Mobile Health and Technology Usage by Patients in the Diabetes,Nutrition,and Weight Management Clinic at an Urban Academic Medical Center

Abstract

Background:

Low-income, minority, and underserved populations are often excluded from mobile health (mHealth) research. This cross-sectional study sought to define how patients at an urban, academic safety net hospital use technology in their daily lives in an effort to incorporate mHealth into clinical care and research.

Methods:

Patients receiving care in the Diabetes and Weight Management subspecialty clinic at Boston Medical Center were asked to complete a 17-question survey on technology usage. It was modeled on a Pew Research Center survey and available in English, Portuguese, and Spanish.

Results:

Of the 394 survey respondents, 279 (70.8%) completed all questions. Majority of respondents were female (76.4%) and between 30 and 49 years old (42.9%). Respondents self-identified primarily as black/African American (35.8%), white/Caucasian (28.2%), and not Hispanic/Latino (46.4%). Over 90% owned a smartphone and more than 85% accessed the Internet on a mobile device at least once per day. Regarding mHealth usage, 33.5% and 23.1% reported current use of health- and weight loss-centric applications (apps), respectively, while only 19.6% of patients with diabetes used smartphone apps as diabetes self-management tools. Nearly three-quarters (73.3%) reported interest in using apps to manage health. Respondents preferred e-mail (48.7%), phone (39.6%), and in-person communication (36.3%) as research recruitment tools.

Conclusions:

The overwhelming majority of an urban, underserved minority population cared for in a subspecialty clinic have access to mHealth-compatible devices and are either using or interested in using mHealth technology.

Introduction

Mobile technology usage has rapidly expanded over the last decade. In 2018, 77% of Americans owned smartphones compared with just 35% in 2011.¹ However, the rate of adoption of new technology is not equal among all Americans. Only 67% of Americans with mean income less than $30,000 have a smartphone, compared with 93% of Americans with mean income greater than $75,000.¹

In line with the ubiquity of technology, digital health has emerged as a priority for the Food and Drug Administration (FDA). Digital health includes mobile health (mHealth), health information technology, wearable devices, telehealth and telemedicine, and personalized medicine. mHealth includes text messages, smartphone applications (apps), and Web portals.^2,3 Within the digital health landscape, digital therapeutics, telehealth, telemedicine, and personalized medicine implement evidence-based treatment programs directed toward specific diseases.⁴ While the FDA is focusing efforts on digital health in its entirety, mobile medical apps and those for general wellness are two particular areas of interest. As the line between technology and personalized medical care is blurred, safety regulations for these systems are needed.

A 2015 study by Krebs and Duncan reported that 58% of smartphone users in the United States had downloaded a health-related app in the past.⁵ The majority of those apps were within the nutrition and fitness categories, although many other mHealth app domains exist, including prevention/lifestyle, self-diagnosis, provider directories, diagnosis and education, prescription filling, and treatment compliance.⁵ Within diabetes, mHealth and, more specifically, apps are now commonly integrated with more traditional diabetes technologies, such as glucometers and subcutaneous insulin pumps.

One of the major proposed benefits of mHealth is the ability to improve remote access to medical care for individuals who live in rural areas or who otherwise have limited ability to see a clinician.⁶ While much excitement over mHealth has focused on patients who live far from medical providers,⁷ similar limitations to medical care access are observed in inner-city areas, particularly among individuals of underserved populations with lower socioeconomic status.^8
–10 Despite the array of literature on varied mHealth topics, very little directly addresses the use of mHealth tools among minority and low-income populations.

As the largest safety net hospital in New England, Boston Medical Center (BMC) serves a diverse urban patient population. Approximately 57% of BMC patients come from underserved populations, 31% do not speak English as a primary language, and more than 50% have an annual income below $20,400.^11,12 Given the diversity of our patient population and an increasing desire to understand our population's mHealth interest and usage, we conducted a survey among outpatients treated for diabetes or weight management within a subspecialty clinic at BMC to help guide integration of mHealth into both patient care and clinical research.

Methods

Participants

Survey respondents were recruited from the BMC Endocrinology, Diabetes, Nutrition, and Weight Management Clinic in Boston, Massachusetts. Our institutional review board determined this study to be IRB-exempt before survey distribution. The only exclusion criteria were age less than 18 years old and receiving care in the clinic for a condition other than diabetes or weight management. A research information document explaining the purpose of the survey was available to all participants in English, Portuguese, and Spanish.

