This is the age of automation and applications. This is the age of healthcare reconstruction. Fuse these trends and you get a force much greater than its individual parts: powerful technology that catalyzes comprehensive, efficient and cost-effective coordinated care for accountable care organizations (ACOs) and other collaborative care models.
How does technology affect coordinated care and population health management?
Technology that enhances performance across the continuum of patient care is not just reserved for ACOs, but for all emerging models of collaborative care (e.g., Medicaid Health Homes) as well.
There are four concepts to think about when strategizing your organization’s approach to coordinated care:
- Population Determination & Outreach
What patients are you trying to target?
- Care Team Assignment & Composition
How will patients be assigned to care teams, and how are those teams going to be composed?
- Coordination Of Care
How is care going to be coordinated across the community?
- Measurement & Improvement
How will the efficacy of the care coordination efforts be measured for continual improvement?
Each stage is essential to proactive and comprehensive care: care that goes beyond the office visit or hospital discharge and lays the groundwork for farsighted, sustainable health management.
But for now, let’s break the sections up and zoom in on that first stage: Population Determination & Outreach.
Population Determination & Outreach
Use the “80/20 Rule.”
The first step is determining your target population for sophisticated coordinated care. This is where the proverbial “80/20 Rule” comes in: Notionally, 20% of patients treated generate 80% of your costs. In other words, 80% of patients in your population will not be candidates for the highest-intensity coordinated care routines.
- That care-intensive 20% represents patients with high costs and potentially bad outcomes: those with chronic diseases, high acuity, high utilization and comorbid conditions. These are the patients who make up your target population for the highest-intensity care coordination.
- The remainder of the population can be stratified into tiers requiring less intensive coordinated care, such as “rising risk” and “wellness maintenance.”
- It’s not enough to simply identify these high-cost, high-care patients. You must also anticipate these patients by considering those trending in a care-intensive direction (i.e., patients who will require coordinated care in the future). Mine that “80%” of your patient population for trends and leading indicators hinting at a possible transition to the high-cost “20%” group.
Build your target population.
Care coordination software brings together a number of different methods and tools to help you build, quantify and assess your target population:
- Predictive Analytics
As its name suggests, this type of analysis is used to predict: forecasting the risk of the patient developing certain conditions or having certain negative outcomes (e.g., hospital readmission). The method involves assessing a number of factors about your patient population and understanding what kinds of outcomes you’re interested in optimizing. For example, you may run a certain statistical algorithm to determine the risk of any given patient falling into a specific category you want to manage.
It’s important to note that predictive analytics gets a bit tricky when you move into a broader community setting (e.g., not just acute patients, but also those from ambulatory services and other providers who are now part of the increasing scale of data contribution). Think about it: Different care provider organizations use different methods to capture data, and while the analytics are able to normalize to a certain extent, there’s some “algorithm training” necessary to get the optimal predictive value. So when working with patients who have been handled by multiple providers, and as you feather in different and new participants from the care continuum, you may have to re-calibrate the analytic algorithms. By “training” the algorithms on the population, you account for those new and varied underlying data-generation algorithms across the evolving care community.
- Population Stratification
Patient stratification, also referred to as “risk stratification,” uses biomarkers to create subsets within a patient population. It’s similar to predictive analytics in its objective – identifying your target population for coordinated care – but may differ in methodology.
By considering different categories of patients (e.g., high length of stay, high costs, high readmission rates, etc.) and combining them in various ways (e.g., juxtaposing groups with overlapping trends, merging specific sets, etc.), you construct your subpopulations of interest. Population stratification is especially powerful because it leverages both algorithmic and human aspects of data collection: quantitative techniques like predictive modeling, plus manual augmentation of patient groups with other population dimensions according to desired program goals.
While not as rigorous as predictive analysis in its “data crunching” and statistical background screening, population stratification is more intuitive and flexible: allowing you to add targeted categories of interest on the fly, as needed. This adaptability is crucial, especially in these nascent stages of collaborative care, where risk stratification methods vary and best practices are still evolving.
- Program Lists
Many organizations receive their lists of patients from their payers. In this situation, payers take the patient population they’re responsible for and dole out those they want considered for coordinated care. While it’s certainly the least data-intensive methodology discussed so far, loading these lists into your system is a viable way to build out your target population for care coordination.
- One-Up Patient Registration
This method is the simplest, yet definitely has its place in patient targeting. Think of it as your on-the-ground model: signing up patients, evaluating them according to your criteria and determining their place in your patient care continuum.
Like most solutions in healthcare, the best approach is not just one approach. As collaborative care coordination continues to evolve, we’re going to see more creative integrations of, variations on and enhancements to these patient-targeting models.
Remember: Data sources are rarely constant – from numbers to timing and type – so you must be nimble when building your target population. Coalesce all the relevant resources, information and technology you have available to get the outcomes you need.
Reach out to your population.
Once you’ve determined the patient population you’re trying to target, it’s time to do some outreach. Patient outreach activities include the following:
Locating patients – a potentially complex activity if you’re targeting transient or disconnected patient populations, like some Medicaid or dual-eligible (Medicare & Medicaid) recipients – is key to effective outreach.
Most communities overestimate their ability to contact patients, not taking into consideration the amount of false data (e.g., fictitious emails or outdated phone numbers) and sparse data in their systems. This has sparked a significant push to improve patient contact and engagement.
This involves educating patients about what the program represents and how you’re going to help them in ways you have not done so before. This is crucial in making patients feel comfortable and proving that your program and organization truly have their best interests at heart.
We’re all familiar with HIPAA regulations regarding patient data. Patients must give consent to share their data, so they must 1) be properly informed about who is going to see their data and 2) have the final say on where their data ends up.
Managing and tracking your outreach activities is key to assessing care coordination program efficacy and goal realization.
THE TAKEAWAY: Technology is a key catalyst to effective coordinated care.
Building and reaching out to your target population become much more manageable, efficient and effective when you take advantage of the health information technology and healthcare apps available to you.
- Automate population stratification to pinpoint the patients who qualify for this level of collaborative care.
- Streamline patient invitation activities with patient portals and personal health records.
- Track outreach activity and set boundaries that prevent wasteful effort on fruitless engagement pursuits.
- Register and enroll patients in a way that allows you to sync the information you gather from your outreach activities with the systems in your care facility.
- Manage consent with ease and accuracy.
- Keep a close eye on enrollment status so you’re always up to date on where each patient is in your system (e.g., whether they’re enrolled, un-enrolled or have withdrawn consent).
Once you’ve determined, built and reached out to your target population, it’s time for the next three steps in strategic collaborative care: Care Team Assignment & Composition, Coordination Of Care and Measurement & Improvement. Stay tuned for tips and best practices to streamline and refine these stages in your program.