Are You a Data Hoarder?

Everyone has moved and had to go through the painstaking process of determining which items are valuable, and which items are better left behind and discarded as junk, right? If not, you are lucky, but you probably have needed to spend at least one Saturday or Sunday per year making your best impression of Marie Kondo and parsing through everything you have accumulated. More than likely, it is surprising how much junk accumulates in a short period. Things that provide no value except to clutter up your space and weigh you down.

Conceptually, this is very similar to the process of healthcare data management. Over time, continuously adding data inevitably results in some data becoming stale and siloed. Some of the most common, detrimental scenarios arise in the Master Patient Index (MPI). Duplicate or fragmented records routinely get created for the same individual. Demographic and clinical data is entered in the incorrect patient record, causing providers to miss critical elements of their patient’s prior care. Inadequate and incorrect data lead to poor treatment plans and a decrease in quality of care and patient safety. Maintaining patient indexes and clinical records is quite a difficult task. Generally, between 0.5-9 full-time resources are needed, depending on the size of the organization. Even then, this protocol pales in comparison to the efficiency of automated data maintenance software.

Analogize the moving scenario above with the task of migrating or merging healthcare IT platforms. Just as you want to find that medium of leaning out, you do not want to discard items only to need them again in the future. What is required is a set of rules and protocols to streamline the process. The same concept applies, identify as much useful and validated data and store it to alleviate the pains and processes of migrating host systems, but don’t bring over the wrong data as that can only lead to decreased efficiency and less satisfactory outcomes for providers and patients.

You probably think that your databases look like the worst Hoarders episode to have ever aired on A&E, and most likely, you correct. The good news, with the right solution, the process is straightforward and effortless. Not only that, but it can also be an economical, one-time effort with ongoing checks and protocols to make sure that you never get back to that unmanageable, out-of-control place.

This beautiful day in age has taught machines how to execute the monotonous, time-intensive, and detail-oriented tasks that would be impossible for a human to ever complete with any semblance of success and accuracy. A machine learner also has a better memory than any individual user, so far less erroneous data or results slip through the cracks. The machine learner can remember [almost] everything it has previously seen. Deterministic and probabilistic algorithms match potential duplicate data values and records. Referential checks on data points via external database APIs identify the stale and obsolete data and determine which records the system should retain. This process efficiently consolidates Master Patient Indexes into one accurate Enterprise Master Patient Index (EMPI). The same logic and practice are transferable to many data sets and can be useful when combining separate Health Plans Masters or reconciling a patient account between multiple systems.

Keeping a clean database conducive to a high quality of care and accurate financial reimbursement is not a one-time task. It will take ongoing effort as maintaining data sets such as patient master indexes is a continual process. New data is pervasively added, allowing the opportunity for new overlays and fragments. Luckily, available software solutions can eliminate the need for expensive and inefficient human capital. IntelyMatch by IntelyHealth relies on cloud-based storage and leverages AI to index data

from multiple host sources, validate and store it in an easily accessible and understandable format. IntelyMatch is applicable to any data set and any quantity of data; the more bad data, the better the learner adapts. The process will pick up on viable conjunctive concepts within the data and use that to parse through and identify problematic areas (duplicate records, invalid data, generic data, etc.) and produce a confidence score on estimated data accuracy. IntelyHealth sits above the traditional healthcare IT platforms which allows your data to be resistant to the familiar pitfalls of user errors, poor data management, and organizational changes. You will no longer have to stress about the quality and accuracy of information your professionals use to make critical decisions.