Diagnostic Identification of Acute Brain Dysfunction in Pediatric Sepsis and Septic Shock in the Electronic Health Record: A Comparison of Four Definitions in a Reference Dataset.

OBJECTIVES: Acute brain dysfunction (ABD) in pediatric sepsis has a prevalence of 20%, but can be difficult to identify. Our previously validated ABD computational phenotype (CPABD) used variables obtained from the electronic health record indicative of clinician concern for acute neurologic or behavioral change. We tested whether the CPABD has better diagnostic performance to identify confirmed ABD than other definitions using the Glasgow Coma Scale or delirium scores. DESIGN: Diagnostic testing in a curated cohort of pediatric sepsis/septic shock patients. SETTING: Quaternary freestanding children's hospital. SUBJECTS: The test dataset comprised 527 children with sepsis/septic shock managed between 2011 and 2021 with a prevalence (pretest probability) of confirmed ABD of 30% (159/527). MEASUREMENTS AND MAIN RESULTS: CPABD was based on use of neuroimaging, electroencephalogram, and/or administration of new antipsychotic medication. We compared the performance of the CPABD with three GCS/delirium-based definitions of ABD-Proulx et al, International Pediatric Sepsis Consensus Conference, and Pediatric Organ Dysfunction Information Update Mandate. The posttest probability of identifying ABD was highest in CPABD (0.84) compared with other definitions. CPABD also had the highest sensitivity (83%; 95% CI, 76-89%) and specificity (93%; 95% CI, 90-96%). The false discovery rate was lowest in CPABD (1-in-6) as was the false omission rate (1-in-14). Finally, the prevalence threshold for the definitions varied, with the CPABD being the definition closest to 20%. CONCLUSIONS: In our curated dataset of pediatric sepsis/septic shock, CPABD had favorable characteristics to identify confirmed ABD compared with GCS/delirium-based definitions. The CPABD can be used to further study the impact of ABD in studies using large electronic health datasets.

Validation of a Computational Phenotype to Identify Acute Brain Dysfunction in Pediatric Sepsis.

OBJECTIVES: To validate a computational phenotype that identifies acute brain dysfunction (ABD) based on clinician concern for neurologic or behavioral changes in pediatric sepsis. DESIGN: Retrospective observational study. SETTING: Single academic children's hospital. PATIENTS: Four thousand two hundred eighty-nine index sepsis episodes. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: An existing computational phenotype of ABD was optimized to include routinely collected variables indicative of clinician concern for acute neurologic or behavioral change (completion of CT or MRI, electroencephalogram, or new antipsychotic administration). First, the computational phenotype was compared with an ABD reference standard established from chart review of 527 random sepsis episodes to determine criterion validity. Next, the computational phenotype was compared with a separate validation cohort of 3,762 index sepsis episodes to determine content and construct validity. Criterion validity for the final phenotype had sensitivity 83% (95% CI, 76-89%), specificity 93% (90-95%), positive predictive value 84% (77-89%), and negative predictive value 93% (90-96%). In the validation cohort, the computational phenotype identified ABD in 35% (95% CI 33-36%). Content validity was demonstrated as those with the ABD computational phenotype were more likely to have characteristics of neurologic dysfunction and severe illness than those without the ABD phenotype, including nonreactive pupils (15% vs 1%; p < 0.001), Glasgow Coma Scale less than 5 (44% vs 12%; p < 0.001), greater than or equal to two nonneurologic organ dysfunctions (50% vs 25%; p < 0.001), and need for intensive care (81% vs 65%; p < 0.001). Construct validity was demonstrated by higher odds for mortality (odds ratio [OR], 6.9; 95% CI, 5.3-9.1) and discharge to rehabilitation (OR, 11.4; 95% CI 7.4-17.5) in patients with, versus without, the ABD computational phenotype. CONCLUSIONS: A computational phenotype of ABD indicative of clinician concern for new neurologic or behavioral change offers a valid retrospective measure to identify episodes of sepsis that involved ABD. This computational phenotype provides a feasible and efficient way to study risk factors for and outcomes from ABD using routinely collected clinical data.

Prevalence of and Associations With Avascular Necrosis After Pediatric Sepsis: A Single-Center Retrospective Study.

