NIH Establishes Early Version of Personalized Medicine Platform

P&T Journal - for the original article go HERE.  For a PDF version go HERE.

Congress has been pumping money into the precision-medicine initiative at the National Institutes of Health (NIH). The first, early results of that spending were announced on May 7, when the All of Us research project went live with three types of aggregate health data on 142,000 individuals. It was the first wave of what is expected to amount to, or hoped to amount to, one million participants. All of Us grew out of the 21^st Century Cures Act and is a favored target for federal funding. President Trump has proposed a 12% cut in NIH funding for fiscal 2020 (starting on October 1, 2019), which Congress is likely to reverse, as it has done in the last few years when similar cuts were asked for by the White House. However, the President’s 2020 budget does contain $313 million for All of Us.

The program was kicked off one year ago in seven cities around the country, with a particular emphasis on convincing members of underrepresented and minority communities to come forward with their health data. The idea is eventually to be able to check genomic data against environmental experience with a view to finding clues on the origin of various health conditions, so as to better target prevention and treatment measures. So far, more than 206,000 people have begun the enrollment process, and more than 142,000 have completed all the steps in the protocol. Most participants (> 75%) are from communities that have been underrepresented in biomedical research. “Those people are often left behind,” said Francis Collins, Director of NIH. “We’re hoping to chip away at vexing health disparities.”

A number of universities around the country have received grants to pay for participant enlistment efforts, including the University of Alabama at Birmingham (UAB). Bruce Korf, MD, Chief Genomics Officer, UAB Medicine, says, “The All of Us research program is creating an unprecedented rich set of medical and biological data on one million participants who reflect the diversity of the United States. It will provide the foundation for the broad application of precision medicine for years to come, including the development of approaches to predict individuals at risk of disease and more precise and effective treatments.”

The NIH program is headed by Eric Dishman, former vice president of the Health and Life Sciences Group at Intel Corporation, where he was responsible for driving global strategy, research and development, product and platform development, and policy initiatives for health and life science solutions. Dishman––and Collins, among others––spoke at the one-year anniversary conference held at NIH on May 6. He explained that All of Us works with a consortium of 2,000 organizations and people from local communities to convince individuals to provide three types of data: electronic health records, responses to survey questions, and physical measurements. The data are available on the All of Us website (https://databrowser.researchallofus.org/.), where one can view, for example, the top 10 health conditions and then drill down in each category to get breakdowns, such as the age of those people affected and other measures.

Dishman was honest in his appraisal of the initial product. “Our tools are pretty crude,” he said. “Over time, the viewer experience will get better.” He noted that the early data can’t be sorted by ethnicity or race, which would seem like a pretty important goal considering that the project’s overarching objective is mining health statistics from the African American and Hispanic populations.

Gary Gibbons, MD, Director of the Heart, Lung and Blood Institute at NIH, posited the potential by explaining that further study of genomic variations among African Americans could help explain the root causes of sickle cell disease. He noted that individuals’ African roots may have caused variations in their hemoglobin genes via interaction with malaria vectors, and ultimately resulted in the sickle cell trait. An individual could have one genetic copy of that trait, or two, and that difference would have health implications. On one hand, the variations could have been protective, against malaria. On the other hand, that same sickle cell trait may predispose African Americans to chronic kidney disease, for which they are at higher risk.

Beyond sickle cell disease, Gibbons noted that genomic/environmental interaction could identify people who are at increased risk for heart disease, hypertension, coronary disease, and asthma. “We’re working toward a day when we can predict that you’ll have atrial fibrillation, which today may only become manifest when you come to an emergency room feeling dizzy and short of breath with a rapid heartbeat. But if we already know from your polygenic risk score that you are [susceptible to] atrial fibrillation, you may be wearing a watch that detects your heart rhythm, allowing us to detect the arrhythmia and provide an electric jolt to short-circuit it.”

Much of the promise of the All of Us initiative appears to rest on genomic exploration, and there are no genomic data available at present. Collins said that the data would be included in 2020 and that the program has a “bold timetable” for enrolling the 800,000 individuals who would complete the cohort of one million.

Author bio: 

Mr. Barlas is a freelance writer in Washington, D.C. who covers issues inside the Beltway.

Possibility of FDA Regulation of Health Information Technology Looms Large

P&T Journal
October 2013 - for a PDF copy of the published version go HERE.

