Poor Data is a Population Health Risk. Ready Access to Quality Data Can Help

Peter Calderone

To make quick, but also well-informed decisions, providers need ready access to high-quality patient data. While this is true for individual patients, it is also the case with population health. A lack of quick access to quality data represents a population health risk.

Specifically, successful population health initiatives require data analytics that help identify populations in need of care, as well as measure the care provided. This ensures the right care is delivered to the right patients.

To cite just one example within population health management, accurate, comprehensive analytics enable providers to identify social determinants of health that affect patients, enabling clinicians to optimize preventive care instead of waiting for patients to become ill.

How Intermountain Health improved population health with analytics

Intermountain Health leveraged a self-service data platform to enable early intervention by deployed algorithms that accurately identify chronic kidney disease (CKD) patients and stratify them based on gaps in care. Notably, this collaboration resulted in an 86% reduction in unplanned admissions for the risk population.

Among patients presenting with stage G3A or G3B CKD, less than 1% progressed to dialysis. Additionally, Intermountain discovered valuable insights for comorbidity management. The success of the CKD program has inspired expansion to other clinical areas, emphasizing the power of technology and data-driven decision-making in healthcare.

Further, Intermountain has closed gaps in cardiovascular care, resulting in reductions of cardiac events, and among newly diagnosed patients with diabetes, the health system has seen a 10% increase in physically active lifestyles.

With a large percentage of its patient population at full-risk, Intermountain focuses intently on value-based care, population health, and high-quality care. The health system’s leadership understood that the first step to improving population health involved gathering and reorganizing patient data. However, that was just the beginning.

Intermountain’s leaders also knew that its population-health-improvement initiative would require a data analytics platform that enabled clinicians and staff to overcome the following four barriers:

Over-mediation: Analysts and clinicians were unable to guide and complete their own queries and had to rely heavily on IT database experts.

Time-intensiveness: The cycle of time to insights was too long and discouraged optimal engagement from clinicians, both at the back end of processing (after a long wait for results) and at the front end (re-engagement with the same time-intensive process).

Unstructured data: Finding meaningful patterns in the data grew more difficult as data sets grew in size, type, and source.

Privacy regulations: Discovering and sharing insights with collaborators was restricted and often not possible due to patient privacy regulations.

Intermountain implemented a self-service data platform to overcome these hurdles. Self-service data platforms are unique in that they enable non-IT users across an organization to directly explore granular patient data in a model that drives more rapid insight discovery. Among Intermountain’s first initiatives was a population health program targeted toward patients with chronic kidney disease (CKD).

Analytics to improve identification of CKD patients

CKD affects 1 in 7 U.S. adults, and the risk of developing the disease is even greater for Americans with diabetes or high blood pressure. However, despite the prevalence of kidney disease in the nation, as many as 9 in 10 adults who have CKD are not aware they have the condition. That is largely because early-stage kidney disease usually has no symptoms, and many people don’t know they have CKD until it is very advanced, according to the National Institutes of Health.

At Intermountain, 70% of late-stage renal disease patients were entering dialysis without knowing they had kidney disease, generating worse patient health outcomes and higher costs. To correct these trends, the health system sought to improve identification and engagement of patients in early stages of CKD, ultimately reducing hospitalizations and preventing unnecessary morbidity and mortality. As a result of early detection of kidney disease, clinicians can deliver critical therapies to slow the disease’s rate of progression in many patients.

Leveraging its data analytics platform, Intermountain managed and quantified, from a cost perspective, the various disease processes associated with CKD. Additionally, the health system used the platform to identify at-risk patients, utilize structured and unstructured data elements to implement care transformation plans for patients, and show potential return on investment from enrolling patients into these plans.

The CKD program has driven significant results, such as:

• Identification and early engagement of CKD patients enabled early intervention. Thus far, less than 1% of patients identified at earlier stages of disease (at or below stage G3A and G3B) progressed to dialysis, and 86% of newly enrolled patients avoiding hospitalizations with an almost 60% reduction in overall admissions since kicking off the program, resulting in savings to the organization at $1.1 million per year since launch.
• Among the patients who presented with stage G3A or G3B, none progressed to dialysis.
• From a research perspective, Intermountain has revealed probabilistic measures in comorbidity management upstream of kidney care as predictors for adverse events and stage progression.

The early success of Intermountain Health’s CKD program has led to other initiatives targeting coronary artery disease, diabetes, hyperlipidemia, and opioids in surgery, all of which are paying significant dividends. Since adopting a new approach of self-service data exploration, the health system has improved speed to answer and uncovered new insights to boost its population health programs.

Peter Calderone is vice president of customer success at MDClone.