Voice Technology: Can It Help Me Take Better Care of Myself and My Baby?

Currently the market for digital assistants is growing, as are voice technology services and products for use in everyday life.

One digital assistant has more than 50,000 skills available worldwide covering a broad array of topics (e.g., money management and lifestyle tips). These skills can be developed by the public, increasing the capabilities of digital assistants making everyday tasks faster and easier for users.

The past decade has provided a rich foundation and launchpad for mobile and ever-present computing in health. For example, mobile health, or mHealth, has the potential to turn mobile devices into personal labs that continuously assess a person's physiology, behavior, social context and environmental exposure.

Mobile-based applications have been used to provide personal health information and behavioral support on topics such as fitness, nutrition and relaxation. While these applications have traditionally relied on users interacting with graphical interfaces (e.g., scrolling, swiping, clicking), deep learning technologies such as natural language processing, speech recognition and artificial intelligence are once again transforming the digital health landscape, providing opportunities for users to interact with digital health tools through strings of text in natural language.

Given the fairly recent surge in digital assistants, terms are still somewhat ambiguous or conflated. For clarity, we provide the following definitions and differentiations:

• Chatbot: typically a single-turn exchange, wherein the utterance is expressed, recognized and mapped onto a task
• Conversational agent: engages the end user for deeper understanding in an attempt to solve a problem
• Virtual assistant, personal assistant or digital assistant: personalized conversational assistant that retains information about the end-user as a basis for interacting with the end-user

For the purpose of this post, we broadly refer to digital assistants as capable of interacting with end-users via voice-activated commands to personalize information within the context of health information, health maintenance and health promotion.

The Potential of Voice Technology

Initially, the growth and penetration of mobile technologies showed great potential to propel healthcare delivery beyond the clinic and into everyday lives of patients.

Similarly, early versions of digital mental health interventions consistently demonstrated as clinically effective for common mental health disorders such as depression and anxiety. However, real-world implementation often failed, highlighting a research to practice gap. Conducting research with digital health technologies is complex, some of the numerous challenges include reliance on telecommunication infrastructure, commercial service providers, evolving technology and diversity among users.

It’s likely the necessity for hands-free interaction with voice technology will be a growing industry across all customer segments. However, empirical support for the design, development and adoption of digital assistants in the healthcare arena are currently limited with substantial room for improvement.

Looking ahead, it is prudent that health interventions using voice technology also address the significant lessons learned from past digital health interventions, or real-world implementation of digital health interventions will continue to be problematic. With increasing affordability and usability of consumer digital assistants, it is important to redefine the scope and development of digital health solutions, including priorities and perspectives from vulnerable populations most in need of health solutions.

HIMSS TV deep-dive into maternal mortality, access to care and the digital divide.

Using Voice Technology in Underserved Populations

To begin addressing these complex issues, we envisioned a digital assistant that offers stress management skills for use among underserved perinatal women and infants.

Both maternal and infant mortality rates serve as indicators of population health and are associated with accessibility of health services. Globally, more than 800 women die daily due to pregnancy and childbirth-related outcomes. In the United States, infant mortality rates are approximately 71 percent higher than comparable countries, and non-Hispanic Black populations experience higher rates of infant mortality compared to non-Hispanic whites.

Maternal child health remains a public health priority and practices to improve outcomes are urgently needed. Recognizing this, the U.S. Health Resources & Services Administration held a global summit to identify innovative and evidence-based approaches to decrease maternal mortality and severe morbidity rates. Key findings recommend improving:

• Access to patient-centered care
• Quality of care
• Types and distribution of health care providers
• Continuity of care following a life course model
• Data quality and availability
• Quality and consistency of review committees through collaboration and technical assistance

To enhance quality and access to health promotion efforts for underserved perinatal women and their infants who may be subject to adverse effects of social determinants of health and chronic stress, our goal is to design, develop and evaluate a theoretically guided, digital assistant.

Our proposed innovation is a digital assistant for use among underserved perinatal women: Stress Management Intervention Life Essentials (SMILE). SMILE is an interactive, voice technology to support the practice of positive stress management skills and positive health behaviors in new parents.

To leverage upcoming and future technologies, SMILE utilizes a commercially available digital assistant to deliver brief, interactive education and stress management skills to women during the perinatal period. This brief, interactive education and stress management skills content would be delivered to women during the perinatal period covering three primary domains:

1. Perinatal education
2. Health behavior skills
3. Social determinants of health/everyday context

Building upon the initial design, future iterations will utilize artificial intelligence to refine and tailor stress management content and provide parenting skills within the context of everyday life. Beginning in Ohio—with the eighth worst infant mortality rates in the country—this project aims to promote stress management skills and positive health behaviors among perinatal women and infants, vulnerable to chronic stress and social determinants of health, to reduce the risk for infant mortality.

The views and opinions expressed in this blog or by commenters are those of the author and do not necessarily reflect the official policy or position of HIMSS or its affiliates.

Originally published January 17, 2019; updated October 12, 2020