predicted while making no assumptions of interdependence between said interventions. This study will also contribute a novel process of sequence prediction using RNNs by incorporating an element of context on top of the sequential data in every training step. An RNN learning the sequence of hemodynamic data comprising a patient’s hemodynamic state would, alone, fit into the realm of sequence prediction. Our innovation is the addition of treatment information with each temporal division of the hemodynamic data. The result is an RNN that combines the task of sequence prediction with sequence translation, the 2 major use cases for RNN learning algorithms.

OBJECTIVES/SPECIFIC AIMS: Allostatic load, the chronic stress-induced wear and tear on the body, has a cumulative deleterious effect in individuals over their lifetime. Recent studies have suggested that socio-economic status, psychological determinants, and biomedical health cumulatively contribute to allostatic load in young adults. Although these finding individually suggest that African American children may be particularly susceptible to the effects of allostatic loading due to racially-based discrimination and economic instability, few studies have shown the effect of exposure to violence on the allostatic load carried by young African Americans. METHODS/STUDY POPULATION: The Biological and Social Correlates of Drug Use in African American Emerging Adults (BADU) data set is composed of young African Americans (n=557 individuals) living in the Washington, DC area, collected from 2010 to 2012. Study participants were sought equally between males and females (n=283, n=274, respectively). This data set provides a rich source of information on the behavioral, mental, and physical health of African American young adults (18–25 year olds) living in the Washington, DC area. Analysis of 6 biomedical markers were measured in BADU study participants: C-reactive protein, cortisol, Epstein-Barr virus IgG, IgE, IgA, and IgM, known to be markers of immune stress and allostatic load. Naive Bayes was used to identify participant responses that were correlated to elevated stress biomarker levels. RESULTS/ANTICIPATED RESULTS: Violence was most closely correlated to elevated EBVVCA IgM and IgE levels. Elevated IgE levels correlated to increased experience of familial violence and sexual abuse; familial drug abuse and depression; violence and community violence. Cortisol is positively correlated to reported emotional state (R=0.072) and perceived individual discrimination (R=0.059). DISCUSSION/SIGNIFICANCE OF IMPACT: Allostatic load appears to be high in individuals who self-report exposure to violence. Both perceived mental health and violence were correlated to elevated stress biomarkers. When Epstein-Barr virus viral capsid antigen IgM was compared with violence features characterized in the data set, we found that internalization of environmental stressors were most strongly correlated to elevated allostatic load markers. This work suggests that internalization of experienced violence may be as important as the actual violence experience.

OBJECTIVES/SPECIFIC AIMS: Our objective was to develop and evaluate a machine learning pipeline that uses electronic health record (EHR) data to predict acute kidney injury (AKI) during rehospitalization for patients who did not have an AKI episode in their most recent hospitalization. METHODS/STUDY POPULATION: The protocol under which this study falls was given exempt status by our institutional review board. The fully deidentified data set, containing all adult hospital admissions during a 2-year period, is a combination of administrative, laboratory, and pharmaceutical information. The administrative data set includes International Classification of Diseases, 9th Revision (ICD-9) diagnosis and procedure codes, Current Procedural Terminology, 4th Edition (CPT-4) procedure codes, diagnosis-related grouping (DRG) codes, locations visited in the hospital, discharge disposition, insurance, marital status, gender, age, ethnicity, and total length of stay. The laboratory data set includes bicarbonate, chloride, calcium, anion gap, phosphate, glomerular filtration rate, creatinine, urea nitrogen, albumin, total protein, liver function enzymes, and hemoglobin A1c. The pharmacy data set includes, for each medication, a description, pharmacologic class and subclass, and therapeutic class. Data preprocessing was performed using Python library Pandas (McKinney, 2011). Top-level binary representation (Singh, 2015) was used for diagnosis and procedure codes. Categorical variables were transformed via 1-hot encoding. Previous admissions were collapsed using rules informed by domain expertise (eg, the most recent age or sum of assigned diagnosis codes were retained as elements in the feature vector). We excluded any patient without at least 1 rehospitalization during the time window. We excluded any admission with or without AKI where AKI was also present in the most recent hospitalization. For comparison, we do not exclude such admissions for an identical experiment in which we considered any AKI event as a positive sample (regardless of AKI presence in the most recent hospitalization). We defined an AKI event as an assignment of any of the acute kidney failure (AKF) ICD-9 codes [584.5, AKF with lesion of tubular necrosis, 584.6, AKF with lesion of renal cortical necrosis, 584.7, AKF with lesion of renal medullary (papillary) necrosis, 584.8, AKF with other specified pathological lesion in kidney, or 584.9, AKF, unspecified]. Since diagnosis codes are believed to be specific but not sensitive for AKI (Waikar, 2006), we supplemented them using creatinine for patients who had laboratory values. Diagnosis was made according to the Kidney Disease: Improving Global Outcomes (KDIGO) Practice Guidelines (AKI defined as a 1.5-fold or greater increase in serum creatinine from baseline within 7 d or 0.3 mg/dL or greater increase in serum creatinine within 48 h). We report preliminary model discrimination via area under the receiver operating characteristic curve (AUC) using k-fold cross validation grouped by patient identifier (to ensure that admissions from the same patient would not appear in the training and validation set). It was confirmed that the prevalence of positive samples in the entire data set was maintained in each fold. Python library Sci-kit Learn (Pedregosa, 2011) was used for pipeline development, which consisted of imputation, scaling, and hyper-parameter tuning for penalized (l1 and l2 norm) logistic regression, random forest, and multilayer perceptron classifiers. All experiments were stored in IPython (Pérez, 2007) notebooks for easy viewing and result reproduction. RESULTS/ANTICIPATED RESULTS: There were 107,036 adult patients that accounted for 199,545 admissions during a 2-year window. Per admission, there were at most 54 ICD-9 diagnoses, 38 ICD-9 procedures, 314 CPT-4 procedures, and 25 hospital locations visited. The admissions were 55% female, the average age was 46±standard deviation 20, and average length of stay was 2.5±8.0 days. We excluded 2360 admissions that involved an AKI event that directly followed an admission with an AKI event and 4130 admissions that did not involve an AKI event but directly followed an admission with an AKI event. In total, there were 4561 (5.3%) positive samples (AKI during rehospitalization without AKI in the previous stay) generated by 3699 unique patients and 81,458 negative samples (non-AKI during rehospitalization without AKI in the previous stay) generated by 31,831 unique patients. When using any AKI event as a positive sample (regardless of whether or not AKI was in the most recent stay), the prevalence was 7.3% (6921 positive samples generated by 4395 unique patients and 85,588 negative samples generated by 33,287 unique patients). Best results were achieved with a code precision of 3 digits for which we had a total of 4556 features per patient. Fitted hyper-parameters corresponding to each classifier were logistic regression with l1 penalty C as 2×10−3; logistic regression with l2 penalty C as 1×10−6; random forest number of estimators as 100, maximum depth as 3, minimum samples per leaf as 50, minimum samples per split as 10, and entropy as the splitting criterion; and multilayer perceptron l2 regularization parameter α as 15, architecture as 1 hidden layer with 5 units, and learning rate as 0.001. Five-fold stratified cross validation on the development set yielded AUC for logistic regression with l1 penalty average 0.830±0.006, logistic regression with l2 penalty 0.796±0.007, random forest 0.828±0.007, and multilayer perceptron 0.841±0.005. In an identical experiment for which an AKI event was considered a positive sample regardless of AKI presence in the most recent stay, we had 4592 features per sample with the same code precision. Five-fold stratified cross validation on the development set with identical settings for the hyper-parameters yielded AUC for logistic regression with l1 penalty average 0.850±0.004, logistic regression with l2 penalty 0.819±0.006, random forest 0.853±0.004, and multilayer perceptron 0.853±0.006. DISCUSSION/SIGNIFICANCE OF IMPACT: Our objective was to investigate the feasibility of using machine learning methods on EHR data to provide a personalized risk assessment for “unexpected” AKI in rehospitalized patients. Preliminary model discrimination was good, suggesting that this approach is feasible. Such a model could aid clinicians to recognize AKI risk in unsuspicious patients. The authors recognize several limitations. Since our data set corresponds to a time-window sample, patients with high frequency of hospital utilization are likely overrepresented. Similarly, our data set contains records from only 1 hospital network. Although we supplement with laboratory-based diagnosis, using diagnosis codes as labels is problematic as numerous reports suggest low sensitivity of codes for AKI. Future work includes calibration analysis, incremental updating (“online learning”), and a representation learning-based (“deep learning”) extension of the model.