OBJECTIVES: Avascular necrosis (AVN) is a rare, but serious, complication after sepsis in adults. We sought to determine if sepsis is associated with postillness diagnosis of AVN, as well as potential-associated risk factors for AVN in children with sepsis. DESIGN: Retrospective observational study. SETTING: Single academic children's hospital. PATIENTS: Patients less than 18 years treated for sepsis or suspected bacterial infection from 2011 to 2017. Patients who developed AVN within 3 years after sepsis were compared with patients who developed AVN after suspected bacterial infection and with patients with sepsis who did not develop AVN. INTERVENTION: None. MEASUREMENTS AND MAIN RESULTS: AVN was determined using International Classification of Diseases, 9th Edition/10th Edition codes and confirmed by chart review. The prevalence of AVN after sepsis was 0.73% (21/2,883) and after suspected bacterial infection was 0.43% (53/12,276; risk difference, 0.30; 95% CI, 0.0-0.63; p = 0.05). Compared with 43 sepsis controls without AVN, AVN in the 21 sepsis cases was associated with being older, having sickle cell disease and malignancy, higher body mass index, unknown source of infection, and low platelet count in the first 7 days of sepsis. Half of sepsis patients were treated with corticosteroids, and higher median cumulative dose of steroids was associated with AVN (23.2 vs 5.4 mg/kg; p < 0.01). Older age at infection (odds ratio [OR], 1.3; 95% CI, 1.1-1.4), malignancy (OR, 8.8; 95% CI, 2.6-32.9), unknown site of infection (OR, 12.7; 95% CI, 3.3-48.6), and minimal platelet count less than 100,000/µL in first 7 days of sepsis (OR, 5.0; 95% CI, 1.6-15.4) were identified as potential risk factors for AVN after sepsis following adjustment for multiple comparisons. CONCLUSIONS: Although rare, sepsis was associated with a higher risk of subsequent AVN than suspected bacterial infection in children. Older age, malignancy, unknown site of infection, and minimum platelet count were potential risk factors for AVN after sepsis.

Evaluating commercially available wireless cardiovascular monitors for measuring and transmitting real-time physiological responses in children with autism.

Commercially available wearable biosensors have the potential to enhance psychophysiology research and digital health technologies for autism by enabling stress or arousal monitoring in naturalistic settings. However, such monitors may not be comfortable for children with autism due to sensory sensitivities. To determine the feasibility of wearable technology in children with autism age 8-12 years, we first selected six consumer-grade wireless cardiovascular monitors and tested them during rest and movement conditions in 23 typically developing adults. Subsequently, the best performing monitors (based on data quality robustness statistics), Polar and Mio Fuse, were evaluated in 32 children with autism and 23 typically developing children during a 2-h session, including rest and mild stress-inducing tasks. Cardiovascular data were recorded simultaneously across monitors using custom software. We administered the Comfort Rating Scales to children. Although the Polar monitor was less comfortable for children with autism than typically developing children, absolute scores demonstrated that, on average, all children found each monitor comfortable. For most children, data from the Mio Fuse (96%-100%) and Polar (83%-96%) passed quality thresholds of data robustness. Moreover, in the stress relative to rest condition, heart rate increased for the Polar, F(1,53) = 135.70, p < 0.001, ηp2  = 0.78, and Mio Fuse, F(1,53) = 71.98, p < 0.001, ηp2  = 0.61, respectively, and heart rate variability decreased for the Polar, F(1,53) = 13.41, p = 0.001, ηp2  = 0.26, and Mio Fuse, F(1,53) = 8.89, p = 0.005, ηp2  = 0.16, respectively. This feasibility study suggests that select consumer-grade wearable cardiovascular monitors can be used with children with autism and may be a promising means for tracking physiological stress or arousal responses in community settings. LAY SUMMARY: Commercially available heart rate trackers have the potential to advance stress research with individuals with autism. Due to sensory sensitivities common in autism, their comfort wearing such trackers is vital to gathering robust and valid data. After assessing six trackers with typically developing adults, we tested the best trackers (based on data quality) in typically developing children and children with autism and found that two of them met criteria for comfort, robustness, and validity.

SARS-CoV-2 Infection in Public School District Employees Following a District-Wide Vaccination Program - Philadelphia County, Pennsylvania, March 21-April 23, 2021.

