Electronic health records are typically touted as providing two primary and vital services: readily accessible patient records and protection against contraindicated medications. But at Intermountain Healthcare, we're benefiting from a growing versatility in the application of electronic health records that has the potential to transform the healthcare landscape far more broadly.
Key to it is trust – in the data, in its utility, and in the people developing and advancing it. Achieving that trust requires an organizational culture from top-to-bottom that is data-driven.
Intermountain Healthcare pioneered electronic health records in the 1970s, so their application is not new to us. Still, technological advances in recent years have evolved significantly, and that evolution has made more versatile applications possible.
Today, the many benefits that we are receiving from electronic health records include the following: analyzing relationships between conditions; improving patient procedures; reducing rates of infection; reducing antibiotic use; tracking vital signs; and reducing supply chain costs.
Extending the benefits and value of electronic health records requires an organization that has a data-driven culture – a culture that understands that what can be measured can be improved and that improvement depends upon data that is both real and trustworthy.
Creating that culture starts at the top: the CEO and the Board of Trustees must be focused on data and view it as essential to the performance of the healthcare system and to their own performance. But building that culture requires two other key groups of people: clinicians, who understand how to define analytics, develop them and use them; and statisticians, who can build algorithms that are useful to those clinicians.
The statisticians – medical informaticists – are more easily trained and hired. The clinicians have to be trained to combine their clinical experience with a comfort with data and how it is extracted, transformed, and loaded. That takes a substantial investment of time, sometimes months or years.
Extracting the data is relatively easy, as is loading it, but transforming it is complex. That's where effective collaboration between clinicians and statisticians is so vital. That's where the clinicians' skill and reputation are key.
Interestingly, I've found nothing generational about what makes a good data-driven clinician. We have many seasoned clinicians – with decades of clinical experience each – who are essential to our electronic health record initiatives. Their experience ensures that the data is real (that it captures what the clinicians want) and that it is trustworthy (that it provides the needed information in a form that is useful and dependable to the practitioner). Their reputation within the healthcare system enables them to command respect and help push proven improvements out across the system.
At Intermountain, we have created a group of clinical programs, conceived by Brent James, MD, executive director of our Institute for Health Care Delivery Research. These are a hierarchical structure of teams of clinicians and statisticians that propose data to be analyzed, work to refine it, test it, and then push proven best practices out broadly across the healthcare system.
Each clinical program has both a physician lead and a nurse lead, additional clinical staff members reflecting a broad array of disciplines, and staff statisticians. Every clinical program has its own budget and reports directly to the chief medical officer and the chief nursing officer.
The clinical programs are so well-respected that their recommendations of best practices are broadly accepted and contribute fundamentally to the trust needed for a data-driven culture to be effective. That culture and the clinical programs generate proven products that practitioners will adopt.
The use of those products is where the value of electronic health records can grow exponentially. With their use healthcare will improve dramatically – through models that can be widely shared and implemented.