Company Overview
An Australian transcription company with medical clinics as clients, processes over 2,000 minutes of audio per day, with client turnaround times of two hours from audio received to finalized transcripts. 

Business Challenge
The company was fielding client complaints from clinics related to turnaround time for the transcription of medical reports, and they had been testing MS Azure and AWS Medical micro-services for options on how to reduce transcript production time. When they ran a small trial with staff for each of these services, they received feedback that it was taking longer to fix errors being made by the speech-to-text systems than typing the files themselves. In addition to issues with accuracy, the staff was having a hard time transitioning from typing to editing text.

Solution
The company engaged with Loom Analytics to evaluate and deploy Claudio Medical Transcription Assistant through an asynchronous DropBox integration followed by an OpenAPI synchronous integration.  

Loom Analytics’ technical integrations team worked with the company’s IT director on both integrations and the customer success team worked with the operations team on onboarding and training the transcription team.

“As our work volumes increased, the decision to switch to Claudio's API was a game-changer. Instantly achieving 2-minute turnaround times for hundreds of files processed in parallel transformed our operations. It was a life-changing experience, eliminating the need for our transcription managers and teams to wait for transcripts.”

— IT Director, Medical Transcription Company, Australia

Transformation
In the first month, the company started a trial of the Claudio Medical Transcription Assistant using the Claudio web app.  In week 1, the company’s Transcription Operations Lead reviewed the transcripts from the first phase of the pilot for accuracy and confirmed that the accuracy levels met the minimum 95% accuracy threshold. In week 2, The Quality Control Lead ran an additional 30 files through in the second phase of the pilot, analyzed the results and requested an update to their self-learning profile to automatically adapt the transcripts based on the feedback.  A week later, the team ran a post-adjustment batch of files to verify that the adjustments were made.  

By the second month, the company’s IT Director was working with Loom Analytics to integrate Claudio directly into their DropBox workflow to accommodate a throughput of 1,500 dictation files over an eight-hour period with transcripts being produced within a 10-minute TAT. This integration was done within two weeks followed by a three-week rollout to a small senior team of medical transcriptionists to measure production times using Claudio. Transcript production times were down by 50% by the end of Month 2 for this senior team.

To further improve Claudio's first draft transcript TATs, the company switched from the DropBox integration to using Claudio OpenAPI, at the beginning of the third month, bringing the TATs down to 2 minutes from audio upload to first draft transcripts available for editing and finalizing. The company onboarded and trained the first four teams using training articles and videos provided by Loom Analytics.  By the fourth month, these teams had transitioned from typing to editing transcripts, eliminating the need for a two-stage process, and reducing turnaround times to bring them within client SLA expectations. 

The remaining teams were transitioned to editing, freeing up half of the staff and doubling production capacity, by the end of Month 6.

Results

• Brought client response times down to 2 hours to meet SLA obligations

• Reduced transcript production costs

• Doubled production capacity
  
  

Security & Compliance
Given the sensitive nature of medical records, HIPAA compliance and the location of data residency was critical. The company set up their Claudio account on Loom Analytics’ Australian servers*. In addition, they opted to go with a 24 hour data retention^ policy to ensure the data was deleted on a timely basis.

Ongoing Support
For the operations and IT teams, the ongoing support from Loom Analytics’ customer support and technical teams was the key to ensuring the success of the transformation from a manual transcription process to one where Claudio assists the transcription team in their transcript production process. The IT team could confidently integrate and deploy Claudio within their workflow system, knowing that they have 24-7 live tech support for any issues that may arise.

*Claudio servers are available in USA, Canada, UK and Australia

^Data retention options include immediate delete, 6 hours, 24 hours, 7 days, 30 days and manual delete