Building the Business Case for an Inclusive Approach to Digital Health Measurement With a Web App (Market Opportunity Calculator): Instrument Development Study

Introduction

Background

Digital technologies are playing an increasing role in health care delivery. Digital health measurement tools represent a key technology class at the center of this movement. These tools use connected sensor technologies to collect physiological measurements and other health information during patients’ daily lives. The resulting measurements offer the potential to change how we define health and treat disease. More frequent measurements outside traditional health care settings can help shift care to more patient-centered models and facilitate earlier diagnosis and better access to care. More diverse, continuous, and high-resolution data can also enable a richer understanding of the impacts of medical products—facilitating faster development of better products. Recently, we have seen rapid growth in the use of digital health measurement tools in both care delivery [1,2] and clinical research and development (R&D) [3].

However, there are well-founded concerns that the benefits of digital health measurement will not be shared equitably [4]. Critically, individuals with the highest unmet need—those who are very sick and those for whom traditional care models have proven inadequate—may find these tools inaccessible or inappropriately designed for their needs [5,6]. Digital health has seen large investments over recent years [7], but there remains a risk that aspirations to fast growth and near-term profits may leave historically underserved groups behind. The existing inequities in health care systems have been well documented, and the COVID-19 pandemic has highlighted many downstream consequences of those inequities [8,9]. Yet, digital tools promise greater access and flexibility, rendering them a promising means to mitigate inequities rather than exacerbate them. As we continue to pursue R&D for new digital health products and measurement tools, health equity must be a central focus.

Against this backdrop, the Digital Medicine Society (DiMe) [10] in collaboration with the Food and Drug Administration’s Center for Devices and Radiological Health [11] has launched the Digital Health Measurement Collaborative Community (DATAcc) to create a “toolkit” of best practices for enabling inclusivity throughout the digital health measurement product development process. As a Center for Devices and Radiological Health strategic priority, collaborative communities convene diverse stakeholders across the medical device ecosystem in precompetitive work to achieve common outcomes, solve shared challenges, and leverage collective opportunities [12].

Introducing the Market Opportunity Calculator

A critical element in fostering inclusive product development is building the business case for why it is worthwhile. While early-stage R&D often happens in nonprofit and academic settings, product development and commercialization are ultimately led by businesses and therefore operate with a focus on commercial benefit. However, the goal of commercial profit is often highly compatible with inclusive product development. Underserved populations, such as racial and ethnic minorities and individuals with disabilities, represent a substantial portion of the US population and therefore a large group of potential beneficiaries of new health tools [13]. In addition, given their higher rates of chronic illness and lower levels of self-reported health, these populations also shoulder a disproportionate share of the US disease burden [14]. In this context, digital health measurement products that are well-suited for these groups will be the only way to reach the total addressable product market. We set out to build a tool to show the value of inclusive digital health measurement technologies. This led to the creation of the DATAcc Market Opportunity Calculator for inclusive development [15].

Methods

Data Sources

The data underlying the DATAcc Market Opportunity Calculator were drawn from publicly available sources. The US Census’s 2017 National Population Projections were used to determine the relative size of population segments across key “inclusion vectors,” such as age groups, race and ethnicity, and levels of educational attainment [13]. Data from the 2019 US Census American Community Survey similarly provided statistics on the size of population segments across household income groups [16]. These data sets were combined with data on disease prevalence and health status from the 2018 Centers for Disease Control and Prevention (CDC) National Health Information Survey [14]. The 2019 CDC Disability and Health Data System provided statistics on the size of disability group segments as well as disease prevalence and health status data for those same groups [17]. Segment sizes were represented as percent shares of the sampled population. To calculate the total market size (in terms of number of potential patients), we multiplied shares by an estimate of the current US population from the US Census Population Clock [18].

Ethical Considerations

Institutional Review Board review was not necessary because our analysis relied solely on publicly available summary statistics from the US Census Bureau and the CDC.

