This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Formative Research, is properly cited. The complete bibliographic information, a link to the original publication on http://formative.jmir.org, as well as this copyright and license information must be included.
Patient-reported outcomes (PROs) are increasingly being used in the management of type 2 diabetes (T2D) to integrate data from patients’ perspective into clinical care. To date, the majority of PRO tools have lacked patient and provider involvement in their development, thus failing to meet the unique needs of end users, and lack the technical infrastructure to be integrated into the clinic workflow.
This study aims to apply a systematic, user-centered design approach to develop i-Matter (investigating a mobile health [mHealth] texting tool for embedding patient-reported data into diabetes management), a theory-driven, mobile PRO system for patients with T2D and their primary care providers.
i-Matter combines text messaging with dynamic data visualizations that can be integrated into electronic health records (EHRs) and personalized patient reports. To build i-Matter, we conducted semistructured group and individual interviews with patients with T2D and providers, a design thinking workshop to refine initial ideas and design the prototype, and user testing sessions of prototypes using a rapid-cycle design (ie, design-test-modify-retest).
Using an iterative user-centered process resulted in the identification of 6 PRO messages that were relevant to patients and providers: medication adherence, dietary behaviors, physical activity, sleep quality, quality of life, and healthy living goals. In user testing, patients recommended improvements to the wording and timing of the PRO text messages to increase clarity and response rates. Patients also recommended including motivational text messages to help sustain engagement with the program. The personalized report was regarded as a key tool for diabetes self-management by patients and providers because it aided in the identification of longitudinal patterns in the PRO data, which increased patient awareness of their need to adopt healthier behaviors. Patients recommended adding individualized tips to the journal on how they can improve their behaviors. Providers preferred having a separate tab built into the EHR that included the personalized report and highlighted key trends in patients’ PRO data over the past 3 months.
PRO tools that capture patients’ well-being and the behavioral aspects of T2D management are important to patients and providers. A clinical trial will test the efficacy of i-Matter in 282 patients with uncontrolled T2D.
ClinicalTrials.gov NCT03652389; https://clinicaltrials.gov/ct2/show/NCT03652389
Uncontrolled type 2 diabetes (T2D) is a significant public health problem in the United States, particularly among vulnerable populations (eg, low-income and racial and ethnic minorities) [
Recognizing the central role patients play in the management of T2D (eg, being aware of its signs and symptoms and engaging in daily self-care behaviors), several national and local organizations have forged initiatives to support the development and use of patient-reported outcomes (PROs) in the evaluation of health and well-being of patients with T2D [
Much of the existing research that incorporates PROs in T2D has been limited to clinical drug trials examining patient tolerance to new treatments [
Systematic reviews of mHealth interventions in patients with T2D have demonstrated positive, short-term benefits on HbA1c levels and self-care behaviors [
The i-Matter (investigating an mHealth texting tool for embedding patient-reported data into diabetes management) trial aims to address this gap in the literature by evaluating the efficacy of an innovative mobile PRO system that incorporates patients’ perspective of their disease into the management of T2D in primary care practices. The i-Matter intervention uses text messaging to capture patients’ self-reported PROs in real time, enhances patient engagement through data-driven feedback and motivational messages, and creates dynamic visualizations of the PROs that can be shared in personalized reports and integrated into the clinical workflow. A future randomized controlled trial (RCT) will evaluate the efficacy of the i-Matter intervention versus usual care on reduction in HbA1c and adherence to self-care behaviors at 12 months among 282 patients with uncontrolled T2D who receive care in resource-limited primary care practices. This paper discusses the iterative process of developing, integrating, and user testing the i-Matter intervention in the formative phase of the trial.
