Before I came to work for Center for Improving Value in Health Care (CIVHC), I thought of databases in the most archaic fashion. Growing up, we had the classic monochromatic set of encyclopedias that most 90’s kids used for school research projects, or as step stools or flower-pressing devices. This was the database I knew, only secondary to the library, which we frequented weekly my entire childhood.
Flash forward to the age of technology…
When I was in elementary school, my dad proudly brought home an encyclopedia software program, a CD that contained all of the information in our encyclopedia set, that could be easily accessed and took up significantly less space. The concept remained the same in my mind – data was a collection of information, it could be added to as more information becomes available each year, but what was previously presented remained static, archived in a series of encyclopedias or programs. As I grew and technology advanced, the internet and college library-based search engines only accelerated my access to the vast amount of information that existed in the world, and I began to have to learn to parse through these databases to access the most credible, relevant, accurate information available.
It wasn’t until recently that I realized I was still holding onto this same framework in regards to the Colorado All Payer Claims Database (CO APCD). The CO APCD, is ever-changing and dynamic, and when the millions of claims that are submitted monthly get processed, new historical data gets incorporated and the entire database gets updated. This essentially means that analyses that haven’t been memorialized in a report or data set might change if the same analysis is conducted on the same time period of data in the future. Every other month when the data warehouse is updated, not only is new information added, but information that may have been corrected or amended by health insurance payers gets processed, creating the most up-to-date and complete view of health care spending and utilization in Colorado. This is true for files that may be resubmitted for the year, or even historical information, as far back as when the payers first started submitting to the CO APCD. This means that there are times that rates, populations, or costs presented in historical data might change ever so slightly, but in doing so, the most current data is the most reliable.
This retrospective processing comes with challenges as well as benefits. As stewards of the CO APCD, CIVHC strives to provide public and non-public data as frequently as possible. If we were to avoid releasing data until we had absolutely finished and complete information included in the CO APCD, we would be in a holding pattern waiting for the next data warehouse update, in perpetuity. Instead, we follow a stringent Quality Control process to make sure the data meets over a hundred checks and balances during submission and processing to ensure it’s actionable and reliable at any given point in time. Many individuals who work with data may understand this detail about database management, but there are also many, like myself, that must rewrite the framework of how databases are ever-changing, with the understanding that it will allow those accessing the data to present or research information that provides the most up-to-date picture.