Traditional predictive analytics aren't enough for truly effective population health management. The healthcare system needs to take an even more proactive approach.
It’s common for many in the healthcare field to think of data analytics as one large, monolithic body of information and analysis. We frequently fail to take a divide-and-conquer approach by breaking analytics down into separate categories with individual disciplines and goals.
When we do so, we find that, traditionally, there have been two primary types of analytics: episodic-predictive analytics, which gathers and uses valuable data when a patient enters the hospital, and long-term predictive analytics, useful for addressing population health management.
But as a practicing clinician, I see the need for a third major paradigm that could be called preventive analytics, or super-predictive analytics.
If implemented properly, this new paradigm of preventive analytics could lead to a healthcare system that improves patient outcomes, reduces hospital admissions and readmissions, and lowers overall costs.
Preventive analytics also holds the promise of streamlining the workload of physicians and other healthcare providers, possibly allowing the care and treatment they offer to become more effective at the same time.
But how exactly would preventive analytics work, and what would it look like? Perhaps it can be best envisioned if you think of an air-traffic control center or a war room, but located in a hospital or perhaps even an off-site location.
An effective preventive-analytics operation would include teams of clinicians sitting in this control center in front of computers and other devices that monitor incoming patient data.
That data would be generated by small telemedical devices that track the vital statistics and other warning signs of at-risk patients wherever they may be: in the hospital, in the home, in an assisted-living facility, or in a nursing home.
Imagine being able to receive regular data on the weight of a patient with congestive heart failure, the blood sugar of someone with diabetes, the respiration of a COPD patient, or the blood pressure of a patient with critical hypertension.
Virtually any type of patient could be monitored in this fashion, several times per day, including those with high admission and readmission potential, chronic illnesses, or high morbidity and mortality potential. And imagine that the monitoring was being done by experienced clinicians who profoundly understand all these conditions and more.
When those in the preventive-analytics command center spot something amiss, they could notify the patient’s physician for possible intervention. Or they could make a triage phone call to the patient to find out what may be causing the abnormal data. Are they feeling okay? Have they taken their medications? Do they need to cut down on fluids?
The command center could even be thought of as something like Life Alert. But instead of sending the message, “I’ve fallen, and I can’t get up,” it would alert clinicians that something bad is about to happen—and can be prevented.
As a former director of both an emergency department and an operating room, I’ve seen countless patients who could have been spared a hospitalization or surgery if only they and their doctors had been aware of an issue before it became a problem. This is the kind of event that a preventive-analytics command center could forestall or avoid altogether.
The system described here offers a wealth of advantages not just to the patients but also to physicians, hospitals, third-party payers and, in fact, virtually the entire healthcare system.
For instance, with the nation’s shortage of physicians, the workload of doctors is often overwhelming, with many caring for several hundred patients. Preventive analytics could ease that burden, alerting physicians only when there is a problem and only when that problem is professionally assessed by experienced clinicians in the command center.
A system of this type could also reduce morbidity and mortality while preventing hospital admissions and readmissions, thus helping to control the overall cost of healthcare and creating a more efficient scenario for third-party payers.
In fact, with a system of preventive analytics in place, it may be possible to envision a future in which everyone in the healthcare field measures success by the number of people not admitted to the hospital, rather than by the number of occupied beds. This would certainly fall right in line with the current widespread emphasis on the prevention of illnesses rather than the treatment of illnesses.
As we all know, the world of healthcare analytics is awash with data, and the power of that data is just beginning to be realized. Fields like population analytics often promise to achieve true effectiveness—and return on investment—in the long term, usually over the course of many years.
People ask when and how that return will come. The answer is often not encouraging, and they often become impatient. But with a system of predictive analytics in place, there could be an almost immediate impact not just on the analytics themselves but also on the quality of patient outcomes, the cost of care and the efficiency of the healthcare system itself.
We have the means to institute widespread use of preventive analytics, and we have the technology. Now, what we need is the vision and the will to take major steps forward in improving care and reducing costs through the preventive-analytics command center.