The federal government is spending nearly $30 billion on electronic health records to improve the nation's healthcare. If I had infinite resources and time, and a greenfield for innovation, here's how I'd design the EHR of the future:
Physicians are on call round the clock, have to be in many different places, and use a variety of computing devices. Therefore, the ideal EHR would be Web-based, browser-neutral, and run flawlessly on every operating system.
It would incorporate decision-support tools and patient-specific preventive care reminders. And it would provide event-driven alerts that send critical data to doctors when immediate action is needed, such as when a patient on digoxin has a low potassium reading that increases the likelihood of dangerous changes in heart rhythm and other toxic effects from the medication.
The EHR would have an easy-to-read summary of all the patient's active problems, medications, visits, and labs. This summary would be exportable to personal health records, such as Google Health and Microsoft HealthVault.
Caregivers would pick from standard, predefined terms to describe patients' problems, and all the patient's clinicians would use specialized social networking tools to collectively maintain these problem lists--a kind of secure Wikipedia for the patient.
An e-prescribing app would link directly to payers' formularies so that doctors would know which medications are covered. It would determine eligibility for high-cost therapies in real time, link to a patients' medication histories, and check for drug interactions and allergies. A pharmacy-initiated workflow would reduce calls to physicians for refills. Here, too, the EHR would use social networking to let caregivers update, change, and comment on patient medications.
Patient visits would be documented with the reason for the visit, the diagnosis, therapies given, and follow-up expected. Notes would be entered using structured and unstructured electronic forms. All data would be searchable. Disease- and specialty-specific templates and macros would make documentation easier. Voice recognition would allow for automated entry of recorded notes. Workflow for signing and forwarding notes to other providers would be easy to use.
Lab results would be displayed by date and type of lab, trended over time. They'd be available in screening sheet format, where disease-specific results are combined with decision support to help doctors see trends and decide next steps. For example, a diabetic screening sheet would include glucose, hemoglobin a1c, lipids, recent eye exam results, podiatry consults, and urinalysis. Clinicians would electronically acknowledge lab result notifications, ensuring that appropriate next steps are taken.
Radiology images would be viewed with a single electronic viewer. They'd be easily retrieved and managed using business rules that ensure compliance with medical record retention rules. Orders for medications, lab tests, radiology, and general care all would be done electronically.
The EHR would be able to retrieve medication lists and clinical summaries from institutions that are part of local and regional healthcare information exchanges. It would also import data from personal health records, such as telemetry data from home devices like glucometers. Every night, data would be exported to data marts to support clinical trials, clinical research, population analysis, performance measurement, and quality improvement efforts.
At Beth Israel Deaconess Medical Center, we've already achieved much of this functionality. But we'll never be done, because the perfect EHR is a continuously evolving target.
Dr. John D. Halamka is CIO of Beth Israel Deaconess Medical Center and Harvard Medical School, chair of the New England Healthcare Exchange Network and the U.S. Health IT Standards Panel, co-chair of the HIT Standards Committee, and a practicing emergency physician. Write to us at firstname.lastname@example.org.