Data Processing

Calculations from the Market Opportunity Calculator are specifically for the US adult population (ie, those older than 18 years, representing roughly 77% of the total US population) due to the US focus of our key data sources. Age group share data were adjusted to reflect the US adult population. Household income data were incorporated using the assumption that “family” households have 2 adults, while nonfamily households have 1. Educational attainment data are for US adults aged 25 years and older. We applied the 25 and over segment sizes to the full adult population under the assumption that they are likely to be representative of the final educational attainment of the full adult population. Finally, race and ethnicity segment sizes were extrapolated based on the entire US population under the assumption that segment shares for adults were roughly consistent.

For each inclusion vector, we defined mutually exclusive, collectively exhaustive segments. For example, we combined race and ethnicity into a single dimension by creating a separate segment for all individuals of Hispanic or Latino ethnicity regardless of race. The remaining non-Hispanic or Latino population was then segmented by race. Additional details on data processing are provided in Multimedia Appendix 1.

Analysis

The goal of the opportunity calculator is not only to report segment sizes and rates of disease prevalence within them but also to quantify the importance of often-underserved patient segments to a given market and, by direct extension, the value of increasing a product’s inclusivity for that group. For example, using statistics on the percent of visually impaired individuals with diabetes and the overall prevalence of diabetes, we can determine the percent of diabetes patients that are visually impaired—a useful statistic in quantifying the importance of expanding the inclusivity of a diabetes digital measurement tool to those who are visually impaired.

Results

DATAcc Market Opportunity Calculator Overview

The DATAcc Market Opportunity Calculator (Figure 1) is a freely available, open-access resource designed for use by those doing R&D in digital health measurement to build a case for incorporating inclusive practices in their product development process. The tool can be accessed from the DATAcc website [15].

‎

Measuring the Importance of Underserved Patient Segments

DATAcc Market Opportunity Calculator users begin by selecting the target health condition their product is intended to address. Current options include arthritis, diabetes, heart disease, high blood pressure, stroke, and fair or poor health status. Next, the user selects an “inclusion vector” along which to segment the population. Current options include age, race and ethnicity, education, household income, and disability. The calculator then displays each segment as a share of the overall US adult population and its share specifically among adults with the target condition (Figure 1A). The calculator output also presents information on which segments are over- or underindexed. For example, while only 3.7% (9.5 million individuals) of the overall US adult population are visually impaired; 8.1% (2.5 million individuals) of those with heart disease are. Thus, among heart disease patients, individuals with vision disabilities are overrepresented by a factor of 2.2.

Quantifying the Benefits of Improving Product Inclusivity

The calculator also enables users to estimate the value of improvements to their products’ inclusivity for specific segments along the chosen inclusion vector. The user first sets a baseline reflecting the user’s understanding of the product’s current inclusivity for each segment (represented as a percentage of the segment the product is accessible and effective for). They then can consider alternative scenarios in a future state where improvements in inclusivity are achieved (Figure 1B). The calculator shows how these improvements manifest in expansions of the product’s accessible market. By setting a unit price or margin for their product, the user can also quantify the economic opportunity associated with the market expansion in dollars (Figure 1C). Consider the example above of a heart disease-focused digital measurement tool where a user explores improving usability for the visually impaired: if the product could serve 20%, rather than 0%, of visually impaired adults with heart disease, that would represent 500,000 more potential patients. With a per unit margin of US $100, such an improvement would expand the addressable market opportunity by US $50 million.

Putting it All Together

After these straightforward steps are completed by users of the calculator, a final readout is provided, which shows the potential value of expanding the inclusivity of the focal product, presents additional growth in market size (Figure 1D), and an associated increase in dollar value (Figure 1E). Users of the market opportunity calculator have the option to generate a pdf of the calculations. No data on user input are stored or collected by the calculator.