OBJECTIVES/SPECIFIC AIMS: To develop an algorithm that identifies post-concussion syndrome (PCS) cases and controls from among patients with mild traumatic brain injury (mTBI) in a large academic biobank. METHODS/STUDY POPULATION: The Vanderbilt University Medical Center’s (VUMC) electronic medical record (EMR) research database includes longitudinal medical record data on 2.5 million people. DNA and genotype data were also available for >225,000 of these individuals. Our algorithm used a combination of billing codes and natural language processing to apply inclusion and exclusion criteria. We defined PCS cases as those with a PCS billing code (ICD-9 310.2 or ICD-10 F07.81) and/or symptoms of PCS within 1–6 months of a qualifying mTBI. We will compare the positive predictive value of our algorithm to that of 2 simpler case selection schemes: (1) 1 instance of the PCS billing code anywhere in the medical record; and (2) 2 or more instances of the PCS billing code anywhere in the medical record. RESULTS/ANTICIPATED RESULTS: An mTBI was diagnosed in 28,720 patients regularly attending VUMC, and 528 of these patients were classified as PCS cases by our algorithm. The characteristics of our EMR sample reflected known risk factors for PCS. Our cases were more likely than controls to be female (49.4% vs. 38.4%), to have sustained a previous TBI (31.0% vs. 12.0%) and to have comorbid mood disorders. Our PCS cases were also more likely than controls to be <18 years of age (42.4% vs. 33.6%) and to have a sports-related keyword associated with the mTBI (44.1% vs. 25.2%), emphasizing the relevance of PCS to young athletes. Nonetheless, the number of PCS cases identified by our algorithm was small, and within the VUMC EMR, there were 5039 patients with 1 PCS billing code, and 2457 patients with 2 or more PCS billing codes anywhere in their EMR. Our next step is to calculate the positive predictive values of each selection scheme by manually reviewing the EMR of a selection of cases. Ultimately, we will implement the selection scheme that maximizes both positive predictive value and sample size, and in future work, we will genotype the selected patients to better understand the genetic architecture of PCS. DISCUSSION/SIGNIFICANCE OF IMPACT: EMR and biobanks are the future of human health research, and we asked whether complex algorithms or simple billing codes were best for studying the genetics of recovery after mTBI within the VUMC EMR. Our results are relevant to other studies of brain injury phenotypes within biobanks, including recovery from moderate or severe TBI, recovery from stroke, or the occurrence of delirium after routine surgery, and will help transform biobanks into fruitful research tools.

