Last week, I met a venture capital firm leader who admitted he knew very little about healthcare but was planning to invest in a new decision and treatment information company. I asked him how the main product would work. He explained that the software would gather a patient’s health data, produce a likely diagnosis or diagnoses, and recommend treatments. When I asked how the doctor fits into the product, he explained that the doctor enters the data, and the algorithm does the rest.
This got me thinking about how technology can help improve our challenged healthcare system.
Having accurate, up-to-date, comprehensive, and analysis-ready data is undoubtedly one key to improving healthcare, a trend reflected by the increasing emphasis on data interoperability and patient access to data by Medicare and healthcare payers. The more tools we can use to improve these data components, the better we can treat patients. That’s why Veradigm is committed to improving access to, utility of, and effectiveness of healthcare data: to achieve better care.
Healthcare treatment can be broken down into three essential components: patient health history, physical exam, and testing. Translating these components into analysis-ready data is a significant challenge. Start with the health history: Imagine a patient with pain who can describe the severity, location, timing, frequency, and type of pain experienced along with associated symptoms. However, when asked what makes the pain better or worse, the patient explains that the pain worsens when they are in certain body positions at certain times of day. This type of information does not lend itself to data capture; it needs a “human element” to interpret and record.
Physical exam findings are more data-friendly but can still be difficult to capture fully. One common example I witnessed while in training in Pediatrics is slipped capital femoral epiphysis (SCFE). A teenager can have pain in the groin, inner thigh, or knee; stiffness and decreased ability to rotate the leg; along with changes in gait. However, many other lower extremity conditions can cause similar symptoms. Making this diagnosis usually requires some experience with the condition and the subtleties of the exam.
Testing generates the types of information most likely to be amenable to data capture. However, even testing has “gray areas.” During residency, I recall frequently having to ask Radiologists to review imaging studies even after the studies were “officially” read. Doctors caring for patients on the floor would describe the patient history and physical findings to the Radiologist, along with the areas of clinical concern and suspicion. The Radiologist would review the studies again with those doctors, answering questions. Occasionally, these meetings even led to changes in the Radiologist’s “official” reading based on the new information.
Although healthcare will always require human expertise and input impossible to capture with data alone, a significant portion of healthcare information can be captured and interpreted in structured data. Making the capture and delivery of such data simple and effective is critical to improving healthcare. Doing so helps healthcare providers see the “big picture” more quickly and clearly, while furnishing more time to spend on the essential human “art” of supporting health.