What users will observe through the use of the calculator is that across disease areas and inclusion vectors there are substantial economic opportunities from increasing inclusivity, whether those be simplifying prompts on a hypertension-focused product to make it more accessible for adults with lower educational attainment, providing Spanish translations to improve use of a diabetes-focused product for the Hispanic and Latino population, or enabling compatibility of an arthritis-focused product with older phone models and low bandwidth connections to make it more accessible to low-income adults (Table 1).

Discussion

Principal Findings

The Market Opportunity Calculator, along with other components of the DATAcc toolkit for inclusivity, supports the implementation of an inclusive approach to digital health measurement product development [15]. Lack of financial incentives is a major barrier to achieving health equity, so creating a business case for inclusive product design is imperative. The Market Opportunity Calculator demonstrates that inclusive product development can result in expanded market size and financial opportunity, providing the necessary incentives for digital health product developers—and their investors—to adopt an inclusive approach.

Results from the Market Opportunity Calculator can be paired with the Library of Evidence Supporting Inclusive Design, also included in the DATAcc toolkit, which highlights specific case studies providing concrete evidence of the benefits of inclusive design, including broadening product usage, improving outcomes, and reducing risk [15]. Once a decision is made to continue on the path of inclusivity, the digital health measurement product development process, modeled after the Food and Drug Administration’s device development process, outlines the key steps in the product development lifecycle where inclusivity elements can be applied. The Framework for Inclusive Development walks through each step and provides recommendations and resources to apply these inclusivity elements. At present, numerous companies, such as AliveCor [19], Verily [20], and Genentech [21], have begun to leverage the Market Opportunity Calculator and other components of the DATAcc toolkit to articulate the value of inclusivity and better incorporate inclusive design principles into their development processes. Additionally, DiMe recently shared the calculator as a resource for the diversity, equity, and inclusion in Digitized Clinical Trials project [22]. The calculator demonstrates how real-world data can be used to identify additional populations who may benefit from inclusion in clinical trials.

In recent years, several studies in the academic literature and popular press have highlighted instances where lack of inclusive design in medical products have resulted in care inequities: pulse oximeters that are inaccurate for certain skin tones [6] and telehealth apps that are inaccessible to the visually impaired [23]. Identifying these issues is the first step toward remedying them. However, ultimately, the goal should be proactive inclusivity, rather than reactive inclusivity. We will only realize the full promise of the digitization of health care if inclusivity is woven into the initial development of the product itself, rather than something to be fixed after the fact. However, as shown by the DATAcc Framework for Inclusive Development, inclusive design takes time and resources—to consult with or hire accessibility experts, recruit diverse populations for usability testing, and implement the requisite features. The Market Opportunity Calculator shows that the investment in inclusive digital product design is worth it, not just from a moral or equity lens but from a business case perspective as well.

Limitations

In this paper, we describe the development of and data from the first iteration of the calculator, which has a number of limitations. First, the calculator currently includes a limited number of inclusion vectors and health conditions. The limited number of health conditions in particular may constrain the use of the tool for companies developing products for conditions outside the selection set. However, in these instances, developers may still use the general “fair or poor rated health” category as a rough proxy for the target condition and still glean insight on the relative importance of historically underserved patient segments. Second, users currently can only select a single inclusion vector at a time, whereas the selection of multiple vectors may enable more granular population segmentation. Third, the data underlying the model were collected from 2017 to 2019; however, we believe these statistics were unlikely to change dramatically over the recent years. Finally, the calculator only provides data for the US adult population.

DiMe, through work by the DATAcc, will update the calculator as additional input data become available. Future iterations will hope to incorporate more recent data, include additional health conditions and inclusion vectors, and allow for the selection of multiple inclusion vectors simultaneously. The target for the next calculator release is sometime in the first half of 2024.

Conclusions

Digital health measurement has grown rapidly over recent years but is still in its infancy. Now is the time to lay a foundation ensuring that inclusivity is woven into the fabric of the industry. Doing so will set the industry up for long-term success: aligning business incentives with the unmet needs of underserved populations and delivering both sustainable commercial profits and maximizing the societal benefits created by digital measurement tools.