Project Spotlight: Aviva (update)

In the intensive care unit (ICU), patient rounds are a vital process that allow healthcare providers to check in with and discuss patients. Aviva is a voice-interactive virtual assistant that listens to ICU team rounding discussions, identifies gaps in the use of evidence-based practices, and selectively prompts the team to consider evidence-based practices. The end goal is to increase evidence-uptake to improve ICU outcomes. After debuting on the blog in 2020, we recently connected with Dr. Andrew King, informatics postdoc in the Department of Critical Care Medicine at the University of Pittsburgh, to get an update on the project’s progress.

Can you please provide a brief refresher on your project and its goals? 

Aviva is reinventing how technology can help ICU teams on rounds. Daily rounds are an essential way for ICU team members to agree on care plans and make key decisions., but these decisions are imperfect. We are creating a voice-based digital assistant that “listens” to multidisciplinary rounding conversations and suggests better to the team. By listening to team discussions, Aviva assembles a timelier and more complete digital representation of an individual patient than decision support tools that only rely on electronic health record data. For example, Aviva turns the monotonous task of reviewing evidence-based practice checklists into an efficient review of the few practices that are indicated for the current patient but have not been discussed by the rounding team. Aviva’s intelligent functions are an important step towards reducing clinician burnout and undesirable variations in patient care.

What was most exciting or unexpected that you learned during the project?

Although it’s difficult to capture and interpret spoken conversations in the ICU, these conversations are the “next frontier” in healthcare quality improvement. To this point, spoken conversation has historically been absent from clinical documentation and computerized clinical decision support. There are important reasons for this – the technologies to automatically capture and interpret the human voice are relatively new. However, we are now in a world where speech-to-text and natural language processing (extracting meaning from text) are cloud-based commodities. With this technology, it’s exciting to witness how information from multi-disciplinary rounds provides a treasure trove of data for a trifecta of beneficiaries: clinical operations, quality improvement, and academic research.

Did you encounter any roadblocks that you did not anticipate?

It probably goes without saying but launching a new project in the ICU during a global pandemic was less than ideal. We were lucky to be able to adjust our procedures to accommodate the new normal: masks on rounds and untold numbers of patients with COVID-19. Yet this roadblock ultimately became an opportunity, since the team-focused solutions we are developing have great potential for improving the care of the critically ill during a future pandemic. The solution is two-fold. First, Aviva makes intensivists (physicians who specialize in ICU care) more efficient, allowing them to care for more patients without sacrificing quality. Second, Aviva can augment the knowledge of non-intensivist physicians so that they can care for patients with greater acuity than they normally do.

What are the biggest takeaways from your project?

Relationships are everything. Getting this project off the ground required the coordination and combined efforts of over a dozen different offices and teams across UPMC and the University of Pittsburgh. I thank everyone who is involved and truly believe that Pittsburgh is one of the only places in the world where we could have achieved as much as we have during the last year.

What are your project’s next steps? 

While year one was proof of concept, we now turn our attention to getting our product in front of clinician users and deployed across UPMC. We are excited to show off what Aviva can do, but I must save that announcement for another day. Stay tuned because our next steps may surprise you.