If you were a moviegoer back in the 1980s, you may recall a certain Ivan Reitman classic…
No, not Ghostbusters.
Sorry, not Stripes either.
Instead, this blog gives Twins its due, albeit in the context of data quality in healthcare. For those of you who haven’t seen it, the premise is quite simple. A top-secret genetic experiment results in twin boys – Julius (Arnold Schwarzenegger) and Vincent (Danny DeVito) – who are then separated at birth. Years later, when Julius learns of his long-lost brother, he sets off to Los Angeles to find him. Hilarity ensues.
Decades later, the movie still holds up and serves as a useful springboard for examining the idea of the electronic health record (EHR) as the patient’s “digital twin.” When the unlikely siblings finally meet, the sight gag is striking. As the tall, muscular Schwarzenegger stares at his diminutive, balding brother through a glass partition, DeVito quips it’s like “looking into a mirror” – which, of course, it isn’t. For many who have seen their electronic chart up close, and for clinicians who use them daily, it can be easy to relate. What we expect to see in our digital twin is rarely what we get.
When data quality in healthcare falls short
Given the volume of available data – from diagnoses and medications to genomics and lab results – it should be fairly easy to paint a clear and accurate picture of a patient…to conjure their digital double. Yet despite the abundance of information, that’s rarely the case since clinical data is often complex, unstructured, and inconsistent.
As patients, we stare through the glass of computer monitors, searching for our reflections but find only fragments or distortions looking back. And for clinicians who seek to understand each patient’s health history and status, the story they find is seldom, if ever, complete.
A host of obstacles impact the quality of healthcare data and stand in the way of creating an accurate digital twin. Our latest insight brief explores a number of them, including:
- Siloed information within the electronic health record (EHR) and between health IT systems
- Variations in data accessibility
- Inadequate capture of detail or specificity
- Inconsistent data structure; and
- Variations in clinical terminology
Then, of course, there’s the sheer volume of data being created. According to one recent estimate, healthcare will generate more than a third of the world’s data by 2025 – making the ability to improve and manage this information more crucial and pressing than ever.
