Innovative App, Software Toolkit Help Harness Data

Innovative App, Software Toolkit Help Harness Data

Bioinformatics experts, programmers, and application developers in CHOP’s Department of Biomedical and Health Informatics (DBHi) build solutions for managing complex biomedical data for both research and clinical use. During fiscal 2013, DBHi’s Application Development team released two innovative applications: an iPad app for drawing family genetic histories, and an open source software toolkit that lets software developers build highly interactive data discovery applications for use by researchers and clinicians.

Similar to family trees, human pedigree diagrams are a critical tool for visualizing relationships within families and recording the occurrence of diseases through multiple generations. Genetic counselors use pedigree analysis to study the inheritance of genes and apply this knowledge to patient-care decision-making. Traditionally, counselors have hand-drawn pedigrees, but they now have a new way to create those diagrams, with just a few finger taps.

The Proband app for iPad enables genetic counselors, geneticists, and researchers to construct even the most complex family pedigrees simply and flexibly. Using simple hand gesture, users diagram the family tree with standard pedigree nomenclature and symbols. In addition, edits can be made seamlessly as they gather new information, even after the interview is completed.

“We designed this app so options appear as you need them,” said DBHi’s Jeff Miller. “Our goal was to make the features contextually relevant, and assist the user as they conduct the interview.”

Mindy Li, MD, a clinical genetics and metabolism fellow, tested the app and compared it to reviewing traditional hand-drawn pedigrees. One advantage of the app is that she did not have to decipher someone else’s handwriting or idiosyncratic abbreviations. She also enjoyed the convenience of being able to set the iPad on her lap while conducting an interview.

The app can capture any level of detail, without the space limitations of paper. And instead of opening up a file cabinet, users can store the pedigrees in a standardized format and export the diagrams to other applications such as electronic medical records. Proband is available in the iTunes App Store.

“As health technology in general is moving toward electronic data, it’s important to have pedigrees that are easy to read and easy to integrate,” Dr. Li said.

DBHi’s software toolkit, meanwhile, will help researchers confronted with large quantities of information in many forms — including vital signs, blood cell counts, lengthy DNA sequences, bar graphs, MRIs, patient demographics, and so much more. With this toolkit, DBHi specialists sought to answer the question of how researchers should assemble, access, and analyze data without having to become specialized database technicians themselves.

Their answer: an open source, highly interactive framework, known as Harvest, which is designed to allow users to navigate quickly among different types and levels of data. “We want to help researchers explore their data, not their database,” said Byron Ruth, lead developer of Harvest.

A key feature of Harvest is the ability to maneuver smoothly among various levels of data, from individual patient records to aggregated reports of all patients in a database, and to subpopulations in between. Users can construct queries to slice and dice data — grouping subjects, for instance, by age or ethnicity, calling up individual blood test results or MRIs, or including or excluding specific diagnoses.

Harvest, said DBHi’s Manager of Translational Informatics Michael J. Italia, “isn’t just shrink-wrapped, ready-to-go software.” He estimates that Harvest typically provides 80 percent of the work, leaving it to any institution’s software developer to adapt the framework to a project’s needs, in collaboration with each project’s principal investigator. Harvest is open source, so users have free access, and in fact are encouraged to customize and contribute to the toolkit.

New Software Aims to Protect Patient Privacy

New Software Aims to Protect Patient Privacy

Healthcare organizations are adept at collecting data, from admissions information to vital signs. In a multispecialty care network like The Children’s Hospital of Philadelphia where clinicians see 1 million patients annually in the ambulatory setting alone, the number of users who are accessing this data stored in patients’ electronic medical records (EHRs) on any given day is massive.

Federal privacy regulations mandate that healthcare organizations have effective privacy programs in place to monitor EHR access logs to ensure that employees are not mistreating that information. If a hospital employee leaks information about a notable patient to the media, for instance, the hospital system could be subject to fines and legal action.

“The challenge is that hospital information privacy officers around the country don’t have the auditing tools they need to look at the volume of data that they’re getting from these EHR access logs,” said CHOP’s Chief Medical Information Officer Bimal Desai, MD, MBI. “It’s like trying to drink from a firehouse.”

Dr. Desai is developing a software application based on sophisticated algorithms that will allow hospital information privacy and compliance staff to detect security breaches among internal users more intuitively. His concept was selected in the spring for CHOP’s Open Canvas project, a one-year program designed to help CHOP employees transform their ideas into accelerated business models.

Currently, hospital information privacy officers use multiple approaches to police access to patients’ medical records. On EHRs’ front end, they help to construct permissions according to users’ roles and profiles. They can build in additional authentication rules and alerts into the EHRs that require providers to attest that they are true participants in a patient’s care each time that they request a particular part of patient’s record.

On the back end, privacy officers scan the voluminous EHR access logs, but it is extremely difficult for them to derive any meaningful information. For example, a single patient who spent two weeks in CHOP’s intensive care unit generated 96,000 rows of data.

“So the task metaphorically becomes looking for the needle in the haystack,” Dr. Desai said.

