top of page

Unraveling the Meaning of Data in Case Management


Not a day goes by where we don’t hear or see an article talking about how we are impacted by data. Sometimes its “small data” but, more often than not, is “big data”. That data permeates our lives. It’s when you order bubble gum from an online retailer, their software will suggest that you might also be interested in purchasing a bubble gum dispenser. That use of personal data can be really helpful for a consumer, and a bit scary at times when we think about how much others know about what we do and how we think.


But the words “big data” miss a subtle but important point. The value of data is not the fact that you have it, or even that you have a lot of it (hence “big”). The value of data is that when you have it, there are means by which you can extract meaning from that data. In my example above, the fact that I purchased bubble gum is an element of data and not all that interesting. But it becomes interesting when combined with lots of other data elements from a lot of other people, and can be used to extract meaning. In this case, the meaning being that because I bought bubble gum I am likely to also want to buy a bubblegum dispenser. 


The distinction between data and meaning is an important one that is often underappreciated in the human services field. We might do a very good job of tracking data elements such as how many meals our home-food-delivery service delivered last year. That is easily measured and quantifiable. The challenge is to extract the meaning: did these meal deliveries help us reduce the number of people facing challenges with food insecurity, did they improve the lives of those receiving the food? And an important one for many agencies when seeking funding, did those deliveries reduce the cost of healthcare?


The challenge then is how do we find meaning in the data we collect? It starts by exploring and understanding how one data element is related to another. All meaning is derived from relationships.


For large on-line retailers that collect data on hundreds of thousands and even millions of transactions a day, finding relationships is possible using any one of many technology solutions. There are artificial intelligence and machine learning algorithms that can take enormous amounts of data elements and using a variety of different models and algorithms, can find relationships and patterns of behavior that would be otherwise undetected by a human even if they looked at that data for thousands of years. The success of these "relationship building engines” has been spectacular. These tools have been able to build models that not only predict who will buy the next bubblegum dispenser, but also very surprising things, like how subtle changes in a person’s speech patterns – changes that are too small for the typical human to detect – can be an early predictor of disease. And that is just one of countless amazing examples. The technology does amazing things.


For the human service agencies, however, we (fortunately) do not have the “luxury” of hundreds of thousands of data elements to work with. When a client walks in the door and needs help with an employment issue, or food insecurity issue, or substance abuse issue, they often have a unique blend of social and medical conditions, mental health and behavioral considerations, and social situation that varies significantly from one person to the next. Given a view of enough of people in these situations, certain patterns begin to emerge, these patterns can fuel various types of artificial intelligence, machine learning, and statistical analysis and give insight into overall patterns and trends within a population. Systems that do this sort of analysis, referred to as Population Health systems, can extract many insights into a community and in some cases, where enough data is available, even into a neighborhood. This type of data can drive community investment and strategic initiatives for a neighborhood,

However, for the case manager, "in the trenches", who is working with one vulnerable and unique individual face-to-face, needs more. For these people, meaning can not be quantifiably extracted from data because there is too little. So how do we support case managers with meaningful data?


To deal with this issue, there is a class of technology solution emerging that focuses on providing frontline case managers with “insight” at the point of care. The Continual Care Solutions imPowr platform is the first of these types of systems and is leading the charge to provide case managers with the tools that support their mission. Because we don’t have the “luxury” of enormous data sets to do predictions at the individual level, we use a more inferential approach that relies on the case manager for final decisions but supports the case manager with critical insights and inferences.


The process works by collecting discrete data from the client and family members, physicians, and other care takers. This can be collected via intake forms, surveys, integrations with EMRs and similar clinical systems, and consumer reported assessments. The highly individualized data is organized and structured, and any obvious data patterns and relationships are identified. The data is then combined with the aggregated population trends generated by the Population Health systems described previously. By comparing individual attributes – and combinations of attributes – with related community level patterns, the system can begin to make some inferences. These inferences may be “soft” to start, including things like level of “impactability” or self-sufficiency relative to a cohort of peers, or simple pareto style assessments and comparisons. When the inferences and insights can be presented to the case manager at the time of need, the moment when the case manager is directly engaging with the vulnerable client, the potential is very powerful. Suddenly what was previously the domain of experienced case managers can now be replicated for those less experienced. New insights that perhaps were never thought of before, now become available. The result: a powerful tool for assisting the case manager in their important direct work with clients.


In summary, data by itself is a means to an end. The “end” is the ability to extract meaning from the data. That is inherently challenging in the human service world where clients are unique and distinct in so many ways, but with the right kinds of tools, and a focus on evidence-based, data-based interventions, it can be done. We are proud to be on the forefront of this evidence-driven approach and to be able to support the important work of case managers in this way.


To learn more about the comprehensive imPowr platform, visit here.

bottom of page