Expert Commentary

Win Whitcomb: Inflexible, Big-Box EHRs Endanger the QI Movement


In “The Lean Startup,” author Eric Ries notes that in its early stages, his gaming company would routinely issue new versions of their software application several times each day. Continuous deployment—the process Ries’ company used—leveraged such Lean principles as reduced batch size and continuous learning based on end-user feedback to achieve rapid improvements in their product.

Ries says companies that learn the quickest about what the customer wants, and can incorporate that information into products more efficiently, stand the greatest chance of succeeding. A software engineer by trade, Ries uses many examples of companies that have succeeded with this approach, none of which are from healthcare.

In stark relief, the chief technology hospitalists interface with daily is the electronic health record (EHR), widely recognized as a system that fails to consider the end-user experience, that is unable to interoperate with other software, and is incapable of using data for quality improvement (QI). The PDSA (“plan, do, study, act”) cycle is the foundation of QI activities and relies on rapidly incorporating observations made by those performing the work to create novel workflows and processes based on learning. EHRs, by digitizing health information, theoretically provide the ideal tool for supporting QI.

The reality is that EHRs have been a colossal disappointment with regard to QI efforts. The space in and around EHR effectively represents “dead zones” for innovation and improvement. Mandl and Kohane note:

EHR companies have followed a business model whereby they control all data, rather than liberating the data for use in innovative applications in clinical care.

Conducting a Google-style search of an EHR database usually requires involvement of a clinician’s information services department and often the specialized knowledge and cooperation of the vendor’s technical teams.

Greg Maynard, MD, MSc, SFHM, senior vice president of SHM’s Center for Hospital Innovation and Improvement, recently provided testimony to the Office of the National Coordinator for Health Information Technology about the challenges current EHRs present to QI efforts and what features EHRs need to incorporate to better serve the needs of patients and clinicians. Dr. Maynard answered a few questions for The Hospitalist:

Q: What is it about current EHRs that make continuous improvement so difficult?

A: EHRs were built for fiscal and administrative purposes, not for quality improvement and safety. The administrative/fiscal roots of today’s IT systems lead to poor availability of clinical, quality, and safety data. In many medical centers and practices, the great majority of information available is months-old administrative data, which does not lend itself to rapid cycle improvement.

Q: Why is the PDSA cycle endangered in most systems?

A: EHRs often do not facilitate rapid-cycle, PDSA-style improvements on a small pilot scale. Most improvement teams get one shot to get the clinical decision support and data-capture tools correct after months of waiting in queue and development time. Any request for revisions and refinements is treated as a failure of the improvement team, and it is often difficult or impossible to pilot new tools in a limited setting.

Q: What features would you like to see in EHRs that would facilitate QI?

A: We need a user-friendly interface for clinicians and for data analysts/reporters. Other industries have common data formats to allow for sharing of information across disparate systems. We need the same capability for clinical information in healthcare. Also, a change in architecture of EHRs and other health IT tools that allows for not just interoperability but substitutable options is required. In the more “app”-like environment, innovation and flexibility would be the rule. An underlying architecture could have different plug-and-play modules for different functions. Some companies are overcoming the current barriers to provide wonderful, easy-to-generate and useful reports, but most are stymied by proprietary systems.


  1. Mandl KD, Kohane IS. Escaping the EHR trap: the future of health IT. N Engl J Med. 2012;366(24):2240-2242.

Dr. Whitcomb is medical director of healthcare quality at Baystate Medical Center in Springfield, Mass. He is a co-founder and past president of SHM. Email him at [email protected].

Help Needed: Open Systems and Modular Architecture

Imagine all the energy we could harness if our most talented engineers wrote modular EHRs instead of “Angry Birds.”

—John Halamka, MD, chief information officer, Beth Israel Deaconess Medical Center, Boston

Today’s EHRs can be thought of as monolithic and closed, with an all-or-nothing, static set of features. On the other hand, think of your smartphone and all the apps (modules) you openly download and, if desired, you delete. This is the vision of a healthy, open, modular EHR ecosystem:

  • Imagine a busy clinician providing real-time feedback about a negative or user-hostile feature in the EHR;
  • Imagine that feedback incorporated—in days or hours—by engineers to create a new version of the application;
  • Imagine a VTE prevention QI team conducting a Google-style search of a group of patients to determine rate of pharmacologic prophylaxis and average VTE risk of that group; and
  • Imagine a hospitalist having five apps to choose from to automatically calculate the readmission risk of a patient: You could choose the best one and delete the others.

The Office of the National Coordinator for Health Information Technology has awarded a series of grants through the Strategic Health IT Advanced Research Projects (SHARP) program to help solve the vexing problems of our closed, innovation-stifling EHR environment. The output of SHARP will be “improvements in the quality, safety, and efficiency of healthcare, through advanced information technology.”

It won’t happen overnight, but perhaps we can hold out hope that there will be a day when EHRs help, not hinder, the QI process.

Next Article:

   Comments ()