Sample size

Based on the annual number of patients seen in the clinic for diabetes or weight management (∼5000), a goal of 360 completed surveys was set to obtain a representative sample with 5% error and 95% confidence interval. Because participants were not required to respond to all questions, the clinic population was intentionally oversampled to try to achieve a representative sample. To minimize clinic disruptions, survey distribution was discontinued after 5 months, at which point 394 surveys had been collected.

Survey

The 17-question survey was adapted from the Pew Research Center Mobile Fact Sheet and Teen Relationships Survey of technology and social media use.^1,13 Given the lack of validated mHealth questionnaires, this survey, in conjunction with Pew Research Center's Mobile Fact Sheet,¹ was used to develop the wording for each question. Assuming that a survey for teenagers has higher readability ease than one for adults, we selected this survey as a baseline reference to increase generalizability based on patients' varying education levels. Although created in English, native speakers translated the survey to Portuguese and Spanish to allow for a more representative sample. Patients were asked to complete the survey before or immediately after a routine clinic visit.

Survey questions assessed technology device ownership, Internet and app usage, social media use, communication preferences, and demographics. There were also questions related to diabetes, diabetes technology (glucometers and continuous glucose monitors [CGMs]), and weight management. Most questions were yes/no or multiple-choice, but participants were provided free text space to write in responses, as necessary. No protected health information was collected as part of this survey. Respondents could skip any survey questions.

Patients were randomly sampled in the clinic to complete the surveys either before or following clinic visits through passive survey administration in waiting areas and examination rooms. Research information sheets and surveys were placed in the clinic waiting areas, examination rooms, and rooms used for group classes. The information sheets contained the investigators' contact information for respondents with questions. Clinicians and staff were notified of survey administration, although their participation was minimal to avoid clinic disruptions.

Results

Demographics

Between April and August 2017, 394 patients participated in completing the technology survey. Of the 394, 279 respondents completed all 17 questions. Many participants skipped or incompletely answered the demographic questions, accounting for a large number of incomplete surveys. After excluding demographic questions, 314 participants (nearly 80%) completed all of the remaining questions.

Among the 394 respondents, 76.4% identified as female (Table 1). The majority of respondents were between the ages of 30 and 49 (42.9%) and 50 and 64 (29.2%). Respondents identified their race primarily as black/African American (35.8%), white/Caucasian (28.2%), and other (17.5%), with 31.7% identifying as Hispanic/Latino and 46.4% as not Hispanic/Latino.

Table 1.

Demographics of Survey Respondents (n = 394)

	n	Percentage of total respondents
Gender
Female	301	76.4
Male	70	17.8
Prefer not to answer	5	1.3
Skipped question	18	4.6
Age range, years
18–29	66	16.8
30–49	169	42.9
50–64	115	29.2
65+	21	5.3
Prefer not to answer	5	1.3
Skipped question	18	4.6
Ethnicity
Not Hispanic or Latino	183	46.4
Hispanic or Latino	125	31.7
Prefer not to answer	44	11.2
Skipped question	42	10.7
Race
Black/African American	141	35.8
White	111	28.2
Asian	10	2.5
American Indian/Alaska Native	6	1.5
Native Hawaiian/Pacific Islander	3	0.8
Other	69	17.5
Prefer not to answer	38	9.6
Skipped question	16	4.1

Technology usage

Over 90% of respondents reported owning a smartphone, and 21.3% reported owning a cell phone that is not a smartphone (Table 2). More than half owned a computer (67.5%) or a tablet, iPad, or similar portable device (57.9%). Over 85% of respondents reported accessing the Internet on a mobile device at least once per day. Among those with smartphones or other mobile devices, nearly 80% used apps at least once per day, with 32.2% using four to six apps daily and 30.2% using one to three apps daily. One-third of respondents used health-related apps, with 23.1% using weight-related apps, more specifically. Current versus lifetime use was not distinguished for this question. Apps commonly reported as used in free text responses included MyFitnessPal, Fitbit, Samsung or iPhone Health, LoseIt, and Weight Watchers. Independent of current or lifetime reported usage, nearly three-quarters of participants reported being somewhat or very interested in using mobile apps to help manage their health.

Table 2.