The School District of Philadelphia reopened for in-school instruction the week of March 21, 2021, and required weekly testing for SARS-CoV-2, the virus that causes COVID-19, for all employees returning to in-school responsibilities. The resumption of in-school instruction followed a mass vaccination program using the Pfizer-BioNTech 2-dose vaccine offered under a partnership between the Philadelphia Department of Public Health and Children's Hospital of Philadelphia to all 22,808 School District of Philadelphia employees during February 23-April 3, 2021.* The subsequent mandatory testing program provided an opportunity to assess the percentage of positive BinaxNow point-of-care antigen tests (Abbott Laboratories) identified among school staff members based on their self-reported vaccination status (i.e., received zero, 1, or 2 vaccine doses) at the time of testing. During the initial 5 weeks after schools reopened, 34,048 screening tests were performed. Overall, 0.70% of tests returned a positive result. The percentage of positive test results was lower among persons who reported receipt of 2 vaccine doses (0.09%) compared with those who reported receipt of 1 dose (1.21%) or zero doses (1.76%) (p<0.001) representing a 95% reduction in percentage of positive SARS-CoV-2 test results among persons reporting receipt of 2 compared with zero doses of Pfizer-BioNTech vaccine. Vaccination of school staff members has been highlighted as an important strategy to maximize the safety of in-person education of K-12 students this fall (1). These findings reinforce the importance of promoting COVID-19 vaccination among school staff members before commencement of the 2021-22 school year.

Identification of Pediatric Sepsis for Epidemiologic Surveillance Using Electronic Clinical Data.

OBJECTIVES: A method to identify pediatric sepsis episodes that is not affected by changing diagnosis and claims-based coding practices does not exist. We derived and validated a surveillance algorithm to identify pediatric sepsis using routine clinical data and applied the algorithm to study longitudinal trends in sepsis epidemiology. DESIGN: Retrospective observational study. SETTING: Single academic children's hospital. PATIENTS: All emergency and hospital encounters from January 2011 to January 2019, excluding neonatal ICU and cardiac center. EXPOSURE: Sepsis episodes identified by a surveillance algorithm using clinical data to identify infection and concurrent organ dysfunction. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: A surveillance algorithm was derived and validated in separate cohorts with suspected sepsis after clinician-adjudication of final sepsis diagnosis. We then applied the surveillance algorithm to determine longitudinal trends in incidence and mortality of pediatric sepsis over 8 years. Among 93,987 hospital encounters and 1,065 episodes of suspected sepsis in the derivation period, the surveillance algorithm yielded sensitivity 78% (95% CI, 72-84%), specificity 76% (95% CI, 74-79%), positive predictive value 41% (95% CI, 36-46%), and negative predictive value 94% (95% CI, 92-96%). In the validation period, the surveillance algorithm yielded sensitivity 84% (95% CI, 77-92%), specificity of 65% (95% CI, 59-70%), positive predictive value 43% (95% CI, 35-50%), and negative predictive value 93% (95% CI, 90-97%). Notably, most "false-positives" were deemed clinically relevant sepsis cases after manual review. The hospital-wide incidence of sepsis was 0.69% (95% CI, 0.67-0.71%), and the inpatient incidence was 2.8% (95% CI, 2.7-2.9%). Risk-adjusted sepsis incidence, without bias from changing diagnosis or coding practices, increased over time (adjusted incidence rate ratio per year 1.07; 95% CI, 1.06-1.08; p < 0.001). Mortality was 6.7% and did not change over time (adjusted odds ratio per year 0.98; 95% CI, 0.93-1.03; p = 0.38). CONCLUSIONS: An algorithm using routine clinical data provided an objective, efficient, and reliable method for pediatric sepsis surveillance. An increased sepsis incidence and stable mortality, free from influence of changes in diagnosis or billing practices, were evident.

Perspective on the Development of a Large-Scale Clinical Data Repository for Pediatric Hearing Research.

The use of "big data" for pediatric hearing research requires new approaches to both data collection and research methods. The widespread deployment of electronic health record systems creates new opportunities and corresponding challenges in the secondary use of large volumes of audiological and medical data. Opportunities include cost-effective hypothesis generation, rapid cohort expansion for rare conditions, and observational studies based on sample sizes in the thousands to tens of thousands. Challenges include finding and forming appropriately skilled teams, access to data, data quality assessment, and engagement with a research community new to big data. The authors share their experience and perspective on the work required to build and validate a pediatric hearing research database that integrates clinical data for over 185,000 patients from the electronic health record systems of three major academic medical centers.

Utility and limitations of exome sequencing as a genetic diagnostic tool for children with hearing loss.

