“Digital first” became a rallying cry in 2020 as the COVID-19 pandemic descended. The concept of virtual care delivery has been around for years – but gained meaningful traction as a viable and valuable alternative when in-person visits became impossible. A second wave of the digital-first mindset is now emerging, specifically around operational workflows that can relieve the pressure on nurses.
Healthcare organizations have long tested strategies to combat burnout and shortages, ranging from increased pay to flexible shift schedules to support groups to stress management. While these approaches have delivered incremental improvements, they have not gone far enough. That’s why more nursing and operations leaders are employing virtual assistants that divert nonclinical, routine tasks to other staff members, freeing nurses to concentrate their time and effort on the clinical care they were educated to provide, often resulting in greater job satisfaction.
Virtual assistants at the bedside
One hospital has equipped inpatient rooms with smart speakers (e.g., Amazon Alexa), for example, supported by a healthcare-specific virtual assistant. Instead of automatically reaching for the call button, patients can speak their request or concern via the device. Software “trained” to recognize conversational speech triages the request, applies artificial intelligence and routes it to the right resource. Nursing staff is immediately alerted to clinical needs – like increased pain, nausea or other symptoms. But other resources such as maintenance or food service field other requests (e.g., television not working or food tray needs to be cleared).
Almost 90 percent of nurses at this hospital reported the virtual assistant helped them respond to patient requests confidently and two-thirds said it improved their ability to prioritize care appropriately. Patients, too, were enthusiastic about the bedside assistant, with 92 percent saying they would recommend other hospitals adopt a smart-speaker option.
Automated post-discharge communication
Another health system in the mid-Atlantic region leverages virtual assistants through an automated post-discharge outreach program. The day after they leave the hospital, patients receive a link to a virtual assistant that poses four basic questions about how they are feeling, if they have filled their prescriptions and made a follow-up appointment, and if they have questions about their care plan.
This virtual assistant is cascaded across communications channels – web, SMS text, and voice – so the patient can interact in a way that is comfortable and familiar for them. Algorithms assess patient responses and escalate the conversation to a triage nurse if needed. In the original pilot program, this health system achieved 86 percent patient engagement, and found that two-thirds of discharged patients were on course for routine recovery. This allowed the nursing staff to focus exclusively on those patients with necessary clinical needs, instead of the entire cohort of discharged patients, saving significant time.
Common misconceptions about virtual assistants
Some healthcare leaders are hesitant to invest more heavily in digital options, particularly in areas traditionally handled by nurses and clinical teams. Their concerns typically fall into three categories: 1) They believe healthcare, by its very nature, requires a human touch and automation interferes with this dynamic. 2) Healthcare uses a specific vocabulary and automated tools aren’t precise enough to understand what is being communicated. 3) Patients don’t like to use virtual assistants and get frustrated when they don’t get the information they need or when they hit a dead end.
Today, virtual assistants are built to avoid these pitfalls and ensure a more elegant and seamless experience.
80% of routine tasks can be managed through a virtual assistant. Many routine healthcare questions and requests can be managed by intelligent virtual assistants – self-scheduling, pre-procedure instructions and navigation instructions, for example. IBM research has shown that 80 percent of routine questions to a call center can be answered by a chatbot. In healthcare, best-in-class virtual assistants have been designed to recognize when an interaction with a patient requires deeper involvement and can escalate the engagement to a nurse or other staff member.
Healthcare-specific virtual assistants can serve as the matchmaker between common medical references and clinical healthcare jargon. Virtual assistants “borrowed” from other industries might not be built with the knowledgebases, artificial intelligence or conversation flows to adequately support patient and provider needs. Increasingly, though, healthcare-specific virtual assistants are focusing exclusively on the needs of healthcare, how consumers speak and think about care delivery, and how to effectively cross-reference lay language into clinical terminology.
When a patient asks for a foot doctor, for example, a good virtual assistant knows he or she seeks a podiatrist. Complex requests such as “I need a migraine doctor who takes Aetna insurance” can be ingested in real time and the response will consider the insurance and even appropriately direct the patient to a primary care physician for a neurology referral.
No more “dumb bots.” Consumers in every industry hate it when automated systems don’t work the way they expect; healthcare is no different. Leaders exploring virtual assistants need to test options rigorously to ensure the functionality genuinely supports anticipated patient (and organizational) needs.
For virtual assistants to truly deliver value, they must be powered by appropriate conversational AI to understand and respond to likely queries. They must leverage machine learning for continuous improvement. Conversational virtual assistants should include cognitive search, a new generation of functionality that uses AI to extract relevant information from multiple, diverse data sets to help find answers to all types of questions. In addition, digital tools need seamless escalation to live agents, so patients can reach a human if desired and not abandon the engagement out of frustration.
Healthcare leaders recognize that the challenges of nurse burnout and shortages are not going away. The “Great Resignation” has not only impacted bedside nursing but also other operational areas like call centers and scheduling that are directly tied to an organization’s ability to acquire/retain patients and generate revenue. Healthcare systems increasingly need to consider a digital-first strategy, leveraging virtual assistants and conversational AI to address today’s complex business challenges.
Patty Riskind is a healthcare tech leader and currently serves as CEO of Orbita. She also held leadership positions with Qualtrics (Head of Global Healthcare) and Press Ganey (Chief Client Experience Officer). She received her BA from Brown University and earned her MBA from the Kellogg School of Management at Northwestern University.