The i-Matter intervention is a blend of 2 frameworks: technology acceptance model (TAM) and capability, opportunity, and motivation model of behavior (COM-B). The TAM is based on the theory of reasoned action and asserts that perceptions of usefulness and ease of use directly influence the intention to use a new technology, leading in turn to its adoption [
Application of capability, opportunity, and motivation model of behavior theoretical constructs to i-Matter design features.
i-Mattera components | COM-Bb constructs | Mechanisms of action | Design features |
PROc assessments |
Capability (comprehension) Motivation (habit formation) |
Rating PROs on a scale helps patients make more realistic assessments of their symptoms and behaviors Daily ratings increase patients’ awareness of their condition on their quality of life and daily functioning Tracking PROs and observing patterns provides patients with reasons to adhere to their self-management regimen |
Daily text message questions Asks patients to complete small doable actions at optimal times |
Feedback messages (insights) |
Motivation (perceptions of illness and emotional response) |
Enables patients to identify changes in PROs that previously went undetected Encourages self-reflection of PRO ratings and their impact on behavior |
Data-driven insights based on PRO ratings, such as: Correlational: association between PRO responses Individual: comparisons of PRO responses across weeks |
Motivational messages |
Motivation Opportunity (perceived support) |
Uses text messages to maintain high levels of engagement in the program |
Text messages that encourage journaling, such as: Response-based: weekly supportive messages based on PRO responses Activity-based: weekly messages based on response rates to the messages Completion-based: messages based on patient duration in the study |
Personalized reports |
Opportunity (patient-provider relationship) Capability (comprehension and ability to plan) |
Facilitates informed discussions with provider Provides provider with succinct and timely data on patient PROs Motivates patients through the gradual completion of the personalized report, with landscape changes every 4 weeks Enables patients to understand and identify patterns in their PROs and to develop behavioral changes to better manage PROs |
Patient reports will include pattern messaging, PRO data visualizations, reflective questions, and tips plus an area for notes Monthly PRO patterns integrated into EHRd, available during and between visits |
ai-Matter: investigating an mHealth texting tool for embedding patient-reported data into diabetes management.
bCOM-B: capability, opportunity, and motivation model of behavior.
cPRO: patient-reported outcome.
dEHR: electronic health record.
We used the evidence-based user-centered design (UCD) approach to conduct the formative phase of the trial [
This study was conducted in a network of primary care practices of New York University Langone Health (NYULH). The practices comprised >1500 ambulatory physicians, nurse practitioners, and physician assistants who care for >800,000 patients in 235 facilities in New York City’s 5 boroughs: Long Island, New Jersey, Westchester County, Putnam County, and Dutchess County. The participating sites include academic practices, many community-based practices, and federally qualified health centers, serving an ethnically diverse population. All primary care practice sites share a single, integrated electronic health record (EHR; Epic).
The target enrollment for the formative phase was 36 patients and 14 providers. To be eligible, patients must (1) have had a diagnosis of T2D for ≥6 months; (2) have had uncontrolled T2D, defined as HbA1c >7%, documented in the EHR at least twice in the past year; (3) be fluent in English or Spanish; (4) be willing to send and receive text messages; and (5) be aged ≥18 years. Patients were excluded if they (1) refused or were unable to provide informed consent; (2) had acute renal failure, end-stage renal disease (ESRD) or evidence of dialysis, renal transplantation, or other ESRD-related services documented in the EHR; (3) were participating in another T2D study; (4) had significant psychiatric comorbidity or reports of substance abuse (as documented in the EHR); (5) were pregnant or planning to become pregnant within 12 months; or (6) planned to discontinue care at the practice within the next 12 months. Providers were eligible if they (1) were a primary care provider (ie, medical doctor, nurse practitioner) practicing at the participating practices and (2) provided care to at least five patients with T2D. The NYULH Institutional Review Board approved this study.
We used 2 approaches to recruit patients and providers into the formative phase. First, potentially eligible patients were identified through a review of the EHR using the diagnosis-related group codes indicating the presence of T2D and receiving care from a primary care provider at one of NYULH practices. After retrieving a list of potentially eligible patients, research assistants (RAs) reviewed patients’ EHR to determine if the patient met the eligibility criteria. Patients that met these criteria were contacted via telephone to confirm eligibility. During the telephone call, the RA gave patients a description of the study, including their role as participants in the study. If the patient remained interested in participating, they were given the option to either complete the focus groups or interviews in-person in a private room or via a remote session using the secure Webex conferencing platform. Providers were sent emails from the study principal investigators inviting them to provide feedback on the development of an interactive mHealth tool that could help enable patients with T2D to take a more active role in their diabetes management. All patients and providers provided written informed consent before participation in the study.
