Whenever working with technology, it’s easy to forget the tremendous amount of work and logic that goes through our brains with little to no effort daily.
Simple tasks such as comparing if two medicines are similar can be done to a certain degree with no expertise in the medical field. But code can’t do so without guidance; the most significant challenge when tackling complex problems is incorporating different lines of thought as pieces of code.
That’s why data extraction in LISA is guided by having human-like decisions in mind while developing and integrating algorithmic solutions to everyday insurance problems.
Search Engines in LISA Healthcare
Implementing case-specific search engines has increased our reliability and understanding of documentation throughout Healthcare. Let’s understand “search engine” as a computer system designed to perform the search of a component in a database. For example, when going through the documentation of an insured from any health line of business, this translates to searching for a prescribed medicine in a medicine-covered database. Let us go through an example:
Let’s say an insurance company covers “Acetaminophen” in any presentation. However, “Tylenol” was listed as the medicine of choice on the medical prescription.
In terms of code, solving a case like this one include questions such as:
- Are milligrams relevant to the medicines covered?
- Are medicines with more than one component valid to repay the insured by paracetamol?
- How can I associate a non-generic medicine name with a generic component?
In LISA, we solved and integrated our medic search engine with these questions in mind in three steps:
First, we looked for public medic databases where we had as much information on medicines as possible, with the name of the medicine, price, and principal component as a minimum.
Then, we contrasted it with a database including the most common generic components and their prices on the market (standardized within Chilean currency).
Lastly, we created a levenshtein distance-based search engine within the root of every word in the medicines database, allowing us to get results like the one showcased below.
Taking into account the steps we take as humans when operating is key to solving everyday problems with code. When attempting to solve a problem, we must keep in mind the different issues that may come with it to adapt and create whatever infrastructure is needed.
In LISA, we do not limit an algorithm’s original purpose to be considered for any solution, giving us an advantage when tackling complex problems.
Did you know that LISA Claims is a low code / no code solution and that is one if it huge advantages? Want to know more? Check this video.