September 3, 2021

Automation for Highly Variable Lab Forms is Now a Reality

Think back to last time you were at a doctor’s office, clinic, or hospital and needed blood work done. Usually the provider pulls out an extremely dense form and ticks off some boxes and scribbles a signature on the bottom. And sometimes that particular test can’t be done in the office so you are told to visit an external site where someone else will mark up that requisition form even more before doing the blood draw, nose swab, or whatever else is needed.

And until now, once your sample is procured, it’s sent off for manual processing and a logistical nightmare begins. Automation for healthcare and laboratory reports has historically been slow to realize due to the inherent nature of requisition forms, which often mix formats and rely on graphical elements as well as text. Legacy automation solutions simply can’t process documents with variable and complex formats, forcing organizations to rely on a disrupted human workforce and supply chain during an unprecedented surge in demand due to the COVID-19 pandemic.

Precise Answers for Complex Processes

Applica leverages cutting edge research based on deep learning methodologies to process documents similar to a human mind, taking into consideration layout, graphics, and semantics – resulting in precise answers for such complex processes. Take an example we recently solved for a US-based leader in healthcare and forensic laboratory sciences.

This company was executing tests in five main categories of lab reports before the pandemic hit. When the demand for COVID-19 tests started in the spring of 2020, their overall volume rapidly increased to 6,000 tests per day and continued unabated. Their process was laborious and had inherent risks because it required a human to rekey the data from each form into their internal system for every requisition. The semi-structured requisition forms include checkboxes and signatures, and over time the documents changed due to updated formatting, new tests added for drug screening, etc. leading to further processing complications.

They had a dedicated team of six people responsible for data entry of the requisition forms, but it was logistically impossible to keep pace with demand, as it would take the time of 20 employees to meet the surge surrounding COVID-19 testing. This resulted in a major bottleneck for operations, because even with each employee spending a maximum of 90 seconds on each form, turnaround times to healthcare providers were significantly delayed.

6,000 Forms Processed in 10 Minutes

Applica was brought in to solve these persistent operational issues by automating the ingestion of the lab requisition forms and automatically extracted and classified the relevant data into a model that could be quickly surfaced up into their internal system. Additionally, with Applica’s solution able to identify any problematic requisitions upon arrival, the error rate went down precipitously, and prevented additional issues arising down the line.

All 6,000 daily requisition forms are now processed in 10 minutes vs. the multiple days it would take the team to work through that volume previously. The team of six have successfully scaled up their operations and are now doing the work of 20 employees, leading to a cost savings of nearly $1 million for the organization.

Does this sound like a solution that could help your organization as well? Connect with an Applica expert today to discuss your team’s specific needs and goals.

Continue reading

Articles
Gaining the Loss Run Advantage
Streamline and expedite analysis of loss runs across the insurance industry.
Research
MLOps - Real-World Application
In the first part of this article, I will try to present and discuss the steps needed to prepare a machine learning model for deployment in production. The second part will cover deployment and automation of the ML system creation process.

MLOps - Real-World Application