Evidence-based user-centered design process for the development of patient-reported outcome text messages.
Steps | Methods | Outputs |
1. Adapt | Patient focus groups and provider interviews |
Thematic analysis of patient and provider needs, preferences, and barriers and facilitators of tracking PROsa Review of existing validated PRO questionnaires by study team based on thematic analysis Initial list of PROs for i-Matterb comprised individual items extracted from existing questionnaires Reduced list of PROs based on importance rankings from focus group participants |
2. Integrate | Design workshop Workflow mapping Problem or opportunity analysis Presentation of PRO list from step 1 EHRc integration |
Refined list of PROs Clinic workflow or patient journey maps Essential features of i-Matter system i-Matterb prototype: PRO text messages and personalized report |
3. Evaluate |
2 rounds of patient user testing sessions Provider interviews |
Finalized PROs and personalized report Fully functional i-Matter intervention |
aPRO: patient-reported outcome.
bi-Matter: investigating an mHealth texting tool for embedding patient-reported data into diabetes management.
cEHR: electronic health record.
The goal of the focus groups was to select the PROs that would be integrated into the i-Matter intervention as it relates to patients’ experiences living with T2D. A trained moderator used a semistructured guide to explore (1) patients’ daily experiences living with T2D, (2) the barriers and facilitators to achieving their diabetes-specific goals, (3) descriptions of patient-provider conversations about T2D and goals for HbA1c, and (4) interest in sharing PRO data with their provider to support treatment of T2D. A trained bilingual moderator also conducted separate focus groups with Spanish-speaking patients to inform the cultural and linguistic adaptation of i-Matter. Before starting each focus group, all patients completed questions about their comfort with using technology and their interest in using mHealth tools for diabetes care.
A trained moderator conducted semistructured individual interviews with primary care providers at the participating practices. The goal of the interviews was to elicit provider feedback on the clinical relevance of the PROs discussed in the patient focus groups for the management of T2D. The interview guide also explored (1) providers’ level of comfort with PRO data, (2) descriptions of patient-provider discussions about diabetes management, and (3) other important PROs not identified in the patient groups.
Results from the thematic analysis of the focus groups and interviews were used to develop a preliminary list of PROs for inclusion in i-Matter [
The study team then recontacted patients from the focus groups to get their feedback on the candidate list and have them rank the perceived importance of each PRO for management of T2D on a 1 (least important) to 6 (most important) scale. The study team used patients’ ratings in concert with the thematic analysis to narrow the list of PROs that would be presented to participants in the design workshop.
The design workshop comprised patients, providers, academic researchers with expertise in T2D and PROs; the digital health company Rip Road; and staff from the NYULH Medical Center Information Technology (MCIT) department. The design workshop used a UCD protocol adapted from the Agency for Healthcare Research and Quality [
Following steps 1 and 2, the study team collaborated with Rip Road to develop a prototype of i-Matter (ie, the beta version of the text message program and personalized report).
We wrote 2 variations of each PRO question to evaluate the wording and response formats that would yield the highest patient response rates and data quality. On the basis of our previous experiences and best practices for data collection via text message [
In addition to prototype development, we created decision rules that would drive the delivery of the text messages and report content. The rules, which were iteratively refined throughout the formative phase, outline the timing and order of the messages, the duration of time patients had to respond to each message (ie, response window), and the conditions that triggered specific motivational text messages and individualized insights displayed on the personalized report (
i-Matter study flow.
User testing was conducted in a purposive sample of patients drawn from the focus groups and those who were naive to the tool (ie, did not participate in previous steps). A rapid-cycle design (ie, design-test-modify-retest in short intervals of time) was used to allow for iterative refinement of the i-Matter prototype between each user test. Patients participated in the user testing sessions for 2 weeks, during which time they received and responded to the PRO questions sent via text message. At the end of the 2-week period, the study team sent patients a copy of the personalized report (
We also conducted interviews with providers to elicit their feedback on preferences for visual displays and placement of the report in the EHR and perceived barriers and facilitators to viewing the reports in clinical practice. The primary outcome of this step was the fully functional i-Matter intervention for testing in the RCT.