OBJECTIVES/SPECIFIC AIMS: The promise and potential of connected personal health records (PHRs) has not come to fruition. This may be, in part, due to the lack of user-centered design and of a patient-centric approach to curating personal health data for use by patients. Co-design with end-users could help mitigate these issues by ensuring the software meets user’s needs, and also engages patients in informatics research. Our team partnered with patients with multiple chronic conditions to co-design a patient-centric PHR. This abstract will describe our experience with the co-design process, highlight functionalities desired by patients, and showcase the final prototype. METHODS/STUDY POPULATION: We conducted 3 design sessions (90 min per session) with patients as co-designers and employed an iterative process for software development. Patients were recruited from Chapel Hill and surrounding areas. The initial design session laid the foundation for future sessions, and began with brainstorming about what patients thought their ideal version of an engaging connected PHR would look like in terms of features and functionalities. After each software iteration, our entire design team, including our patient co-designers, was shown the prototype during a subsequent design session. Once the final prototype was developed, usability testing was conducted with patient participants. Our team then conducted a final design session to debrief about the final prototype. RESULTS/ANTICIPATED RESULTS: We started with an initial group of 12 patients (6 males) who all had diabetes and an additional comorbidity such as hypertension and hyperlipidemia. Age of participants ranged from 30 to 77 years with an average age of 56. The majority of participants were Caucasian with 1 Asian and 2 African Americans. Hemoglobin A1c values ranged from 6.0% to 9.2% with approximately half having A1c values less than the goal of 7.0%. Half the patients were aware of PHRs, majority had smartphones, and all participants had access to the Internet and used email. Two of the patients were retired engineers who had prior experience with software design. The other sessions had between 7 and 8 participants at each session, and 7 patients completed the 90-minute usability testing session. There was a core group of 7 patients who were engaged in the design and testing sessions throughout the entire 9-month study. Key features of the PHR that emerged from design sessions included the following: (1) allow for annotation of data by patients (particularly important for lab values like glucose or for physical activity); (2) calendars, to do list, and reminder functions should be linked so that an entry in one of these allows for auto-population of this data within the other sections; (3) notifications whenever new data from the electronic health record or other sources are pushed to the PHR account; (4) allow for drag and drop of photos of pills/medications taken via smartphone or from other sources so that medication list has photo of actual pills or pill bottle; (5) allow for patients to customize the order of sections in the PHR dashboard so that the sections most important to the individual patient can be displayed more prominently; (6) allow for notifications from pharmacies to be pushed to the PHR (eg, confirmation of receipt of prescription requests or alert that prescription is ready to pick up); and (7) graphical display of trends over time (patients would like to select the measures and time frames to plot for display). Patients cited the importance of data provenance so that patient-entered data Versus provider or electronic health record data could be easily differentiated. Patients also highlighted the importance of having this PHR be a “one-stop shop for all their health data” and to have meaningful data dashboards for the different types of information needed to comprehensively manage their health. Patients wished for a single PHR that could easily bring together data from multiple patient portal accounts to avoid having to manage multiple accounts and passwords. They felt that heat map displays such as those used on popular fitness tracking websites were not intuitive and that the color-coding made interpretation challenging. Participants noted that engagement in the design process made them feel that they contributed towards developing software that could not only positively impact them individually but others as well. Every patient indicated the desire to participate on future design projects. Of the 19 tasks evaluated during usability testing, only 5 tasks could not be completed (eg, adding exercise to the calendar, opening the heat map, etc.). Patients felt that the overall PHR design was clean and aesthetically pleasing. Most patients felt that the site was “pretty easy to use” (6 out of 7). The majority of participants would like to use this PHR in the future (5) and would recommend this PHR to their friends/family to use (6). DISCUSSION/SIGNIFICANCE OF IMPACT: Involving patients directly in the design process for creating a patient-centric connected PHR was essential to sustaining engagement throughout the software life cycle and to informing the design of features and functionalities desired by patients with chronic conditions.

OBJECTIVES/SPECIFIC AIMS: Key factors causing irreproducibility of research include those related to inappropriate study design methodologies and statistical analysis. In modern statistical practice irreproducibility could arise due to statistical (false discoveries, p-hacking, overuse/misuse of p-values, low power, poor experimental design) and computational (data, code and software management) issues. These require understanding the processes and workflows practiced by an organization, and the development and use of metrics to quantify reproducibility. METHODS/STUDY POPULATION: Within the Foundation of Discovery – Population Health Research, Center for Clinical and Translational Science, University of Utah, we are undertaking a project to streamline the study design and statistical analysis workflows and processes. As a first step we met with key stakeholders to understand the current practices by eliciting example statistical projects, and then developed process information models for different types of statistical needs using Lucidchart. We then reviewed these with the Foundation’s leadership and the Standards Committee to come up with ideal workflows and model, and defined key measurement points (such as those around study design, analysis plan, final report, requirements for quality checks, and double coding) for assessing reproducibility. As next steps we are using our finding to embed analytical and infrastructural approaches within the statisticians’ workflows. This will include data and code dissemination platforms such as Box, Bitbucket, and GitHub, documentation platforms such as Confluence, and workflow tracking platforms such as Jira. These tools will simplify and automate the capture of communications as a statistician work through a project. Data-intensive process will use process-workflow management platforms such as Activiti, Pegasus, and Taverna. RESULTS/ANTICIPATED RESULTS: These strategies for sharing and publishing study protocols, data, code, and results across the spectrum, active collaboration with the research team, automation of key steps, along with decision support. DISCUSSION/SIGNIFICANCE OF IMPACT: This analysis of statistical methods and process and computational methods to automate them ensure quality of statistical methods and reproducibility of research.