PURPOSE: Hearing loss (HL) is the most common sensory disorder in children. Prompt molecular diagnosis may guide screening and management, especially in syndromic cases when HL is the single presenting feature. Exome sequencing (ES) is an appealing diagnostic tool for HL as the genetic causes are highly heterogeneous. METHODS: ES was performed on a prospective cohort of 43 probands with HL. Sequence data were analyzed for primary and secondary findings. Capture and coverage analysis was performed for genes and variants associated with HL. RESULTS: The diagnostic rate using ES was 37.2%, compared with 15.8% for the clinical HL panel. Secondary findings were discovered in three patients. For 247 genes associated with HL, 94.7% of the exons were targeted for capture and 81.7% of these exons were covered at 20× or greater. Further analysis of 454 randomly selected HL-associated variants showed that 89% were targeted for capture and 75% were covered at a read depth of at least 20×. CONCLUSION: ES has an improved yield compared with clinical testing and may capture diagnoses not initially considered due to subtle clinical phenotypes. Technical challenges were identified, including inadequate capture and coverage of HL genes. Additional considerations of ES include secondary findings, cost, and turnaround time.

Anticipated responses of early adopter genetic specialists and nongenetic specialists to unsolicited genomic secondary findings.

PURPOSE: Secondary findings from genomic sequencing are becoming more common. We compared how health-care providers with and without specialized genetics training anticipated responding to different types of secondary findings. METHODS: Providers with genomic sequencing experience reviewed five secondary-findings reports and reported attitudes and potential clinical follow-up. Analyses compared genetic specialists and physicians without specialized genetics training, and examined how responses varied by secondary finding. RESULTS: Genetic specialists scored higher than other providers on four-point scales assessing understandings of reports (3.89 vs. 3.42, p = 0.0002), and lower on scales assessing reporting obligations (2.60 vs. 3.51, p < 0.0001) and burdens of responding (1.73 vs. 2.70, p < 0.0001). Nearly all attitudes differed between findings, although genetic specialists were more likely to assert that laboratories had no obligations when findings had less-established actionability (p < 0.0001 in interaction tests). The importance of reviewing personal and family histories, documenting findings, learning more about the variant, and recommending familial discussions also varied according to finding (all p < 0.0001). CONCLUSION: Genetic specialists felt better prepared to respond to secondary findings than providers without specialized genetics training, but perceived fewer obligations for laboratories to report them, and the two groups anticipated similar clinical responses. Findings may inform development of targeted education and support.

Utilization of the Tablet Application Proband in Pedigree Construction and Assessment.

Many medical institutions have converted to a digital model for record keeping due to the Health Information Technology for Economic and Clinical Health Act. This Act provides incentives to health care systems to accelerate and encourage the adoption of electronic health record (EHR) systems. The pedigree as a tool in medicine provides an efficient method to assess and represent an individual's health and family health risks that may otherwise not be apparent in the medical record in a clearly identifiable way (Schuette, J. L., & Bennett 2009). Many clinicians continue to construct pedigrees using pen and paper method despite findings of improved identification of at risk patients with similar electronic intake tools (Arar et al. in Personalized Medicine 2011 8:523-32). The goal of this study was to explore the patient and practitioner experience with electronic pedigree programs using Proband, an application developed at The Children's Hospital of Philadelphia for genetic counselors to construct pedigrees during genetic counseling sessions directly on iPads. The first part of this study looked at the patient experience and assessed time to take the pedigree and the impact of using an electronic pedigree tool on the relationship between participant and genetic counselor. This involved 50 participants and was compared with the traditional paper method of taking a pedigree. There was no statistical significance found between the two different mediums in accuracy, speed, and rapport with provider. The second part of the study assessed the usability of Proband by ten genetic counselors. Overall, the application received a system usability score of 90/100 with a majority (7/10) of counselors agreeing that they would use this application in their clinic. The positive outcome of this study encourages future work to assess the impact and usability of programs on a larger scale as they continue to integrate into current electronic health records.

Genomic decision support needs in pediatric primary care.

Clinical genome and exome sequencing can diagnose pediatric patients with complex conditions that often require follow-up care with multiple specialties. The American Academy of Pediatrics emphasizes the role of the medical home and the primary care pediatrician in coordinating care for patients who need multidisciplinary support. In addition, the electronic health record (EHR) with embedded clinical decision support is recognized as an important component in providing care in this setting. We interviewed 6 clinicians to assess their experience caring for patients with complex and rare genetic findings and hear their opinions about how the EHR currently supports this role. Using these results, we designed a candidate EHR clinical decision support application mock-up and conducted formative exploratory user testing with 26 pediatric primary care providers to capture opinions on its utility in practice with respect to a specific clinical scenario. Our results indicate agreement that the functionality represented by the mock-up would effectively assist with care and warrants further development.