He aptly named the software application that he created, “Haystack.” It uses computational techniques to filter the EHR access logs and distinguish between normal clinical behavior patterns and potential privacy breaches categorized as high, medium, or low risk. Instead of a spreadsheet with thousands of rows of data, the software generates interactive graphs that enable privacy professionals to view the network of providers involved in the patient’s care and then drill down to investigate any discrepancies.

“Ultimately, it will allow privacy professionals to focus on a very, very small subset of access events that look suspicious,” Dr. Desai said.

“Every patient deserves the highest level of rigor around privacy of their healthcare data, and organizations like CHOP and others have an obligation to help enforce that,” Dr. Desai said. “Every hospital wants to do the right thing; they just need the tools to do it.”

Novel Gene Test to Improve Transplantation, Research

Novel Gene Test to Improve Transplantation, Research

The Children’s Hospital of Philadelphia has long been a leader in the world of genomic medicine. CHOP Research investigators regularly publish groundbreaking studies that span the gamut of childhood disease, from those focused on autism to studies of the childhood cancer neuroblastoma, to investigations of common conditions like obesity and more rare diseases.

During fiscal 2013, Children’s Hospital immunogenetics experts developed a unique laboratory test to characterize the genes that encode human leukocyte antigen (HLA) molecules, which are complex proteins on cell surfaces that are essential to immune function. By using faster, more comprehensive gene sequencing technology to type HLAs, the new test may improve transplantation outcomes through a more refined assessment of donor compatibility.

“This faster, more thorough technology allows us to better account for subtle genetic differences between individuals,” said Dimitri Monos, PhD, director of the Immunogenetics Laboratory in the Division of Genomic Diagnostics. “We expect this knowledge to yield clinical benefits, by facilitating more precise matches between transplant donors and recipients, and assessing the significance of mismatches in genomic regions of the HLAs that were previously uncharacterized.”

HLA genes are the most complex gene family in the entire human genome, and current tests often provide ambiguous and limited results. In addition, preliminary HLA testing must often be followed by a second level of testing, adding expense and time to the process. However, the new CHOP test — a single, comprehensive test — will provide the highest resolution possible by covering the full HLA genomic region, Dr. Monos said. CHOP is the first hospital to offer this new comprehensive HLA-typing test, based on extensive research by Dr. Monos and colleagues.

In addition to allowing for refined assessments of donor compatibility, the test also will expedite the donor selection process from bone marrow registries. And it provides an advanced tool for research in immunological diseases, infectious diseases, and pharmacogenomics — the field that studies the influence of genetic variations on drug efficacy and toxicity.

Indeed, “by focusing on fine details of immune responses, this technology can advance our understanding of how specific individuals respond to infectious diseases, to vaccinations, and to particular drugs. This test represents a potentially powerful tool in personalized medicine,” Dr. Monos said.

“This is a new, disruptive technology, with the potential to transform research and clinical practice, in transplantation and other fields,” said Robert Doms, MD, PhD, pathologist-in-chief at CHOP.

Technology Focuses on Acne Evaluation, Management

Technology Focuses on Acne Evaluation, Management

It is challenging for clinicians who are treating a patient with acne to accurately measure disease activity from visit to visit. Often, they must rely on personal memory and the patient’s perspective to determine if the skin problem is improving, which can be an imprecise and time-intensive process.

In the U.S., it is estimated that 60 million people have acne, and one-fourth will seek out medical services; however, pediatric dermatology services are particularly difficult for patients to access with referral wait periods that can exceed more than three months.

Albert Yan, MD, Section Chief of the Division of Dermatology at The Children’s Hospital of Philadelphia, and co-developers Elena Bernardis, PhD, and Jianbo Shi, PhD, are creating a software program using sophisticated computer vision algorithms that will expedite clinical visits for acne and allow physicians to spend more valuable time on patient education.

“We are excited about this project because of the great potential to streamline the efficiency of primary care management of this disease in a way that adheres to accepted evidence-based guidelines and reduces unnecessary referrals of less severe disease to specialists,” Dr. Yan said.

First, patients will complete an electronic survey to update their clinical history and current symptoms. In the exam room, clinicians will take pictures of the patient’s skin lesions using an inexpensive smartphone or smart device, and then send those images to the software system that Dr. Yan named COMEDO, a clever acronym for “computer optimized management and evaluation of disease outcomes.” The term also carries the double meaning of blackhead or whitehead, so COMEDO represents the initiative’s target disease as well.

COMEDO will analyze the images and generate an overall assessment of acne severity. Next, it will produce a treatment plan based on a set of standardized pediatric acne guidelines that Dr. Yan helped to author for the American Acne and Rosacea Society.

“Formulating an acne treatment plan can be challenging for primary care clinicians because we all have to navigate myriad medications out there,” Dr. Yan said. “When you consider all of the topical and oral medications available, there are literally thousands of different combinations that you can come up with for any individual patient. This system will take into account patients’ skin types, medication history, etc., and then it will create treatment recommendations that are customized to that particular patient.”

At subsequent patient visits, COMEDO will update the acne assessment and treatment recommendations data. In doing so, it also will actively compile information based on the cumulative experience of all patients in the system.