OBJECTIVES/SPECIFIC AIMS: Historically, logistic regression algorithms (LRAs) have failed to differentiate strangulated small bowel obstructions (SBOs) from nonstrangulated SBOs. Our hypothesis is that a machine learning algorithm (MLA) can differentiate strangulated from simple SBOs better than an LRA can. METHODS/STUDY POPULATION: We used records of patients presenting with acute SBO and managed with exploratory laparotomy to test and train algorithms. We compared MLA to LRA via area under the receiver operating characteristic curve (AUROC) and cut-off points maximizing sensitivity and specificity. RESULTS/ANTICIPATED RESULTS: With 192 patient records, the AUROC of the MLA was 0.85. At the sensitivity cutoff, the MLA had 100% sensitivity and 55% specificity. At the specificity cutoff, the MLA had 45% sensitivity and 100% specificity. We anticipate improvements as more records are incorporated, and that LRA will underperform MLA across all measures. DISCUSSION/SIGNIFICANCE OF IMPACT: Our MLA represents a significant improvement over past LRAs, and may provide decision assistance to surgeons managing SBO. If this MLA maintains its high sensitivity, it may be used in the future to prevent unnecessary surgeries.

OBJECTIVES/SPECIFIC AIMS: The objective of this partnership was to create a global network of clinical and public health researchers and communities conducting technology-assisted research in noncommunicable disease. METHODS/STUDY POPULATION: The University of Rochester’s Clinical and Translational Science Institute (CTSI) has successfully leveraged the informatics core’s capacity into an emerging network of organizations that focus on technology and health in settings outside of the mainland United States. The CTSI coordinated with another NIH-funded infrastructure program [the RCMI Translational Research Network (RTRN)] to identify partner institutions interested in technology and health. RTRN identified the University of Puerto Rico and the University of Hawaii, both of which serve as hubs for common research interests in technology and health throughout the Caribbean and the Pacific. This network was formalized as the CDC’s Coordinating Center for its Global and Territorial Health Research Network (the “Global Network”), with additional US partners (Yale, University of Illinois at Chicago, University of North Caroline Chapel Hill, and the University of South Florida) within a wider scope of the CDC’s Prevention Research Centers (PRC) program. RESULTS/ANTICIPATED RESULTS: Through combining 2 main NIH-funded research infrastructure networks (CTSA and RTRN), with a large CDC-funded PRC, the University of Rochester’s Informatics Core was successful in establishing a new productive global health network throughout Latin America and the Caribbean, and in the Pacific, garnering additional research support from NIH Fogarty and other programs. The resulting network not only supports locally-important research in technology and health on compelling health issues (eg, diabetes, ZIka, participation in research), but also facilitates community engagement through local partnerships and the cores of the involved networks. In addition, much of the information and communications technology (ICT)-related research and learnings from the Global Network activity is immediately applicable to populations in the United States, served by the various collaborative networks. In total, while new, the Global Network supports a wide range of projects and engagements throughout the world that expand local informatics capacity and use of technology in the research process and to address global health problems, further enhancing the CTSI’s informatics core to serve the needs of its own constituency and promote research engagement with technology within this population. Local research collaborative projects reinforce the utility of the network and its resources, evidenced by tools, publications, partnerships, and conference presentations that have arisen. Lessons to date from this Global Network collaboration include: specific global research projects provide opportunities for partnership building and meaningful collaboration, team science is of central importance in distributing the work of the network, synergy is multidirectional with expertise and need flowing in all directions, and project team members in all locales learned and contributed substantially in ways that carried into their other responsibilities. DISCUSSION/SIGNIFICANCE OF IMPACT: The overall partnership has created opportunity for South-South collaboration, for adaptation of projects among locales, and has helped boost reputational value for all partners involved. Implications for other CTSA awardees include: global collaboration can serve core research and technical needs for the CTSA itself and its local partners, CTSA status can be leveraged to access resources to support local research, and collaboration in other federally-funded research networks helps expand the insight, scope, and potential for new research.

OBJECTIVES/SPECIFIC AIMS: To learn the edit distance costs of a symbolic univariate time series representation through a stochastic finite-state transducer to predict patient outcomes in intensive care units. METHODS/STUDY POPULATION: High frequency data of patients in intensive care units were used as a data set. The nearest neighbor method with edit distance costs (learned by the FST) were used to classify the patient status within an hour after 10 hours of data. Several experiments were developed to estimate the parameters that better fit the model regarding the prediction metrics. RESULTS/ANTICIPATED RESULTS: Different metrics were obtained for the several parameters. These metrics were metrics (ie, accuracy, precision, and F-measure). DISCUSSION/SIGNIFICANCE OF IMPACT: Our best results are compared with published works, where most of the metrics (ie, accuracy, precision, and F-measure) were improved.

OBJECTIVES/SPECIFIC AIMS: The goal of this study is to assess how quality of life scores change in menopausal women before and after implementation of this aid. In addition, we are also interested in 2 process evaluation objectives: (1) determine if MyChart, the patient portal, is an effective way for this patient population to provide insight their quality of life to their providers and (2) to evaluate providers use of and reactions to the decision support tool. METHODS/STUDY POPULATION: This project is a collaboration between University of Rochester Medical Center and S.U.N.Y. Upstate Medical University. Participants were recruited through Upstate’s Family Medicine and OB/GYN practices via a MyChart invitation sent by the practices. Participating patients will be asked to complete a survey, through MyChart, every 3 months for 18 months. Participating health providers will be trained to use the decision support tool and participate in 3 interviews with the researchers to gain insight into the usefulness and effectiveness of the tool. RESULTS/ANTICIPATED RESULTS: Of the 465 eligible women, 117 women responded to our MyChart invitation to join our study. Of these, 105 agreed to participate and 98 met eligibility criteria. Only half of the women currently enrolled in our study had spoken to a provider about menopause related symptoms (56.1%) prior to study enrollment. DISCUSSION/SIGNIFICANCE OF IMPACT: The goal of this study is to improve menopause related symptoms in women, thus increasing their quality of life, but it will also provide important process evaluation for using EPIC and MyChart for future research studies.