Technology Ownership Among All Survey Respondents (n = 394)

Do you, personally, have any of the following items?	n	Percentage of total respondents
a. A smartphone
Yes	357	90.6
No	37	9.4
b. A cell phone that is not a smartphone
Yes	84	21.3
No	89	73.4
c. A desktop or laptop computer
Yes	266	67.5
No	120	30.5
d. A tablet, iPad, or other similar device
Yes	228	57.9
No	158	40.1
Number of patients who reported owning
All four devices	24	6.1
Three devices	181	45.9
Two devices	116	29.4
One device	64	16.2
No devices	9	2.3

Of the 361 respondents who answered the question about diabetes, 148 (37.6%) reported having diabetes. Of those with diabetes, the most used self-management tool was a glucometer (66.2%) (Table 3). Additionally, 25.7% used the BMC patient portal system (EPIC MyChart), 20.9% used a CGM, and 19.6% used a smartphone app to help manage their diabetes. More than one respondent wrote in that they were using Dexcom, Samsung Health, and iPhone Health, while only single respondents named other diabetes management apps.

Table 3.

Technology Usage Among Respondents with Diabetes (n = 148)

Do you, personally, have and use any of the following items to manage your diabetes?	n	Percentage of patient with diabetes (n = 148)
a. Smartphone app
Yes	29	19.6
No	89	60.1
Skipped question	30	20.3
b. Blood sugar meter (glucometer)
Yes	98	66.2
No	25	16.9
Skipped question	25	16.9
c. CGM
Yes	31	20.9
No	75	50.7
Skipped question	42	28.4
d. BMC (EPIC) MyChart Patient Portal
Yes	38	25.7
No	67	45.3
Skipped question	43	29.0

app, application; BMC, Boston Medical Center; CGM, continuous glucose monitor.

Communication methods

Participants were also surveyed on their use of social media platforms (Table 4). Most respondents reported using Facebook (75.6%) and Google+ (54.3%), followed by Instagram, Snapchat, and Twitter (37.3%, 24.6%, and 15.0%, respectively). Respondents wrote in other social media platforms, with majority of free responses being WhatsApp and Messenger.

Table 4.

Preferences Regarding Communication Methods (n = 394)

	n	Percentage of total respondents
What is your preferred way to communicate with people?^a
Text messages	288	73.1
Phone calls	275	69.8
E-mails	188	47.7
In person	171	43.4
Social media	97	24.6
Other: ______________________	0	0.0
What is your preferred way to communicate with your health care team (doctors, nurses, dietitians, and any other medical professionals)?^b
Phone calls	280	71.1
In person	182	46.2
Text messages	153	38.8
E-mails	155	39.3
Messages through a secure online portal	59	15.0
Other ______________________	0	0.0
What is your preferred way to hear about a research study?^c
E-mail	192	48.7
Phone call	156	39.6
In person (doctor in clinic)	143	36.3
Text message	142	36.0
In person (research coordinator in clinic)	113	28.7
Flyer	59	15.0
Social media	48	12.2
Other ______________________	1	0.3

Respondents were allowed to choose all answers that apply.

The two most common options selected in combination by a single subject for general communication were phone call and text message.

For health care provider communications, phone call and in person options were selected.

For research, text message and e-mail options were selected.

Preferred communication methods were categorized into communication with all people, with their health care team, and regarding research studies. Respondents could select multiple answers for each question. Most participants preferred communicating with all people through text messages (73.1%) and phone calls (69.8%). With their health care team, most respondents preferred phone calls (71.1%), with less than half preferring any other communication method. Participants preferred to hear about research studies by e-mail (48.7%), followed by phone calls (39.1%), in person from a physician (36.3%), and text messages (36.0%). For in-person research recruitment, 7.6% more respondents preferred hearing about a research study from a physician compared with a research team member. Only 15.0% of respondents reported wanting to hear about a research study through flyers, with even fewer (12.2%) preferring social media as a means of research recruitment.

Discussion

While a variety of mobile technologies are available to patients, little is known about adoption of such technologies in a safety net hospital setting. We sought to better characterize our clinic patients and their interest in technology as this population may benefit from expansion of mHealth integration into routine care, but only if patients have access to and are interested in using such technologies.

