Three Steps to Actionable Healthcare Data Analytics
As a result of the Affordable Care Act, healthcare is pivoting from a quantity-based model to a quality-based model and the initial results are encouraging. Indeed, in 2013, overall healthcare expenditures grew at the lowest level since 1960. One major aspect of this revolution is a portfolio of technology solutions that enable Population Health Management (PHM). These solutions aggregate patient data across multiple resources and analyze it into a single, actionable patient record so that care management, collaboration, patient engagement, performance management and demand management can be improved both clinically and financially.
Current State of Population Health Management
PHM initiatives started gaining traction about two years ago. Most participating healthcare organizations, which can include hospitals, clinics, physicians and insurers, are in the beginning steps of the process. The Department of Health and Human Services (HHS) formed the Healthcare Payment Learning and Action Network in March 2015 with the goal of tying 30 percent of fee-for-service Medicare payments to quality and value through alternative payment models by 2015, and 50 percent by 2018. This mean that the healthcare industry needs to mature capabilities around analytics and patient engagement technologies that enable population health.
With rates of serious health issues still on the rise in the U.S., it’s critical for healthcare organizations to ramp up quickly in PHM. Nearly a third of the population is considered obese and one-fifth still smokes, the combination of which is closely correlated with many forms of cancer. Asthma rates have grown by 25% over the past decade, and more than 30% of adults are affected by hypertension, a leading risk factor for heart disease. And according to the CDC, an estimated 9.3% of the population has diabetes, and a third of the population has prediabetes, with 9 in 10 not realizing they have it. Because of the new regulations, in the next 4–5 years every healthcare organization will need to have either a mature PHM model or at least have implemented a model and have learnings about what worked, what didn’t work, and where they want to reinvest in this capability.
So what are the steps necessary for organizations to build effective Population Health Management capabilities? Ultimately, an advanced state of population health will need to factor in every single disease state and variable and make insights very actionable. Success begins with the right strategy.
The key is actionable analytics.
First step: Look closely at how your organization approaches analytics
Analytics must be viewed across the entire organization. Most organizations understand they need analytics, but they also mistakenly believe they can start building all their capabilities off one of their core systems, such as an EHR system. In reality, because population health is much broader in scope, multiple systems must be brought together in order to understand the dependencies and relationships between data sets. For example, disparate critical administrative and financial data along with data from the organization’s ecosystem must be harvested and integrated in a data repository with patient records and more. Therefore, most organizations need help on where to house the data and how to structure it.
Once this data has been aggregated, analytics in PHM programs can be used in prescriptive and predictive fashions. For example, a patient with a history of heart failure and hypertension would be categorized as high risk. If that patient has more than two hospital visits per year, predictive analytics would indicate a high probability that the person would be hospitalized within the next 12-month period. And prescriptive analytics might reveal that if the patient is likely to be hospitalized, she/he can be monitored daily at home and provided coaching regarding hypertension and diet, with the goal of avoiding the hospitalization. This results in better patient care and cost savings. Of course, this is just one example. There are various predictive risk models, such as the one outlined in Health Quality Ontario’s guide to HARP (Hospital Admission Risk Prediction).
With actionable analytics being a lynchpin to delivering PHM, an effective analytics strategy needs to leverage a long-term investment plan, technology and data assets, platform choice, data governance, build vs. buy decisions, and program delivery.
Second step: Get the right tools
Once you understand the inventory of the data you have to work with, you need to determine the right tools to extract and infer insights from them. What is the right technology to overlay on top of these data sets in order to start extracting and presenting information in a meaningful way? This will highly depend on the size of your organization. For example, mid-to-large healthcare organizations might use Cognos, an SAP BI platform, Tableau, or QuickView. Some organizations might use an off-the-shelf solution, such as Crimson Population Health, IBM Explorys EPM Suite, Optum One Population Health, Wellcentive Advance Outcomes Manager, or Verisk Health Population Health Analytics. Technology decisions will also rely on assessments involving current staff capabilities, organizational structure, technology infrastructure, and more.
Third step: Build a roadmap for successful alignment of patient care and IT
Every organization’s PHM roadmap will depend on the contracts between health delivery organizations and payers regarding various models built around patient types. For example, managing diabetes patients first, and then expanding capabilities to manage patients with both diabetes and hypertension, and then branching further to include other disease states and combinations. The roadmap will define a portfolio of analytics applications that support each stage.
Note that organizations must be acutely aware of government regulations surrounding data capture and delivery. HIPAA regulations must be critically regarded when physicians and their healthcare partners pull data from various sources, and they should be a mandatory element of any PHM roadmap.
Population Health Management is becoming an imperative for healthcare organizations and will undoubtedly be a key driver in the evolution of healthcare in the next several years, with the goals being the delivery of enhanced patient care at reduced costs. Organizations with mature PHM capabilities are already seeing these benefits. Regardless of your present state in PHM, developing a comprehensive strategy and roadmap with the insightful analytics and tool choices based on your organization’s specific needs and abilities is the first step to putting you on the right track for success.