OBJECTIVES/SPECIFIC AIMS: To share lessons learned from implementing a health survey to a global sample of mTWs. METHODS/STUDY POPULATION: mTWs were paid $0.50 for taking a 15 minute survey to ascertain attitudes and intentions toward participating in genetic research. Two phases included: pilot survey targeting 7 global regions and a large-scale implementation in English in United States, India, and other countries and in Spanish in Spanish speaking countries. Administrative and descriptive information were collected and analyzed by region/country including: completions by location, demographics, time to complete, and survey satisfaction. RESULTS/ANTICIPATED RESULTS: There are 4 key lessons: (1) MTurk is fast. The US sample (n=505) accrual took <2 days and the Indian sample (n=505) took 11 days, while the response from other countries (n=118) generally exceeded 30 days. (2) Using Amazon country specification was the best way to ensure responses from specific countries and regions. (3) Demographic differences exist in mTWs between countries. For example, US mTWs were significantly more likely female (60.1%) compared with India (30.2%) and other countries (34.2%). (4) mTWs found the survey understandable/acceptable. mTWs reported high understandability and acceptability of the survey, which did not vary significantly across countries or by language. DISCUSSION/SIGNIFICANCE OF IMPACT: mTurk provides an efficient platform for survey research from diverse US and Indian samples. In other countries and in Spanish, the mTurk mechanism yielded a smaller sample more slowly but was still effective.

OBJECTIVES/SPECIFIC AIMS: Issues with recruiting the targeted number of participants in a timely manner often results in underpowered studies, with more than 60% of clinical studies failing to complete or requiring extensions due to enrollment issues. The objective of this study is to develop and implement a scalable, organization wide platform to enhance accrual into clinical research studies. METHODS/STUDY POPULATION: We are developing and evaluating an informatics platform called Utah Utility for Research Recruitment (U2R2). U2R2 consists of 2 components: (i) Semantic Matcher: an automated trial criterion to patient matching component that also reports uncertainty associated with the match, and (ii) Match Delivery: mechanisms to deliver the list of matched patients for different research and clinical settings. As a first step, we limited the Semantic Matcher to utilize only structured data elements from the patient record and trial criteria. We are now including distributional semantic methods to match complete patient records and trial criteria as documents. We evaluated the first phase of U2R2 based on a randomized trial with a target enrollment of 220 participants that compares 2 treatment strategies for managing back pain (physical therapy and usual care) for individuals consulting a nonsurgical provider and symptomatic <90 days. RESULTS/ANTICIPATED RESULTS: U2R2 identified 9370 patients from the University of Utah Hospitals and Clinics as potential matches. Of these 9370, 1145 responded to the Back Pain study research team’s email or phone communications, and were further screened by phone. In total, 250 participants completed a screening visit, resulting in the current study enrollment of 130 participants. Forty-three of 1145 patients refused to participate, and 50 participants no-showed their screening visit. DISCUSSION/SIGNIFICANCE OF IMPACT: A recruitment platform can enhance potential participant identification, but requires attention to multiple issues involved with clinical research studies. Clinical eligibility criteria are usually unstructured and require human mediation and abstraction into discrete data elements for matching against patient records. In addition, key eligibility data are often embedded within text in the patient record. Distributional semantic approaches, by leveraging this content, can identify potential participants for screening with more specificity. The delivery of the list of matched patient results should consider characteristics of the research study, population, and targeted enrollment (eg, back pain being a common disorder and the possibility of the patient visiting different types of clinics), as well as organizational and socio-technical issues surrounding clinical practice and research. Embedding the delivery of match results into the clinical workflow by utilizing user-centered design approaches and involving the clinician, the clinic, and the patient in the recruitment process, could yield higher accrual indices.

OBJECTIVES/SPECIFIC AIMS: To create a searchable public registry of all Quality Improvement (QI) projects. To incentivize the medical professionals at UF Health to initiate quality improvement projects by reducing startup burden and providing a path to publishing results. To reduce the review effort performed by the internal review board on projects that are quality improvement Versus research. To foster publication of completed quality improvement projects. To assist the UF Health Sebastian Ferrero Office of Clinical Quality & Patient Safety in managing quality improvement across the hospital system. METHODS/STUDY POPULATION: This project used a variant of the spiral software development model and principles from the ADDIE instructional design process for the creation of a registry that is web based. To understand the current registration process and management of quality projects in the UF Health system a needs assessment was performed with the UF Health Sebastian Ferrero Office of Clinical Quality & Patient Safety to gather project requirements. Biweekly meetings were held between the Quality Improvement office and the Clinical and Translational Science – Informatics and Technology teams during the entire project. Our primary goal was to collect just enough information to answer the basic questions of who is doing which QI project, what department are they from, what are the most basic details about the type of project and who is involved. We also wanted to create incentive in the user group to try to find an existing project to join or to commit the details of their proposed new project to a data registry for others to find to reduce the amount of duplicate QI projects. We created a series of design templates for further customization and feature discovery. We then proceed with the development of the registry using a Python web development framework called Django, which is a technology that powers Pinterest and the Washington Post Web sites. The application is broken down into 2 main components (i) data input, where information is collected from clinical staff, Nurses, Pharmacists, Residents, and Doctors on what quality improvement projects they intend to complete and (ii) project registry, where completed or “registered” projects can be viewed and searched publicly. The registry consists of a quality investigator profile that lists contact information, expertise, and areas of interest. A dashboard allows for the creation and review of quality improvement projects. A search function enables certain quality project details to be publicly accessible to encourage collaboration. We developed the Registry Matching Algorithm which is based on the Jaccard similarity coefficient that uses quality project features to find similar quality projects. The algorithm allows for quality investigators to find existing or previous quality improvement projects to encourage collaboration and to reduce repeat projects. We also developed the QIPR Approver Algorithm that guides the investigator through a series of questions that allows an appropriate quality project to get approved to start without the need for human intervention. RESULTS/ANTICIPATED RESULTS: A product of this project is an open source software package that is freely available on GitHub for distribution to other health systems under the Apache 2.0 open source license. Adoption of the Quality Improvement Project Registry and promotion of it to the intended audience are important factors for the success of this registry. Thanks goes to the UW-Madison and their QI/Program Evaluation Self-Certification Tool (https://uwmadison.co1.qualtrics.com/SE/?SID=SV_3lVeNuKe8FhKc73) used as example and inspiration for this project. DISCUSSION/SIGNIFICANCE OF IMPACT: This registry was created to help understand the impact of improved management of quality projects in a hospital system. The ultimate result will be to reduce time to approve quality improvement projects, increase collaboration across the UF Health Hospital system, reduce redundancy of quality improvement projects and translate more projects into publications.

