Intro
If Siri or Alexa can be a virtual assistant at home, why not in healthcare? That’s the question asked by nurses from a major U.S. hospital network. DFFRNT found many opportunities for voice technologies to support nurses’ workflow and enhance their workday. Medtech voice solutions have proliferated in the last few years. The tech has powerful capabilities, but there are challenges to implementing voice solutions across multiple healthcare sites or different specialties within one care setting.
The Client
The client is one of the leading non-profit hospital systems in the U.S. It operates more than 2600 care sites, including 139 hospitals. This extensive presence enables clinicians and caregivers to share best practices and insights. An innovation group within the hospital network supports the organization’s commitment to evidence-based clinical improvements. Recently, this innovation group requested the assistance of DFFRNT to research and advise on the areas where the company could leverage voice technology to support nurses.
The Challenge
The innovation group had theoretical questions about how voice interaction could improve the workflow for nurses.
The current state of voice technologies is promising, with artificial intelligence for healthcare and machine learning fueling rapid improvements and complex speech recognition. Voice solutions are already employed in many professional settings to capture information. Voice assistants are moving toward an omnichannel experience, connecting with an increasing number of platforms and products.
Some of the emerging voice technology trends in healthcare, as identified by Gartner, include AI for emotional recognition, ambient digital scribe tools, experiential wayfinding and clinical communication and collaboration.
DFFRNT’s objective was to explore nursing use cases and discover opportunities to apply voice technology successfully. DFFRNT’s principals have decades of experience designing for healthcare settings and recognize that nurses carry a heavy cognitive load in their day-to-day work. Nurses are constantly multitasking and being interrupted. Their active memory is overloaded. They must frequently look up data (What was patient XX’s blood pressure this morning?) and reference material.
The Solution
DFFRNT employed a mixed-methods approach for the research phase of this project. Ethnographic observation was paired with remote and asynchronous interviews and an environmental scan. The ethnographic field studies were conducted at two of the client’s hospital sites in Wisconsin, observing nurses in a medical-surgical ward.
Ethnographic observation immerses the researcher in the user’s environment. This type of study provides an in-depth understanding of user groups, their goals, task flows, use contexts and pain points. It provides factual data on which to base decisions about design and features, and helps create a compelling experience that aligns with their needs and expectations.
Based on the findings from the ethnographic research, interviews allowed DFFRNT to identify potentially unexplored, tangential topics that affect nurses and gather in-depth feedback about motivations, experiences and goals.
This deeper exploration allowed the team to provide more informed design recommendations based on observation and interaction.
The third prong of this research effort was an environmental scan, which included emerging trends for voice technology in healthcare and its use in other business and personal settings.
The Outcome
Voice solutions already exist in many areas applicable to nursing. Most of these have been built as mobile apps or omnichannel products. The more significant players target specific functions, such as communication and patient care. Importantly, DFFRNT’s research found no voice solutions built for nurses as the primary user.
Following the ethnographic field studies and interviews, it was clear that the initial focus for this client should be on solutions for nurses that reduce repetitive tasks and help find information more quickly. Nurses told the researchers the ability to take verbal notes quickly, in an organized way, would also be valuable.
Voice solutions already exist in many areas applicable to nursing. Most of these have been built as mobile apps or omnichannel products. The more significant players target specific functions, such as communication and patient care. Importantly, DFFRNT’s research found no voice solutions built for nurses as the primary user.
Following the ethnographic field studies and interviews, it was clear that the initial focus for this client should be on solutions for nurses that reduce repetitive tasks and help find information more quickly. Nurses told the researchers the ability to take verbal notes quickly, in an organized way, would also be valuable.
Nurses who participated in the research imagine that voice technologies will be able to aid in assessments, ordering medications, updating charts, requesting chart info, contacting members of the care team or family, communicating detailed information about alerts and retrieving data.
Keeping in mind the working conditions of nurses, the key to successfully adopting voice technology solutions is reducing cognitive load. New systems should not require nurses to remember a lot of commands. Also, customization is essential to allow for many different workflows and personal preferences.
It is essential to understand the healthcare context for any voice technology. Many people consider healthcare monolithic, but practices and rules vary from state to state, province to province, hospital to hospital, and between disciplines. It is a challenge to make various systems come together and interact. Drug dispensing systems don’t necessarily exchange info with the EHR system. Nurses and other caregivers are often the translators for these systems.
The Results
DFFRNT developed many recommendations for the adoption of voice technology to support nurses in their work. Eight areas of further investigation were identified.
Among the recommendations in the final report were:
Strengthen communications: Choose solutions that support the nursing culture of interpersonal communication.
More than just charting: Voice solutions have the potential to support nurses beyond basic charting tasks. There is an opportunity to reimagine the documentation paradigm for nursing.
Omnichannel capability: Voice solutions in healthcare can create an omnichannel experience by integrating with and executing tasks across multiple systems. They can act as the “glue” to integrate disparate systems. With this ability to integrate disparate systems, voice technology may be able to mobilize nurses and untether them from their desks and WOWs.
Use voice for output: There is an opportunity for voice technology to provide valuable data about a situation to support decision-making. A voice solution that provides nurses with detailed information about a bedside alarm or a patient’s condition can help nurses make better decisions and save time.
Pain points for nurses fall into three general categories: Charting, time pressures and coordination. If voice-based medical technologies can alleviate some of these challenges without unduly adding to a nurse’s cognitive load, they will be a welcome addition to the healthcare workplace.
Time observations when following a nurse for one hour
- The nurse cumulatively spent 27 minutes charting (5 – 6 pm).
- There were 13 alarms.
- There were four human and ten system interruptions (such as the portable phone).
- There were two phone calls.
- The nurse accessed the Pyxis (automated dispensing system) machine three times.
Excerpt from an ethnographic report for this healthcare project.
Team involved:
Shaun Illingworth
Dominira Saul
Elizabeth Rousseau
Services
Ethnographic research
Innovation advisory