Example of a final personalized report after two rounds of user testing.
Patient use of mobile technology: before the focus groups, patients completed a survey created for this study that assessed the frequency of mobile phone use, capabilities of their mobile phones (eg, Wi-Fi connection, Bluetooth, and mobile data plan), the most commonly used functions (eg, text messaging, phone calls, email, and apps), comfort with using their mobile phone to manage T2D, interest in enrolling in a text messaging diabetes program, and challenges to using their mobile phone for diabetes self-management.
Sample size estimates for the formative phase were based on best practices for maximizing the information power of qualitative research, which recommends beginning with 6 to 8 participants per qualitative method and adding to the sample, as needed [
Focus groups and interviews were audiotaped, translated where necessary, and transcribed verbatim. Both data sources were analyzed using the constant comparative method, in which text was categorized into themes with the use of codes developed iteratively to reflect the data [
After each round of user testing, the study team employed the best practices for instant data analysis of usability data for each PRO [
Following the analysis of use data, the research team categorized each issue with usability as either critical (abandon or remove), severe (significant delay or frustration in task completion requiring revision), or cosmetic (minor issue). Each of these issues were mapped onto the interview transcripts and survey responses to provide specific and detailed recommendations for refining i-Matter before proceeding to the next testing session.
We invited 55 patients with T2D (22 male and 33 female) to participate in the focus groups, of which 35 (64%) declined participation, leaving 20 potential participants. Reasons for declining participation included being too busy, limitations owing to other comorbid conditions, personal or family constraints, and lack of interest in participating in the research. Of the 20 people who agreed to participate, 12 (60%) attended one of the focus groups, 1 did not attend owing to a scheduling conflict with work, and 7 stopped responding to the RA’s outreach calls. We held 4 focus groups: 2 for English-speaking patients (n=6) and 2 for Spanish-speaking patients (n=6).
Analysis of the focus groups identified 4 core themes: (1) patients felt as though their lives were controlled by their blood sugar values; (2) patients’ greatest fear of having T2D were vision loss, kidney failure, or risk of amputation, and avoiding these consequences served as motivators for behavior change; (3) important goals for patients were being in control of their T2D, feeling well, living a long healthy life, and eventually not needing medications for T2D (owing to concerns about the negative long-term effects); and (4) forgetfulness, poor dietary adherence, physical inactivity, tiredness or fatigue, and poor emotional health were viewed as major barriers to keeping blood sugar in control. Patients in the Spanish-speaking focus groups also spoke about God being an important source of strength and motivation to improve their health.
We conducted 6 provider interviews (3/6, 50% female; 4 primary care providers, 1 endocrinologist, and 1 general surgeon and weight management specialist). Analysis of the interviews identified the central theme that providers want PRO data that are specific and actionable and can help them focus the clinic visit on what is most important for their T2D patients’ care. All providers felt that an asset of a program like i-Matter would be having patients systematically track data such as dietary intake and medication adherence that they cannot reliably assess within the time constraints of a clinic visit. All providers liked the idea of showing correlations between PROs being tracked in i-Matter and clinical data that are already stored in the EHR, such as HbA1c values. Providers varied on the importance of tracking patient functional status, quality of life, and psychosocial health, with two-thirds of the providers commenting that it was central to understanding patients’ behaviors, whereas the remaining one-third felt they were
Next, the study team selected individual items from existing PRO measures that best represented themes derived from the focus groups and interviews. This resulted in the selection of items representing 8 categories of PROs: diabetes-related symptoms, quality of life, emotional health (eg, depression, mood, and distress), treatment-related symptoms, treatment satisfaction, diabetes-related functional status, medication adherence, and lifestyle behaviors. Patient ranking of the items further reduced the number of PRO categories to 5: diabetes-related symptoms, quality of life, emotional health, medication adherence, and lifestyle behaviors (
Sociodemographic characteristics and comfort with technology survey responses among focus group participants (n=12).