OBJECTIVES/SPECIFIC AIMS: Due to scope and breadth of research activity and infrastructure capacities at academic medical centers, the discipline of Biomedical Informatics is often deployed in a decentralized manner through geographically dispersed and unrelated organizational units. As a result, without a conscious strategy, an academic medical center risks redundant effort and gaps in resources, and perhaps poor coordination. A mechanism to bring together disparate organizational entities to identify, discuss, and negotiate Informatics-related concerns may produce a better institutional research environment. The University of Rochester (UR) has implemented such a strategy of Informatics governance, adapting tactics from team science, diplomacy, and deliberative engagement. METHODS/STUDY POPULATION: Based on current needs and institutional Informatics priorities, the UR’s Clinical and Translational Science Institute (CTSI) established 6 Informatics “clusters” in distinct but deliberately overlapping focal areas: (1) Data—capture, management, and analysis of all types of data for research. (2) Analytics—quantitative research across the spectrum of translational research. (3) Infrastructure—technical and computing infrastructure to support informatics. (4) Electronic health records (EHR)—(i) features within the EHR explicitly designed to address the needs of research; (ii) accessing and procuring EHR data for research. (5) Population health—Informatics design and systems expertise relevant to population health research (a key CTSI focus area). (6) Education—development, deployment, and assessment of Informatics learning opportunities for learners at all levels. Each cluster facilitates access to expertise and resources around the institution, promotes collaboration, identifies redundancy, and serves as a forum to strategize regarding institutional needs related to Biomedical Informatics. A CTSI faculty or staff member leads each cluster. To maximize effectiveness of the cluster, other members are decision-makers in the organizations they represent, or serve in a critical staff function. Clusters meet in person on a quarterly basis with more frequent electronic interaction. The clusters share documents via Box, a secure online file sharing app. The cluster coordinators meet as a group on a biweekly basis to monitor progress and make plans. RESULTS/ANTICIPATED RESULTS: There were 45 different people representing 46 distinct centers, departments or offices, and 2 outside agencies agreed to participate in the clusters. In total, 20 people represented a single organizational unit; 15 represented 2 units; 8 represented 3 units, and 2 represented 4 units. The richness and complexity of these organizational linkages illustrates the decentralized nature of Informatics at the institution and the promise of the cluster approach. DISCUSSION/SIGNIFICANCE OF IMPACT: Adapting to a decentralized Informatics environment, the CTSI established clusters that recognize and respect autonomy and capacity of a wide range of units throughout the university, creating a collaborative atmosphere for steering and implementing an overall Informatics vision. As Informatics capacity rapidly expands throughout growing biomedical research institutions without a centralized Informatics hub, this distributed, deliberative approach could offer an effective governance solution that promotes cooperation. In this model, the CTSI provides the leadership and staffing necessary to ensure progress at the institutional level around Informatics and creates a venue for communication and coordination on Informatics-related topics.

OBJECTIVES/SPECIFIC AIMS: (1) Obtain publically available citation data, funding data, and generate multiple networks topologies based on dynamic queries of individual faculty. (2) Determine successful pathways that lead to tenure, and career advancement, in addition to determining the effect of CTSA programs on faculty collaboration. (3) Develop publically available commercial interface for the study of faculty networks METHODS/STUDY POPULATION: For our study we included all available citation and funding data publically available on all CTSA programs (as of 2015) with historical data dating back to 2005. We then included the top 25 collegiate institutions who may not have had a CTSA program (eg, Princeton University). We then developed network topologies for each university network, and explore the evolution of individuals in these networks, and the effects of faculty development—as an example in the University of Rochester network, we singled out the directors of the CTSA program there to understand their level of centrality and overall impact on network development, with key observations being that early publications across varying domains lead to stronger network performance. Although individuals who did not benefit from such development, may have succeeded but if they did were likely to leave the institution for elsewhere. RESULTS/ANTICIPATED RESULTS: A secondary goal of this project is to evaluate the effectiveness of the Clinical & Translational Science Institute (CTSI) since its inception in 2006. The mission of CTSI is to advance the field of translational science and research, to link other departments at URMC and community stakeholders by research collaboration, publication, and goals to improve population health, and provide translational education and training to students, researchers, and physicians. To determine how the induction of CTSI affects collaboration within the URMC network, we examined the role of funding in the CTSI network. This was done around the second successful funding around 2013. In doing so we can see that not only did the funding request affect the network topology, but opened new collaborations which were not present prior to the request. DISCUSSION/SIGNIFICANCE OF IMPACT: We have developed an automated method, which is superior to manual methods necessary for citation generation and funding data analysis of faculty growth in citation networks. This technique is applicable to all institutions, not just those in a CTSA environment, but demonstrates the benefit of cross-collaborative efforts, in the case of the URMC network we can state the following. The key takeaway is for individuals to succeed in the URMC collaborative environment they should create their own network and expand it and eventually rise to prominence. There are 2 pathways to this you can take the Dewhurst approach which is to seek out collaborations among internal peers and scale up. Or you can take the Nedergaard approach which is develop the special network, and gain enough public recognition outside of the network that you are capable of leaving it (Fig. 2d). In either case, collaborations among communities and diverse out-degree networks allow faculty to succeed in their given field. Given the wealth of data which has been curated in this fashion, there are numerous explicit questions that can be asked of the data. One of the unique approaches of this data is that is highly reproducible, which allows various questions to be asked. Future work would try to determine what optimal pathways are in a given network to success, and who are ideal collaborators, and collaborations to avoid. Given this information, custom pathways to career success for individual faculty can be developed, moving beyond purely institutional level co-citation networks, which do little to advance faculty development at scale. In Figs 1c and d, the network increased by 75% in terms of graph density (0.007) and decreased by 18.8% (16) in terms of diameter. What this suggest in that the interconnectivity of the network grew dramatically, while the ability for new members to integrate into it increased. This also apparent when one examines the modularity of the network down by 3.6% (0.857), this suggest that the network has as many communities but these communities are less isolated that those in the previous funding year, meaning fields are becoming more transdisciplinary in their collaborations. This was the result of the presence of a CTSA program, thus demonstrating the effectiveness of such institutions, however, our analysis also lays the framework for applying this to other institutions which may be considering a CTSA. Or maintaining the success of a given CTSA program, and ultimately determining where faculty should place their efforts and choose which programs to pursue career advancement.