The biorepository portal toolkit: an honest brokered, modular service oriented software tool set for biospecimen-driven translational research.

BACKGROUND: High throughput molecular sequencing and increased biospecimen variety have introduced significant informatics challenges for research biorepository infrastructures. We applied a modular system integration approach to develop an operational biorepository management system. This method enables aggregation of the clinical, specimen and genomic data collected for biorepository resources. METHODS: We introduce an electronic Honest Broker (eHB) and Biorepository Portal (BRP) open source project that, in tandem, allow for data integration while protecting patient privacy. This modular approach allows data and specimens to be associated with a biorepository subject at any time point asynchronously. This lowers the bar to develop new research projects based on scientific merit without institutional review for a proposal. RESULTS: By facilitating the automated de-identification of specimen and associated clinical and genomic data we create a future proofed specimen set that can withstand new workflows and be connected to new associated information over time. Thus facilitating collaborative advanced genomic and tissue research. CONCLUSIONS: As of Janurary of 2016 there are 23 unique protocols/patient cohorts being managed in the Biorepository Portal (BRP). There are over 4000 unique subject records in the electronic honest broker (eHB), over 30,000 specimens accessioned and 8 institutions participating in various biobanking activities using this tool kit. We specifically set out to build rich annotation of biospecimens with longitudinal clinical data; BRP/REDCap integration for multi-institutional repositories; EMR integration; further annotated specimens with genomic data specific to a domain; build application hooks for experiments at the specimen level integrated with analytic software; while protecting privacy per the Office of Civil Rights (OCR) and HIPAA.

Temporal bone radiology report classification using open source machine learning and natural langue processing libraries.

BACKGROUND: Radiology reports are a rich resource for biomedical research. Prior to utilization, trained experts must manually review reports to identify discrete outcomes. The Audiological and Genetic Database (AudGenDB) is a public, de-identified research database that contains over 16,000 radiology reports. Because the reports are unlabeled, it is difficult to select those with specific abnormalities. We implemented a classification pipeline using a human-in-the-loop machine learning approach and open source libraries to label the reports with one or more of four abnormality region labels: inner, middle, outer, and mastoid, indicating the presence of an abnormality in the specified ear region. METHODS: Trained abstractors labeled radiology reports taken from AudGenDB to form a gold standard. These were split into training (80 %) and test (20 %) sets. We applied open source libraries to normalize and convert every report to an n-gram feature vector. We trained logistic regression, support vector machine (linear and Gaussian), decision tree, random forest, and naïve Bayes models for each ear region. The models were evaluated on the hold-out test set. RESULTS: Our gold-standard data set contained 726 reports. The best classifiers were linear support vector machine for inner and outer ear, logistic regression for middle ear, and decision tree for mastoid. Classifier test set accuracy was 90 %, 90 %, 93 %, and 82 % for the inner, middle, outer and mastoid regions, respectively. The logistic regression method was very consistent, achieving accuracy scores within 2.75 % of the best classifier across regions and a receiver operator characteristic area under the curve of 0.92 or greater across all regions. CONCLUSIONS: Our results indicate that the applied methods achieve accuracy scores sufficient to support our objective of extracting discrete features from radiology reports to enhance cohort identification in AudGenDB. The models described here are available in several free, open source libraries that make them more accessible and simplify their utilization as demonstrated in this work. We additionally implemented the models as a web service that accepts radiology report text in an HTTP request and provides the predicted region labels. This service has been used to label the reports in AudGenDB and is freely available.

Increasing cumulative exposure to volatile anesthetic agents is associated with poorer neurodevelopmental outcomes in children with hypoplastic left heart syndrome.