Pharmacies and Their Vendors Worry About Quality

The hospital drive to implement health information technology (HIT) systems, driven in good part by Medicare and Medicaid incentives (and penalties), has certainly affected pharmacy systems, and will continue to do so as the definition of "meaningful use" is expanded to incorporate more required medication tasks. So anyone who works in a hospital pharmacy ought to be interested in how the federal government regulates these new technologies, be they hardware or software. Retail pharmacies, too, even though they are not being pressured by incentives, are concerned about how any regulation would affect mobile medical devices such as glucose monitoring/diabetes management, at home hypertension monitoring/management, medication management, and medication reconciliation at transitions of care. 

Pharmacy organizations such as the Pharmacy e-HIT Collaborative are communicating with a work group within the Department of Health and Human Services (HHS) Office of the National Coordinator for Health Information Technology (ONC) which will be making HIT recommendations to Congress imminently. That report will provide directions to the Food and Drug Administration (FDA).

Surescripts, too, is concerned about the failure of some pharmacy systems vendors to police the quality of their products. Surescripts provides the e-prescribing electronic backbone which software vendors plug into. Those vendors sell plugged-in e-prescribing systems directly to end-user pharmacies. David Yakimischak, General Manager, E-Prescribing, says the company has worked hard to encourage its vendors to adopt quality programs leading to fewer pharmacy errors. "Vendor responsiveness has been relatively random," he explains. "Quality/patient safety performance does not appear to play heavily in end-users’ purchasing decisions. Our conclusion is that in the absence of new incentives and penalties, there is currently little or no business case for vendors to make significant investments in high quality patient safety programs."

The FDA is under a congressionally-mandated deadline of January 2014 to produce a regulatory framework for HIT. The key issue is whether the FDA should be allowed to regulate all HIT as it does medical devices. Or should some HIT, especially software where pharmacists and physicians interact with a system to make clinical decisions or observations, even be regulated at all. There are those who think systems requiring clinical intervention should be certified, not regulated, by a non-federal body such as The Joint Commission, with that certification process overseen by the ONC. The ONC already has a few years of experience certifying HIT eligible for incentives (which go to the purchasers, such as hospitals) decreed by the 2009 stimulus bill. "The ONC is well positioned to coordinate oversight among these various organizations to ensure patient safety, prevent overlap and continue to foster innovation and adoption of health IT by providers," states Ann Richardson Berkey, Senior Vice President, Public Affairs, McKesson Corp. She says organizations such as the National Committee for Quality Assurance (NCQA), The Joint Commission or URAC have track records in the rigorous accreditation of healthcare software.

The American Hospital Association (AHA) thinks products should be regulated based on the risk they pose to patients. Key factors to be considered include the potential for harm, the extent of harm, and the extent to which software is automating and or guiding clinical decision-making. "For example, when drug dosage data are sent from an order entry system to a pharmacy information system, it is crucial for safety that both the data points and their units of measure are accurate within each system and across systems," says Linda E. Fishman, Senior Vice President, Public Policy Analysis & Development, AHA.

Last February, the Bipartisan Policy Center, a think tank of sorts meant to meld Democratic and Republican views, produced a report called An Oversight Framework for Assuring Patient Safety in Health Information Technology. It stated: "The FDA’s current regulatory approach for medical devices is generally not well-suited for health IT." The report recommends that HIT products be divided into three categories according to the relative risk to patients and the opportunity for clinical intervention. Those where there is no or little opportunity for clinical intervention represent a higher potential risk of patient harm. Such devices are currently regulated by the FDA as Class I, Class II, or Class III medical devices. The FDA would continue to regulate these devices.

The lowest risk category would be administrative software which supports the administrative and operational aspects of healthcare but is not used in direct delivery of clinical care. Population analytics, back office billing systems, claims payment systems, and prescription drug refill reminders are all examples of software that are not used for patient specific treatment or diagnosis. The BPC recommends no additional oversight for this category.

The middle category would include products which can be used to recommend a course of care; call this category clinical software. Very few participants in the debate, at least on the industry side, think this category should be regulated by a federal agency.

    

Meanwhile, the ONC's ability to translate its considerable HIT experience into political weight in any upcoming battle with the FDA over new regulations is compromised by the absence of a National Coordinator, given the departure of Farzad Mostashari, M.D. The position needs to be filled quickly, and for a lot of reasons.