OBJECTIVES/SPECIFIC AIMS: This research project envisions the integration of Homeless Management Information System (HMIS) and UI Health Cerner electronic medical record (EMR) system with the following goals: (1) enable sharing of data about the status of the housing insecure and homeless. (2) Identify and match patient record accurately. (3) Record housing insecurity or homelessness information with structured data elements in the EMR. METHODS/STUDY POPULATION: We created a Master Person Index (MPI) of the homeless individuals from HMSI using OpenEMPI software package, which is an open source implementation of an Enterprise Master Patient Index (EMPI). An entity model was generated based on the selective data elements from HMIS database, which were relevant for the patient identity management and healthcare service management. An automated script was implemented to extract data from HMIS and load it into OpenEMPI to build the MPI. Once the MPI is setup, the Emergency Department users were able to perform patient identity matching and confirm housing insecure or homeless status of their patients by querying the index using the web-based tool. We developed structured data elements to record homelessness information, which will allow us to measure the prevalence of this risk among patients. We are also exploring the possibility to integrate the systems the using the IHE PIX/PDQ profile, which provides ways for healthcare applications to query a patient information server for a patient based on user-defined search criteria, and retrieve a patient’s information directly into the application. RESULTS/ANTICIPATED RESULTS: We implemented a MPI of homeless individuals, which would allow the emergency department users to perform patient identity matching of housing insecure or homeless patients, without undue privacy intrusions. We are confident that IHE PIX/PDQ profile is able to support the integration of healthcare and housing and homeless services systems and enable the data sharing in an efficient way. DISCUSSION/SIGNIFICANCE OF IMPACT: The project addressed the gap in the sharing of data about housing insecure or homeless persons between healthcare and housing and social services that will result in improvements in coordination of care, reduce the cycle time from recognition of risk to the referral to housing and services and improve health outcomes and residential stability. Successful completion of this integration project will give us a model that we can scale to many other communities.

OBJECTIVES/SPECIFIC AIMS: Research on cancer difference is of significant scientific and practical value. For leukemia, the survival disadvantage of the Blacks has been suggested in multiple studies. However, the existing epidemiologic analysis has multiple technical limitations. The goal of this study is to more accurately quantify so as to better understand different sources of racial differences in leukemia survival. METHODS/STUDY POPULATION: A new statistical method, which is based on robust regression and resampling, is developed. Data are obtained from the SEER (Surveillance, Epidemiology, and End Results) database. Using the “classic” epidemiologic methods as well as the new method, analysis is conducted on the prognosis of 4 leukemia subtypes (ALL, CLL, AML, and CML) for 4 major racial groups (White, non-Hispanic White, Black, and Asian and Pacific Islander). RESULTS/ANTICIPATED RESULTS: After effectively removing differences caused by the observed clinicopathological and demographic factors, the survival disadvantage of the Blacks persists for the following patient groups: ALL and age>14, CLL and age>14, and ALL and age≤14. The quantitative results are significantly different from those from classic epidemiologic analysis. Such observed racial differences are more attributable to the unobserved risk factors and cancer disparity. DISCUSSION/SIGNIFICANCE OF IMPACT: This study provides a more effective and more direct quantification of racial difference in leukemia prognosis. The survival disadvantage of the Blacks which is observed for certain subtypes/age groups deserves further attention but should not be overstated. More data collection and analysis are needed to more accurately decipher racial differences in leukemia and other cancer types.

