SpeechFlow 1.0

Hospital discharge summaries communicate information necessary for continuing patient care. We implemented an automated voice dictated discharge summary process in a health care system.

Comment 13

Inspiration

In today's digital world, the volume and complexity of physical document exchange in healthcare is surprising. An essential part of this process is the documentation of a discharge summary. A discharge summary is a clinical report prepared by a health professional at the conclusion of a hospital stay or series of treatments. It is considered a legal document and it has the potential to jeopardize the patient’s care if errors are made. Delays in the completion of the discharge summary are associated with higher rates of readmission, highlighting the importance of successful transmission of this document in a timely fashion.

During the current pandemic situation, discharging patients from a hospital became a very complex task. It could be because of the availability of the physician due to busy schedules, frequent visits to other clinical centers as well. This inspired us a lot to come with a low code process which can automate the manual discharge creation for health care staff. And also, I have been using Apple Siri in setting alarm and other basic things. We thought of using the same speech based automation in workflow built in Pega.

What it does

It can automate the manual process of discharge creation using Pega App Studio and cloud based dictation service provided by Philips speech live. Once a healthcare staff creates a discharge request, it goes to the physician worklist. Physician can either capture the details using the web application or through a mobile channel built Pega low code studio. Physician can also provide the discharge dictation using advanced voice enabled cloud based dictation solution powered by Philips Speech Live. Once the speech based dictation is captured using any of the speechlive devices like SpeechMike Air, Pocket Memo, SpeechAir mobile app or even by a web browser it will be passed through Pega text analytics - Natural Language Processing to extract the critical information in it. Once the critical details were extracted, it will be routed to a typist work list for any corrections before sending the final copy to the patient.

Physicians doesn't need to be at the hospital site as they can provide the dictation through speech from anywhere, at any time. It's not only easier to talk than type, it's also much faster.

How we built it

We have used the Pega low code app studio and created a discharge request micro journey. This micro journey is integrated with Zapier - A no code automation platform. This solution uses Philips SpeechLive - A cloud based dictation and transcription solution which converts speech to text. We created a new zap which can automate the voice dictation captured using any of the Phillip's speech live devices like SpeechMike Air, Pocket Memo, SpeechAir. This unstructured speech text will be passed through Pega text analytics - a natural language service to extract critical information required by the discharge summary report.

Challenges we ran into

Getting started with Philips SpeechLive and integrating this Pega Process flow was a challenge thing. Later we found a way using Zapier which makes automation easy without any code. We created a new Zapp and integrated SpeechLive with our discharge summary.

Accomplishments that we're proud of

1)We were able to automate a complex documentation process which is often the primary mode of communication between the hospital care team and aftercare providers. 2)We have integrated Pega with Zapier automation platform. 3)This complete solution was built without writing any complex code.

Few numerical values are not converted into text by speech algorithm. we have utilized the speech live fields to pass the numerical case id.

What we learned

We can automate the process using cloud based voice dictation software powered by low code Pega app builder. We used to use Siri, Alexa to set basic things like daily alarm or any notifications. One we adapt voice based solutions in our micro journeys, it will improve the case processing time with higher productivity.

This is the converted text from the speech by Philips speech live service, which is sent to Pega text analytics to extract important information, before processing the case, to typist. We have used the Pega out of the box, extract rules and also created few based on our use case. We are routing this case to typist if there are any words, which are not correctly translated,or extracted by Pega NLP.

What's next for SpeechFlow 1.0

We are exploring speech flow 2.0 to complex use cases across healthcare, finance, insurance, legal etc. This will help to solve many industry problems, using Pega low code, no code capabilities with voice automation workflows. It's not only easier to talk than type it's also much faster which results in higher productivity.

In fact, research has found that speaking is seven times faster than typing. With speech we can save valuable time, cost savings, mobility, and speed. Pega digital process automation, powered with cloud based dictation software's, text analytics are going to play a crucial role in driving digital transformation. Thanks for giving this opportunity in participating in Pega Hackathon 2021.

Note: We have used ROYALTY FREE music and video clips in the video demo.

Built With

+ 48 more
Share this project:

Updates