The Future of HIT and Medical Informatics

Here’s a speech I recently delivered to a graduate Medical Informatics class at Temple University.

Hello, my name is Dan Foltz. I am the owner of Parnassus Consulting, a local healthcare and life sciences consulting firm. I also serve as Chair of the organizing committee for the Mid-Atlantic Chapter of the American Medical Informatics Association. We just launched the chapter in April. I am pleased to say that this chapter is the first local Chapter created in AMIA’s history, and now several other regions across the nation are following suite. You can learn more about the chapter at www.AMIA.org/chapters

AMIA’s slogan is “informatics professionals leading the way.” It’s going to take a special breed of professionals who understand healthcare, data and computer science to transform healthcare with health information technology. Increasingly they also need to understand healthcare as a business that faces harsh economic realities.

I recently had the opportunity to develop a commercialization strategy for an academic medical center that has built a new health information technology platform. Now listen to this description, this platform manages rules that are executed against molecular and clinical data accessed in real time from a variety of data sources and presented as decision support within existing EHR systems and other applications. I like to highlight this example because it’s loaded with implications for the future of health information technology. In one sentence you heard reference to personalized medicine, multiple data source access, rules based clinical decision support and real-time data use. There was also a plural reference to applications implying an application ecosystem that works through a standard approach to liberate data that heretofore has been locked into silos. You see in the rapidly coming future, it will not be all about any single EHR vendor.

You also heard reference to rules based clinical decision support algorithms that exist outside of the EMR, implying a web-service that is managed outside of the traditional EHR but interoperates with it. Certain functions simply need to be optimized in certain places by certain groups of experts. And clinical decision support is an example of one such function. The market is pretty well aligned around this point.

Times they are a changing. But it is complex. The issues around interoperability and usability remain the bain of health information technology for clinicians that want to provide the best care possible, but still find that HIT sometimes gets in the way. It truly will take informatics professionals to lead the way in shaping health information technologies.

I want to leave you with three thoughts about changes coming in healthcare technology that present opportunities for you to embrace and pursue in your careers. 1. Consumerism will fundamentally change healthcare as we know it. 2. Electronic health records will evolve into communication hubs within an application ecosystem 3. All healthcare stakeholders will have appropriate access to healthcare data to create value and manage risk

So first, let me give you some examples about how consumerism will fundamentally change healthcare as we know it. Have you heard of Scanadu? This new company has transferred NASA sensor technology from the MARS space program to create multi-functional consumer health sensors. They are currently bringing two consumer devices to market for testing next year. One device called Scanadu Scout is being called the first medical Tricorder. You hold it to your forehead and in 20 seconds it measures your heart rate, blood pressure, core body temperature, ECG, oxymetry, respiratory rate and emotional stress. It sends this data to a smart phone app that can track normal fluctuations and spot abnormal changes. A second device, Scanaflo, will test urine for levels of glucose, protein, leukocytes, nitrates, blood, bilirubin, urobilirubin, specific gravity, pH and pregnancy. It is designed to give people early warning if something is wrong with their liver, kidneys, urinary tract or metabolism; and do so at a fraction of the time and cost of conventional testing.

Scanadu’s goal is to obtain FDA clearance as a medical grade device for consumers. Last week, I signed up to be one of their research subjects. You can too. I actually paid for this opportunity which is turning the tables on conventional research funding. You see it’s a direct to consumer model. It’s a disruptive technology that will change how we think about at least some aspects of healthcare.

The big picture here is that things like cheap health sensors will generate more data over the next five to ten years than has been generated in the history of healthcare. Johns Hopkins estimates that 160 million people will use health sensors to monitor at least one chronic medical condition by 2020.

The utility of this data, the uses it can support and the extent to which it becomes seamlessly integrated into healthcare in a way that improves outcomes and reduces costs will be one of the most exciting challenges in medical informatics. We’ve all heard of the opportunity to monitor patient conditions like diabetes, hypertension, congestive heart failure and others in order to preempt acute medical events. You may have even heard that intelligence agencies can identify you not just by your smartphone, but they can confirm it’s you holding your smartphone by your location and walking gait. Yes, walking gait. I know this is a bit creepy, but imagine a healthcare application that identifies that your mother or father has a remarkably different pattern of movement today when compared to the every other day, and alerts you or perhaps a visiting nurse to check if everything is OK. Will there be an app for that in the next 5 years? I can almost guarantee there will be a consumer app in use before physicians know what to do with it. That’s the nature of consumerism.

Even genetics is going consumer. For $99 you can get your personal genotype from 23andMe. They test for approximately 1,000,000 genetic variants that can be used to inform ancestry, health risks and drug response based upon known genetic biomarkers. But the utility of this test will be determined in part at least by the extent to which this kind of testing gets integrated into healthcare and becomes a part of shared decision making between physicians and consumers. I have had discussions with this company about their efforts to build bridges into healthcare and can tell you that it’s coming. But when you know that this type of test and the resulting information is available TODAY, such as whether or not you are at risk for a particular disease, or how will you respond to a particular drug therapy, then you begin to think seriously about wanting this information.

Determining how a new technology impacts healthcare and deciding when a healthcare organization should begin to use it will be interesting challenges for the medical informatics profession to address.