OBJECTIVES/SPECIFIC AIMS: Clinical guidelines recommend using predicted atherosclerotic cardiovascular disease (ASCVD) risk to inform treatment decisions. The objective was to compare the contribution of changes in modifiable risk factors Versus aging to the development of high 10-year predicted ASCVD risk. METHODS/STUDY POPULATION: Prospective follow-up of the Jackson Heart Study, an exclusively African-American cohort, at visit 1 (2000–2004) and visit 3 (2009–2012). Analyses included 1115 African-American participants without a high 10-year predicted ASCVD risk (<7.5%), hypertension, diabetes, or ASCVD at visit 1. We used the Pooled Cohort equations to calculate the incidence of high (≥7.5%) 10-year predicted ASCVD risk at visit 3. We recalculated the percentage with a high 10-year predicted ASCVD risk at visit 3 assuming each risk factor [age, systolic blood pressure (SBP), antihypertensive medication use, diabetes, smoking, total and high-density lipoprotein cholesterol], one at a time, did not change from visit 1. RESULTS/ANTICIPATED RESULTS: The mean age at visit 1 was 45.2±9.5 years. Overall, 30.9% (95% CI 28.3%–33.4%) of participants developed high 10-year predicted ASCVD risk. Aging accounted for 59.7% (95% CI 54.2%–65.1%) of the development of high 10-year predicted ASCVD risk compared with 32.8% (95% CI 27.0%–38.2%) for increases in SBP or antihypertensive medication initiation and 12.8% (95% CI 9.6%–16.5%) for incident diabetes. Among participants <50 years, the contribution of increases in SBP or antihypertensive medication initiation was similar to aging. DISCUSSION/SIGNIFICANCE OF IMPACT: Increases in SBP and antihypertensive medication initiation are major contributors to the development of high 10-year predicted ASCVD risk in African Americans, particularly among younger adults.

OBJECTIVES/SPECIFIC AIMS: Traditional hospice focuses on symptoms and quality of life (QOL) at the very end of life. Clinical symptoms and QOL in the last 1–2 years of life are also important and may be affected by dementia. Our objective was to characterize how symptoms differ between people with and without dementia in the last years before death and whether symptoms impact social dimensions of QOL. METHODS/STUDY POPULATION: We studied 1270 community-dwelling participants who died between 2011 and 2015 in the National Health and Aging Trends Study, a nationally representative cohort of older adults. From the last interview before death, we examined sensory (vision; hearing), physical (pain; problems with breathing, chewing/swallowing, speaking, upper or lower extremity strength/movement, and balance/coordination), and psychiatric (depression; anxiety; insomnia) symptoms by dementia status. We examined associations between symptoms and participation restrictions (visiting family/friends, attending religious services, participating in clubs/activities, going out for enjoyment, and engaging in favorite activity). RESULTS/ANTICIPATED RESULTS: Low energy (69%), pain (59%), and lower extremity strength/movement problems (56%) were most common. People with dementia (37.3% of decedents) had higher prevalence of all symptoms (p≤0.01), except pain, breathing problems, and insomnia. Dementia and greater symptom burden were independently associated with greater odds of participation restrictions (p<0.05). Problems speaking were significantly associated with limitations in all activities except for attending religious services. Balance/coordination, energy, and strength/movement problems were associated with limitations in 3 activities. DISCUSSION/SIGNIFICANCE OF IMPACT: Sensory, physical, and psychiatric symptoms are common in the year before death, with greater symptom prevalence in people with dementia. Both dementia and symptoms are associated with restrictions in participation. Older patients may benefit not only from earlier emphasis on palliative care but also programs and assistive devices that accommodate physical impairments.

OBJECTIVES/SPECIFIC AIMS: The objective of this research is to determine under what conditions endpoints based on estimated glomerular filtration rate (eGFR) slope or on relatively small declines in eGFR provide valid and useful surrogate endpoints for pivotal clinical trials in chronic kidney disease (CKD) patients. METHODS/STUDY POPULATION: We consider 2 classes of surrogate endpoints. The first class includes endpoints defined by the average rate of change in eGFR during defined portions of the follow-up period of the trial, following initiation of the randomized treatment interventions. The second class includes composite endpoints defined by the time from randomization until the occurrence of a designated decline in eGFR or kidney failure. The true clinical endpoint is considered to be the time from randomization until kidney failure, irrespective of the trajectory in eGFR measurements prior to kidney failure. We apply statistical simulation to determine conditions under which alternative endpoints within the 2 classes are (1) valid surrogate endpoints, in the sense of preserving a low probability of rejecting the null hypothesis of no treatment effect on the surrogate endpoint when there is no treatment effect on the clinical endpoints and are also (2) useful surrogate endpoints, in the sense of providing increased statistical power that allows significant reductions in sample size and/or duration of follow-up. Input parameters for the simulations include (a) characteristics of the joint distribution of the longitudinal eGFR measurements and the time to occurrence of renal failure, (b) characteristics of the short-term and long-term effects of the treatment, and (c) design parameters, including the duration of accrual and follow-up and the spacing of eGFR measurements during the follow-up period. We use joint analyses of 19 treatment comparisons across 13 previous clinical trials of CKD patients to guide the selection of input parameters for the simulations. We apply longitudinal mixed effects models for analysis of endpoints based on eGFR slope, and Cox regression for analyses of the composite time-to-event endpoints. RESULTS/ANTICIPATED RESULTS: We have previously shown that surrogate endpoints defined by eGFR declines of 30% or 40% can provide valid and useful alternative endpoints in CKD clinical trials for interventions that do not produce short-term effects on eGFR which differ from the longer-term effects of the interventions. Other factors influencing the validity and utility of these endpoints include the average baseline eGFR, the mean rate of change in eGFR, and the extent to which the size of the treatment effect depends on the patient’s underling rate of eGFR decline. We will extend these results by presenting preliminary results describing conditions under which outcomes based on eGFR slope provide valid and useful alternatives to the clinical endpoint of time until occurrence of kidney failure. DISCUSSION/SIGNIFICANCE OF IMPACT: The statistical simulation strategy described in this research can be used during the design of clinical trials of chronic kidney disease to assist in the selection of endpoints that maximize savings in sample size and duration of follow-up while retaining a low risk of producing a false positive conclusion in the absence of a true effect of the treatment on the time until kidney failure.