The rate of smartphone ownership by respondents (over 90%) is much higher than the rates of smartphone usage reported by the Pew Center for individuals with income less than $30,000, even though more than half of BMC patients have an average income lower than this.¹² Although nearly 75% of respondents were interested in using mobile apps for self-management, current self-reported practices do not reflect this interest as only 33% and 20% of all patients and patients with diabetes, respectively, used health-related apps. Many of the apps that patients reported using for diabetes management also overlapped with apps used for weight loss.

Clinically, less than 25% of patients with diabetes were using any technology aside from a glucometer for diabetes self-management. It is also striking that almost 40% of respondents with diabetes were not consistently using a glucometer to support self-management. This is particularly concerning in regard to patient safety and health care costs, as a 2005 review found that nearly 70% of patients in academic medical center endocrinology clinics are using insulin.¹⁴ Patients must know their preprandial blood glucose to administer an appropriate dose of bolus insulin; thus, lack of appropriate glucometer usage indicates an area of improvement for diabetes self-management. Interestingly, 20.9% of patients with diabetes reported the use of CGM for diabetes management similar to the most recent estimate of 22% reported in the type 1 diabetes (T1D) exchange,¹⁵ although our survey included patients with both T1D and type 2 diabetes (T2D) mellitus. Although we did not ask about diabetes type as part of the survey, our clinic's diabetic population consists of 87.2% T2D, 10.2% T1D, and 2.3% atypical diabetes patients. It is interesting that a similar percentage (∼20%) reported app or CGM usage with only six patients using both apps and CGM in conjunction.

Although this survey captured only a small percentage of our clinic patients with diabetes, these data are consistent with prior studies reporting lower rates of technology use among individuals with chronic disease and more specifically diabetes.¹⁶ Given the growing ubiquity of chronic diseases, it is important to remember that self-management tools are changing and patients may not know of or have access to the newest technologies.

The data suggest a wide discrepancy between mHealth interest and use, and the reason for this is likely multifactorial. Possible contributing factors include patient comfort with using technology for health care, concerns about security in using mHealth to manage personal health information, and patient preferences regarding apps. Additionally, it is important to consider provider comfort and knowledge of mHealth resources and their ability to recommend and educate patients on proper use of such technologies. Providers may not be familiar with mHealth resources or have a valid scientific rationale for prescribing technology as a means of enhancing patient self-management.

This study has several clinical and research-related implications. The rapid rate of advancement in this field and limited data available to support recommendation of specific technologies contribute to the lack of adoption by both patients and providers. Even if patients are interested in using technology to manage their health, their ability to integrate mHealth into routine care may be limited without clear guidance from a health care provider. It is essential to educate both patients and providers on mHealth self-management resources, such as smartphone apps and CGM, to ensure the best possible care. It is also critical that patients and providers share mHealth goals to be successful. For instance, while providers may prefer patient portal systems to communicate laboratory results, patients may prefer secure e-mail, and this disconnect may impede successful integration of mHealth into care.

When considering research recruitment strategies, alternative communication methods should be used to reach wider populations. Research recruitment historically relied on the use of flyers, but respondents in our survey preferred e-mails, phone calls, texts, or in-person communication. While participants favored hearing about research from a provider over a research coordinator, the survey did not specify whether the physician sharing the research in person was the patient's own provider. Thus, it is unclear whether there would be a difference if the recruiting physician were not engaged in the patient's routine medical care. At the same time, although social media is widely used for communication in daily life, respondents were least interested in learning about studies on social media, so there is still some lag in technology usage and preferred communication methods for health care information and medical research.

The data also have an implication on clinical procedures as clinic staff should verify contact information beyond mailing address during appointment check-in. By obtaining patient e-mail addresses, research centers can implement opt-out policies for research recruitment, where all patients will hear about studies unless they choose to unsubscribe from such e-mail notifications. In addition to recruitment notifications, research centers should utilize e-mail to increase transparency with patients by communicating research findings as they become publicly available. This communication could enhance trust between patients and the research team, which is often a barrier to research participation among minority populations.