We are now close to reaching a point where a single medical decision can be informed by a thousand data points. Just 20 years ago, it was more like 5-10 data points. Healthcare decision making is exceeding human cognitive capacity. Healthcare providers and consumers alike will need good information technology to incorporate all available data into healthcare decision making.

Next, I said that electronic health records will evolve into communication hubs within an application ecosystem. The EHR of today is not the EHR of tomorrow. We often hear about the need to accurately phenotype and genotype patients. EHR’s were not really designed to do that nor could they easily do that in a fragmented healthcare system. Slowly but surely we are going to see electronic medical records evolve into an ecosystem of interconnected applications built upon interoperability standards that support an incredible array of clinician centered workflows, patient centered workflows, research and surveillance. All of our buzzwords will be supported; e.g. ACO’s, Patient Centered Medical Homes, Population Health, Meaningful Use, Active Surveillance, Translational Research, etc.

It’s the job of the medical informatician to shape and guide this evolution. It’s not just about health information exchange or care coordination across EHR platforms in the sense that is conventionally discussed. The change coming is much more profound and we are already starting to see the effect in some areas.

You see the EHR is going to evolve into a communication platform for clinicians, patients, researchers and others. These platforms are going to increasingly integrate applications and content from 3rd parties, not just the EHR vendors. Here’s some examples that exist or are soon coming: • Today, PDR Network integrates with over 30 EHR’s. Some of the functions it performs within existing EHR workflows includes 1.) updates to drug information in real time as the FDA approves new drugs and drug label changes, 2.) processes Patient Assistance program applications to provide financial support to those who cannot afford their medications and 3.) enables physicians to more easily report adverse drug events to the FDA • Smart applications will automatically link to remote biometrics sensors and medication monitoring devices via inference engines that generate actionable alerts or trend reports based on the streaming data. Similar to the way sensors in manufacturing equipment and elevators have been notifying engineers of preventative maintenance needs for the past 20 years • Clinical decision support platforms containing evidence-based algorithms established by expert physician panels will link into EHR’s. As medical evidence changes, the algorithms are updated once and accessible instantly in all Electronic Health Records. Some things only should only have to be done once, instead of by thousands of healthcare organizations. This is one of those things. • Automatically link with consumer genetic testing/molecular diagnostic company repositories to access interpretation and decision support • Facilitate access to a personal health assistant who can quickly offer help 24×7 based on your personal health profile, predictive models and situational awareness protocols. • Alert patients and physicians to clinical trial opportunities. Everyone, especially those with terminal disease, should know all of their options. • Know if severely chronic patients have missed taking their meds….and text, email or call them….or notify their health manager.

I’m excited for all of this convergence to occur. But it’s not going to come from a single electronic health record vendor. It’s going to take a village of interoperable and diverse solutions.

The final trend I want to talk about today is the data.

All healthcare stakeholders will have appropriate access to healthcare data to create value and manage risk. All means all. This includes patients, providers, payors, regulators, drug and device manufacturers and others who are obligated to deliver value and manage risk. But to do so, they need to have appropriate access to data…..or really the ability to gain insights from data with mechanisms to protect patient privacy. While everyone needs access to insights from data, not everyone really needs to see your medical record. This is an important distinction and existing policies, processes and technologies can enable this to occur. It already is starting to happen but there are still too many uninformed decision makers and influencing forces.

Actually this trend is starting to come full circle for me. I was led to medical informatics based on personal experience. In 1997, I was working in drug safety at the company that marketed a drug called Redux, better known as Phen-Fen, when the FDA pulled it from the market. In spite of successful clinical trials, which demonstrate that a product works and is safe in several thousand patients, this drug was found to cause heart valve damage after it was released into the market. The drug was pulled from the market after less than 20 adverse event reports were reported to the FDA and the Cleveland Clinic presented data from an observational study of about 240 patients.

Once the drug was pulled from the market, the resulting media attention created broad public awareness. Within 24 hours, thousands of calls were made to the drug company by patients and physicians reporting suspected adverse events. In the end, about 14,000 people were determined to have had heart valve damage caused by the drug. The facts that a drug was prescribed, heart valve disease was diagnosed and surgery was performed were documented, sitting in claims databases and medical records, albeit mostly paper based at the time. But no one was looking at all of this data in a proactive way. And no individual patient or doctor could have known that it was the drug that caused heart valve disease. This event impacted me and led to my involvement in initiatives like the Observational Medical Outcomes Partnership and the FDA Mini-Sentinel Initiative. The latter is a pilot program funded by the FDA and managed by Harvard Pilgrim Healthcare Institute that is now actively monitoring drugs, vaccines and blood products using healthcare claims (mostly) and some electronic medical records covering 130 million lives in the US.

We are learning a lot about the strengths and weaknesses of the data, and its overall utility. We’ve got a long ways to go. The data, the systems and the science will get better. The extent to which we can nurture appropriate uses of the data and overall demand for data will lead to improvements in the data.

But this whole story illustrates why all stakeholders need access to data. Not just individual doctors and their patients. The roles medical research and public health surveillance are critical too. Their discoveries can be translated into decisions that impact millions of people. They too need data.

In closing, I encourage all of you to think broadly about the value to be derived from health information technology, the resulting data and all of its potential uses. As you participate in important initiatives, I encourage you to think out of the box and take bold, but calculated steps. Help to make the future of healthcare, present today.

Thank for the opportunity to share some thoughts with you today.

Good luck and remember that medical informatics professionals are leading the way!