OBJECTIVES: Despite improved survival in children with hypoplastic left heart syndrome (HLHS), significant concern persists regarding their neurodevelopmental (ND) outcomes. Previous studies have identified patient factors, such as prematurity and genetic syndromes, to be associated with worse ND outcomes. However, no consistent relationships have been identified among modifiable management factors, including cardiopulmonary bypass strategies, and ND outcomes after cardiac surgery in infancy. Studies in immature animals, including primates, have demonstrated neurodegeneration and apoptosis in the brain after certain levels and extended durations of anesthetic exposure. Retrospective human studies have also suggested relationships between adverse ND effects and anesthetic exposure. METHODS: Cumulative minimum alveolar concentration hours (MAC-hrs) of exposure to volatile anesthetic agents (VAA) (desflurane, halothane, isoflurane, and sevoflurane) were collected from an anesthetic database and medical record review for 96 patients with HLHS or variants. ND testing was performed between ages 4 and 5 years, including full-scale IQ, verbal IQ, performance IQ, and processing speed. Four generalized linear modes were hypothesized a priori and tested using a Gaussian (normal) distribution with an identity link. RESULTS: Cumulative VAA exposure ranged from 0 to 35.3 MAC-hrs (median 7.5 hours). Using specified covariates identified previously as significant predictors of ND outcomes, statistically significant relationships were identified between total MAC-hrs exposure and worse full-scale IQ and verbal IQ scores (P's < .05) alone and after adjusting for relevant covariates. CONCLUSIONS: Increased cumulative MAC-hrs exposure to VAA is associated with worse ND outcomes in certain domains in children with HLHS and variants.

Clinical phenotype-based gene prioritization: an initial study using semantic similarity and the human phenotype ontology.

BACKGROUND: Exome sequencing is a promising method for diagnosing patients with a complex phenotype. However, variant interpretation relative to patient phenotype can be challenging in some scenarios, particularly clinical assessment of rare complex phenotypes. Each patient's sequence reveals many possibly damaging variants that must be individually assessed to establish clear association with patient phenotype. To assist interpretation, we implemented an algorithm that ranks a given set of genes relative to patient phenotype. The algorithm orders genes by the semantic similarity computed between phenotypic descriptors associated with each gene and those describing the patient. Phenotypic descriptor terms are taken from the Human Phenotype Ontology (HPO) and semantic similarity is derived from each term's information content. RESULTS: Model validation was performed via simulation and with clinical data. We simulated 33 Mendelian diseases with 100 patients per disease. We modeled clinical conditions by adding noise and imprecision, i.e. phenotypic terms unrelated to the disease and terms less specific than the actual disease terms. We ranked the causative gene against all 2488 HPO annotated genes. The median causative gene rank was 1 for the optimal and noise cases, 12 for the imprecision case, and 60 for the imprecision with noise case. Additionally, we examined a clinical cohort of subjects with hearing impairment. The disease gene median rank was 22. However, when also considering the patient's exome data and filtering non-exomic and common variants, the median rank improved to 3. CONCLUSIONS: Semantic similarity can rank a causative gene highly within a gene list relative to patient phenotype characteristics, provided that imprecision is mitigated. The clinical case results suggest that phenotype rank combined with variant analysis provides significant improvement over the individual approaches. We expect that this combined prioritization approach may increase accuracy and decrease effort for clinical genetic diagnosis.

Harvest: an open platform for developing web-based biomedical data discovery and reporting applications.

Biomedical researchers share a common challenge of making complex data understandable and accessible as they seek inherent relationships between attributes in disparate data types. Data discovery in this context is limited by a lack of query systems that efficiently show relationships between individual variables, but without the need to navigate underlying data models. We have addressed this need by developing Harvest, an open-source framework of modular components, and using it for the rapid development and deployment of custom data discovery software applications. Harvest incorporates visualizations of highly dimensional data in a web-based interface that promotes rapid exploration and export of any type of biomedical information, without exposing researchers to underlying data models. We evaluated Harvest with two cases: clinical data from pediatric cardiology and demonstration data from the OpenMRS project. Harvest's architecture and public open-source code offer a set of rapid application development tools to build data discovery applications for domain-specific biomedical data repositories. All resources, including the OpenMRS demonstration, can be found at http://harvest.research.chop.edu.

Harvest: a web-based biomedical data discovery and reporting application development platform.

Biomedical researchers share a common challenge of making complex data understandable and accessible. This need is increasingly acute as investigators seek opportunities for discovery amidst an exponential growth in the volume and complexity of laboratory and clinical data. To address this need, we developed Harvest, an open source framework that provides a set of modular components to aid the rapid development and deployment of custom data discovery software applications. Harvest incorporates visual representations of multidimensional data types in an intuitive, web-based interface that promotes a real-time, iterative approach to exploring complex clinical and experimental data. The Harvest architecture capitalizes on standards-based, open source technologies to address multiple functional needs critical to a research and development environment, including domain-specific data modeling, abstraction of complex data models, and a customizable web client.