This study also had several limitations, namely the single-site setting, lack of a validated questionnaire, and use of self-reported data without researcher probing or verification of nonduplicate respondents. In addition, the goal for completed surveys to achieve a representative sample was not achieved, although the distribution was intentionally discontinued after oversampling to avoid clinical disruptions. Despite the diverse sample, significantly more women completed the survey than men; thus, it is more difficult to generalize study results to male patients, although the survey population is reflective of our clinic's demographics. Because surveys are subject to accessibility and bias issues, native speakers translated the original English surveys into Spanish and Portuguese. Despite this attempt, most returned questionnaires (83%) were in English, with the remaining 10.4% being in Spanish and 6.6% being in Portuguese. However, internal clinic data demonstrate that among patients speaking the three languages that the survey was available in, 79% speak English, 19% Spanish, and 2% Portuguese. Thus, even though most of the surveys returned were in English, this closely parallels languages spoken by patients in this clinic, although it still leaves the possibility that the results do not apply as broadly to non-English-speaking patients.

Nonetheless, knowing that patients in a diverse safety net population have access to and are interested in using technology for their health, further studies must be undertaken to assess both adoption and benefits of increasing mHealth access in this population.

Conclusions

Although demographically diverse, most patients at our urban safety net hospital have access to and utilize technology frequently. In contrast with current mHealth research, which frequently excludes lower socioeconomic status and non-English-speaking populations, the patients in this subspecialty clinic have access to and are interested in mHealth interventions and modern communication methods, such as e-mail and text messaging for health care. Use of mHealth resources to increase self-monitoring of chronic diseases, such as diabetes, may improve short- and long-term health outcomes for patients from a variety of backgrounds, and further studies should assess mHealth usage in more diverse populations.

Footnotes

Author Disclosure Statement

The author(s) declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Pew Research Center: Mobile Fact Sheet. www.pewinternet.org/fact-sheet/mobile Published 2018. (accessed August 31, 2018 ).

Modzelewski

, Stockman

, Steenkamp

: Rethinking the endpoints of mHealth intervention research in diabetes care. J Diabetes Sci Technol, 2018; 12:389–392.

Food and Drug Administration: Digital Health. www.fda.gov/medicaldevices/digitalhealth Published 2018. (accessed January 30, 2019 ).

Natanson

: Digital Therapeutics: The Future of Health Care Will Be App-Based. www.forbes.com/sites/eladnatanson/2017/07/24/digital-therapeutics-the-future-of-health-care-will-be-app-based/#2e76e1da7637 Published 2017. (accessed January 30, 2019 ).

Krebs

, Duncan

: Health app use among US mobile phone owners: a national survey. JMIR Mhealth Uhealth, 2015; 3:e101.

Colorafi

: Connected health: a review of the literature. Mhealth, 2016; 2:13.

Mallow

, Theeke

, Barnes

, et al.: Using mHealth tools to improve rural diabetes care guided by the chronic care model. Online J Rural Nurs Heal Care, 2014; 14:43–65.

Robertson

, Cousineau

: Health status and access to health services among the urban homeless. Am J Public Health, 1986; 76:561–563.

Hussein

, Diez Roux

, Field

: Neighborhood socioeconomic status and primary health care: usual points of access and temporal trends in a major US urban area. J Urban Health, 2016; 93:1027–1045.

10.

Ahmed

, Lemkau

, Nealeigh

, Mann

: Barriers to healthcare access in a non-elderly urban poor American population. Health Soc Care Community, 2001; 9:445–453.

11.

Boston Medical Center: BMC Facts. 2017. http://internal.bmc.org/about/documents/INT_79_BMC-FACTS_2017.pdf (accessed July 31, 2018 )

12.

Boston Medical Center: 2018. interview day diversity and inclusion presentation. Boston Medical Center employees. 2018.

13.

Pew Research Center: Teens, social media and technology overview. www.pewinternet.org/2015/04/09/teens-social-media-technology-2015/pi_2015-04-09_teensandtech_06 Published 2015. (accessed July 31, 2018 ).

14.

Grant

, Buse

, Meigs

: Quality of diabetes care in U.S. academic medical centers: low rates of medical regimen change. Diabetes Care, 2005; 28:337–442.

15.

DeSalvo

, Miller

, Hermann

, et al.: Continuous glucose monitoring (CGM) and glycemic control among youth with type 1 diabetes (T1D): international comparison from the T1D Exchange and DPV Initiative. Pediatr Diabetes, 2018; 19:1271–1275.

16.

Fox

, Purcell

: Chronic Disease and the Internet. Washington DC: 2010. Pew Research Center. http://pewinternet.org/Reports/2010/Chronic-Disease.aspx. (accessed July 31, 2018 ).