As we’ve said before, life expectancy underwriting requires correct and complete information. Health care data alone is neither of these things. That doesn’t mean it’s entirely useless, but it’s a means to an end –not a complete solution.
Health care data includes the medical information stored in all sorts of systems, patient portals, and databases. A lot of this raw data is gathered together by health information exchanges (HIEs). There are over 6,800 of them used in the U.S., but most of them do NOT make the data useful. Think of a landfill in which there are, in the aggregate, a lot of precious metals spread throughout. The problem is how to sift through the trash to get to the gold.
Compiling the Dirty Data
There are other issues with a data-only approach as well. Having large amounts of “data,” does not mean the data is either correct or complete, or even useful. The original sources of most data are predominantly human beings. Initially, people enter information into computers, by typing, talking, checking boxes, filling out forms, etc. In many cases, this data is neither error proof, nor “clean.” It can be repetitious, full of errors and contradictions, and a lot of that “noise” is passed along to the HIEs. However, because most purveyors of data can sell it “as is,” they do.
From Noise to Knowledge
To make data useful, you not only need to get it, but you need it in a form that is useful. Generally speaking, more work is needed to turn data into useful information. Remember our favorite acronym, DIKA, which is data becomes information, information becomes knowledge, and knowledge enables action.
There are some uses for data that is minimally processed or “normalized” such as providing a user with clues as to what might be affecting a given insured person, or where more complete information can be found. These approaches are fast, low cost means of filtering out potentially unqualified applicants for life settlements and maybe beginning the process of deciding where and how to obtain the most useful information for underwriting purposes.
Different Kinds of Data
Other challenges to consider include the fact that there are many different kinds of data such as medication data, test data, which includes laboratory test results (e.g., blood tests), and billing data. Billing data can be harder to collect but it may include procedures, diagnoses and other billable events related to using health care services. In our next article, we’ll touch on the various types of data, why they can be difficult to obtain, and how some data can be used to identify clues to other better, but less easy to work with, forms of health care information.
Learn more about how ISC Services digs through decades of healthcare records and data to create accurate, individualized life expectancy assessments by contacting a member of our team today.
