EMR Interoperability and the cost of Healthcare
Healthcare in the United States costs far more than in any other country in the world, and yet the quality of the care Americans receive is rated 11th among first world countries. The high price of healthcare burdens employers, increases the national debt, and historically has left many without access to healthcare. We all want cheaper, better healthcare, but it is difficult to agree on specifics of how to get there. Why is healthcare so expensive?
Of the many factors contributing to the high cost of healthcare in the United States, each comes with their own set of challenges. Drug costs, litigation, use of expensive treatments, and physician pay are commonly cited as major drivers of healthcare costs. These factors are important. However, they often receive more attention than they deserve. In actuality, inefficiencies in the system and the high cost of healthcare administration are by far the largest source of the problem. Indeed, administrative overhead accounts for a full 25% of the cost of healthcare in the US. If we want to attack the problem of healthcare costs, we need to start by making the system more efficient.
Healthcare efficiency is a broad topic, and there are many problem areas. There are structural, technological, and organizational inefficiencies in the system. Congress enacted the Affordable Care Act (ACA) in 2009 to try to address some of these problems. The many changes introduced by ACA legislation included incentives and penalties associated with the use of Electronic Medical Record (EMR) systems, as well as incentives for health providers and hospitals to form organizations called Accountable Care Organizations (ACOs) to improve coordination and efficiency. These two provisions of ACA have introduced radical changes to the healthcare industry. As we shall see, the EMR and ACO changes are causing a new set of technological problems and opportunities.
First a bit of background on EMRs, also referred to as EHRs.* The first EMRs appeared in the early 70’s. With the proliferation of personal computers in the 90’s, electronic storage of patient data by providers went into widespread use. This led to security concerns, and in 1996 congress passed the Health Insurance Portability and Accountability Act (HIPAA) to protect patient data. Over the years, market and regulatory pressures resulted in more and more providers moving from paper to electronic formats. In recent years and with the rise of Software as a Service (SaaS) solutions, many EMRs have moved to the cloud. Today we find a huge number of EMR vendors (405 by one count). This has led to segregation of data, vendor lock-in, and data interoperability issues. In today’s world of EMRs, every vendor has their own data storage scheme, and there is no standard format that will allow migration of one EMR to another.
EMR solutions come in all shapes and sizes. On one end of the scale are massive implementations with thousands of database tables such as those provided by EPIC Systems for large organizations such as Kaiser, UCSF, and the Mayo Clinic. These require database administrators, secure servers, and an IT department with specialized training. On the other end of the scale are “free” cloud-based solutions for small providers such as San Francisco based startup Practice Fusion, which relies on advertising for revenue. Clearly, an EMR that is appropriate for a large hospital will not work for a small group of health providers.
The lack of interoperability between EMRs has large consequences for providers. One effect of ACA was the introduction of three stages of Meaningful Use (MU) requirements as published by the Center for Medicare and Medicaid Services (CMS). MU requirements are a sweeping set of reporting, claim, and other requirements that are difficult to meet. The cost of implementing MU changes for EMR vendors is significant, and many EMRs have not kept up. Unfortunately, this puts providers in a difficult situation. They cannot simply drop an EMR that has not met regulatory requirements, as this would affect billing and patient care. In addition, changing from one EMR to another requires a migration path which no EMR vendor has implemented. In order to move data, providers would need to hire technical experts to perform a custom extract transform load (ETL) operation, often at significant cost. EMR vendor lock-in results in providers failing to meeting MU requirements, which results in government penalties.
Interoperability issues result in even larger problems when the structure of health organizations change due to mergers and acquisitions. As mentioned above, ACA strongly incentivizes providers to merge together into Accountable Care Organizations (ACOs) for the purpose of increasing cooperation between providers. Over the past 5 years, the number of ACOs has grown from 64 to 838, covering a total of 28.3 million people. These newly formed ACOs have found themselves hamstrung by the variety of EMRs under their jurisdiction, with some reporting over 20 distinct EMR systems. The decision to migrate or consolidate is a major decision for these ACOs. They must weigh the efficiency benefit against the risks and costs of performing a custom cleanup and migration. Because the number and reach of ACOs is continuing to grow, understanding and addressing inefficiencies that affect ACOs means addressing inefficiencies that affect the future of US healthcare as a whole.
From an ACOs perspective, the downside of maintaining so many different EMRs is clear. The CMS defines an ACO as follows:
“Accountable Care Organizations (ACOs) are groups of doctors, hospitals, and other health care providers, who come together voluntarily to give coordinated high quality care to their Medicare patients. The goal of coordinated care is to ensure that patients, especially the chronically ill, get the right care at the right time, while avoiding unnecessary duplication of services and preventing medical errors.” source
ACOs that support multiple EMRs are at a disadvantage in achieving their primary objective, which is to provide coordinated care. By storing a patient’s data in fragmented fashion across distinct silos, a complete picture of a patient’s status cannot be known. This puts patients at risk of receiving duplicative treatment. In addition, medical errors can occur when a provider cannot view a patient chart in its entirety.
An ACO’s Chief Information Officer (CIO) is confronted with a series of difficult choices. Probably the most important decisions involve EMR selection and EMR migration and deprecation. Some considerations include the following:
- • Will the EMR solution be supported in 5 years (is the vendor’s business likely to survive)?
- • Does the EMR meet MU requirements and appear likely to meet future requirements?
- • Is the EMR interface simple and intuitive?
- • Does the EMR meet ACO billing requirements?
- • Does the EMR include a patient portal that supports outreach and education?
- • Does the EMR have a clear plan for interoperability?
While we wait for interoperability, ACOs must decide how to cobble together a working system, making targeted efficiency improvements where it makes sense.
What is being done about EMR interoperability? The first and less promising solution lies with Health Information Exchanges (HIEs). HIEs are a way of querying patient data from multiple EMR sources. Unfortunately, they are limited in being primarily EMR-specific and/or geographical in nature. A regional HIE might provide a good stopgap solution for an ACO, but there is work needed to integrate with the HIE, should it even exist in the ACO’s region.
A better option might be to use an “integration engine” such as Orion Health’s Rhapsody Integration Engine, which relies on various communication protocols to join patient data together. Many EMRs support the HL7 standard, which is an older text-based means of transmitting patient data. Web-based solutions are reluctant to share this approach, and are more interested in a modern web-based interoperability solution.
More promising (but more long term) is the work towards a National HIE that a consortium of groups called the Sequoia Project is spearheading. A number of large EMRs have agreed to adopt Sequoia’s Carequality Interoperability Framework. Still, we are at best many years away from seeing interoperability at the national level. While the Department of Health and Human Services (HHS) has a goal of making national data exchange the norm by 2024, this goal depends on the cooperation of EMR vendors, as well as technology and other hurdles.
* Electronic Health Record (EHR) refers to a digital representation of a patient chart, as opposed to EMR which is a record of a patient’s health related data. Some use these acronyms interchangeably, while others treat them as complementary but distinct concepts. For the purpose of this writing we will use EMR and assume interchangeability.