In a recent press conference, Google Cloud’s Aashima Gupta shared results from a new survey conducted by Harris Poll and Google Cloud regarding administrative burdens on U.S. healthcare workers. Gupta discussed in a panel with several subject matter experts how generative AI (gen AI) could help with these challenges.
The survey, Measuring the administrative burden on U.S. healthcare workers—and how generative AI can help, found that clinicians spend almost 28 hours weekly on administrative tasks. Most healthcare providers and payors found that the administrative burden contributes to burnout and staffing shortages.
Eighty percent of providers reported that the burden takes away time spent with patients, and the majority feel this impacts the quality of care. Some respondents expressed concern about the risk of human error in administrative tasks, while many noted they felt optimistic about using Gen AI for these tasks.
Panelist Dr. Tony Farah, M.D., chief medical and clinical transformation officer with the nonprofit Highmark Health, noted that physician burnout had been a crisis for a long time before COVID-19. “Covid just accelerated that concern.” He reiterated that time itself is of concern when taking care of patients. Farah mentioned that the prior authorization process adds to the burden on healthcare professionals. “What we need to do is spend time on how to reduce the denial rates and therefore eliminate all the administrative burden that is placed on both payers and providers.” Farah added that his organization has been able to automate about 30 percent of their authorizations. “We've reduced our cost to process authorizations by 50 percent on the provider side.”
Regarding denials, Farah stated, “We've been able to significantly reduce the time our medical directors spend on doing this work by about 40 percent. This is time reduced, and 90 percent of our letters right now are generated through this process thanks to the help of generative AI.” His organization uses ambient intelligence and virtual care in clinics and in-patient settings.
Michael J. Schlosser, M.D., MBA, FAANS, SVP, with HCA Healthcare, spoke about his organization’s care transformation and innovation strategy launched three years ago. Tackling the administrative burden was a key focus. “The more administrative burden we remove, the higher the quality and the experience of care.” He believes that Gen AI will help with handoff procedures. His organization partnered with the Google research team to develop a critical summary model. Data from nurses gave the team and the model feedback on the quality of the job it was doing in creating these summaries. The nurses rated the model as 90 percent helpful.
Helen Waters, EVP and COO of MEDITECH, discussed the partnership with Google and how AI solutions are incorporated into electronic health records (EHR). “We're actually using an AI-generated model 1.5 with Google under the Vertex AI platform to bring a very strong, summarized view of a patient's record to the forefront, and then using the Gen AI capabilities to correlate conditions and explore those conditions to see what might be most relevant to the next level of clinical decision making.”
During the press conference, Gupta announced the general availability of Vertex AI Search for Healthcare and its Healthcare Data Engine on Google Cloud. “Vertex AI Search for Healthcare helps developers build better assistive technology for healthcare workers to alleviate administrative burdens, and Healthcare Data Engine helps organizations build an interoperable data platform, the foundation of gen AI,” an October 17 press release stated.