“The idea is to make it a smart system that continues to get smarter as time goes on,” Dr. Yan said. “We’re going to build a big data repository of clinical response to disease, and as we accumulate that data, it will inform the system and make it smarter in terms of predicting what are going to be the best available treatments and what improvements clinicians should expect to see for their patients.”

In addition to helping pediatricians and primary care clinicians who grapple with assigning acne treatment plans on a daily basis, COMEDO’s database could spark interest from insurance company representatives who decide how to prioritize medications for their formularies or from pharmacies to help patients find appropriate over-the-counter products while patients are in the store. Eventually, patients could use COMEDO to provide symptom updates in between visits, allowing physicians to provide immediate feedback and encouragement to stay on track with the treatment plan.

Dr. Yan anticipates that a pilot prototype of COMEDO will be available in six to 12 months to be used in clinical trials within CHOP’s primary care practices that will compare the system’s accuracy with expert clinicians to establish good concordance. The development team also is working closely with the Food and Drug Administration to ensure that the device is compliant with the agency’s latest guidelines for medical apps.

The COMEDO project received start-up funds and helpful institutional resources through CHOP’s Chair’s Initiatives, which awards two-year internal grants to support the piloting of different care models to deliver accessible, high quality care at lower cost and coordinate complex, accountable care across disciplines.

Tech Transfer By the Numbers

Tech Transfer By the Numbers

List of Patents

List of Patents

U.S. Patent No. 8,497,285

Mark I. Greene, MD, PhD; Wayne W. Hancock, MBBS, PhD; Makoto Katsumata, PhD; and Yan Xiao, PhD

This invention is in the fields of immunology and autoimmunity. More particularly, it concerns methods of treating patients with compounds, which are useful agents for inhibiting the functions of TIP60 in the treatment of an individual suffering, for example, from ulcerative colitis and other irritable bowel diseases.

U.S. Patent No. 8,557,762

Rodney M. Camire, PhD, and Mettine H. A. Bos, PhD

Compositions and methods for the treatment of coagulation disorders using Factor V variants are provided. Preferred disorders include hemophilia A and B.

U.S. Patent No. 8,569,229

Michael M. Grunstein, MD, PhD

Compositions and methods for the treatment of asthma and inflammatory ocular disorders are disclosed.

U.S. Patent No. 8,580,545

Ivan S. Alferiev, PhD; Robert J. Levy, MD; and Ilia Fishbein, MD, PhD

A method and a composition for delivery of a biomaterial to an animal cell or a tissue. The composition includes (a) a biomaterial; (b) a biodegradable cross-linker portion having a hydrolyzable bond, wherein the biodegradable cross-linker portion is covalently bound to the biomaterial; and (c) a substrate, wherein the substrate is covalently bound to the biodegradable cross-linker portion, provided that the biodegradable cross-linker is adapted to hydrolyze by breaking the hydrolyzable bond and thereby release and deliver the biomaterial. A process of making the composition is also provided.

U.S. Patent No. 8,575,120

Hakon Hakonarson, MD, PhD; Struan F. A. Grant, PhD; Jonathan P. Bradfield, BS; and Constantin Polychronakos, MD

Compositions and methods for the detection and treatment of T1D are provided.

U.S. Patent No. 8,562,505

Robert J. Levy, MD; Boris Polyak, PhD; Michael Chorny, PhD; Ivan S. Alferiev, PhD; Gennady Friedman, PhD; Darryl Williams, PhD; and Ilia Fishbein, MD, PhD

Systems and methods for magnetic targeting of therapeutic particles are provided. Therapeutic particles comprise one or more magnetic or magnetizable materials and at least one therapeutic agent. Therapeutic particles are specifically targeted using uniform magnetic fields capable of magnetizing magnetizable materials, and can be targeted to particular locations in the body, or can be targeted for capture, containment, and removal. Also provided are bioresorbable nanoparticles prepared without the use of organic solvents, and methods for therapeutically using such bioresorbable nanoparticles.

U.S. Patent No. 8,618,242

Robert J. Levy, MD; Ivan S. Alferiev, PhD; and Stanley J. Stachelek, PhD

Heart valve disease affects millions, and at this time can only be treated by valve replacement or repair surgery. Cardiac valve prostheses have not significantly improved in decades, and one problem is a progressive decline in performance of heart valve replacements over 10 to 15 years after implant, necessitating additional surgery or resulting in morbidity and mortality. Additionally, valve prostheses prepared from polymeric materials, typically polyurethanes, have limited ability to bond to the living tissue to which they are attached. Therefore, methods of providing replacement valves having good longevity in vivo, including effective attachment to living tissues, would be of considerable medical value.

U.S. Patent No. 8,642,265

Carolyn A. Felix, MD; and Donald A Baldwin, PhD

Compositions, methods, and kits for detecting DNA topoisomerase II-DNA complexes are disclosed.

U.S. Patent No. 8,703,733

John H. Wolfe, VMD, PhD; and Carlos Gay-Antaki, BS

Compositions and methods for treating neurological diseases and disorders are disclosed.