Sociodemographic characteristics | Values | ||
Age (years), mean (SD) | 62.5 (5.6) | ||
HbA1ca, mean (SD) | 7.95 (0.8) | ||
Female, n (%) | 8 (67) | ||
Employed, n (%) | 4 (33) | ||
Retired, n (%) | 4 (33) | ||
Annual income <US $25,000, n (%) | 7 (58) | ||
Hispanic, n (%) | 7 (58) | ||
|
|||
|
White | 5 (42) | |
|
Black | 3 (25) | |
|
Asian | 1 (8) | |
|
Other | 4 (25) | |
|
|||
|
Less than high school | 1 (8) | |
|
High school degree | 4 (3) | |
|
Some college | 2 (17) | |
|
College degree | 5 (42) | |
|
|||
|
Currently uses text messaging | 7 (58) | |
|
Has an unlimited text messaging plan | 12 (100) | |
|
Always has mobile phone with them | 9 (75) | |
|
Comfortable downloading apps on their mobile phone | 7 (58) | |
|
Comfortable receiving and responding to text messages about T2Db | 8 (67) | |
|
Interested in using mobile phone to help keep track of T2D | 7 (58) | |
|
|||
|
Cost of receiving messages | 2 (17) | |
|
Unreliable internet access | 1 (8) | |
|
Do not use mobile phone regularly | 3 (25) | |
|
Unsure of benefit | 4 (33) | |
|
Concerns about privacy and security | 2 (17) |
aHbA1c: hemoglobin A1c.
bT2D: type 2 diabetes.
Patient ranking of perceived importance of initial list of candidate patient-reported outcomes.
PROa categories | PRO statements | Mean importance score, range 1 (low) to 6 (high) |
Symptom | Over the past week, did you experience tingling or prickling sensations in hands or feet owing to your diabetes? | 1.8 |
Symptom | Over the past week, did you experience dry mouth owing to your diabetes? | 3.0 |
Symptom | Over the past week, how often were you bothered by blurred vision? | 3.6 |
Symptom | Over the past week, how would you rate your level of fatigue owing to your diabetes? | 4.3 |
Symptom | Over the past week, how often did you experience increased thirst and frequent urination owing to your diabetes? | 4.1 |
Emotional health | Over the past week, how often were you been bothered by emotional problems such as feeling anxious, depressed, or irritable owing to your diabetes? | 4.2 |
Emotional health | How often over the past month, did you feel overwhelmed by the demands of living with diabetes? | 2.75 |
Lifestyle behavior | On average, how many days did you participate in at least 30 min of physical activity over the past 7 days? | 3.13b |
Lifestyle behavior | Over the past week, how often did you eat (favorite unhealthy food)? | 4.6 |
Lifestyle behavior | Over the past week, how often did you eat (favorite healthy food)? | 2.25 |
Lifestyle behavior | How would you rate your sleep quality over the past 7 days? | 4.8 |
Quality of life | I was able to keep my blood sugar in good control today. | 4.6 |
Medication adherence | Over the past week, how often were you able to take your diabetes medication on time? | 4.9 |
Medication adherence | How many days did you miss or skip at least one dose of your diabetes medication over the past 7 days? | 2.9 |
aPRO: patient-reported outcome.
bDespite the lower score, physical activity was added as a PRO after review of transcripts and notes from patient and provider interviews.
A total of 17 stakeholders participated in the design workshop. The following themes emerged when the group discussed the candidate list of PROs: (1) PROs should show variability in patients’ responses over time and be actionable by both patients and providers, (2) PROs should be taken from validated questionnaires to increase provider confidence in the data patients report and be comparable with HbA1c levels, (3) choosing fewer PROs would help increase patient response rates and reduce the burden on providers to view the data, (4) tracking PROs that focus on adherence to lifestyle behaviors were most appealing to patients, and (5) PRO content should be general (eg, “how are you feeling today?”) as opposed to diabetes-specific (“how much does diabetes interfere with your life?”). The group reasoned that questions that were too specific may not be relevant to all patients and could lead to disengagement or missing data. Alternatively, a broader question could be used as a way to show care for patients’ overall well-being and as an entry point for more diabetes-specific questions that may uncover new or different concerns the provider should be aware of.
On the basis of these discussions, the group generated several ideas for potential visualizations of PRO data. These included defining a threshold that patients’ data can fall above or below and depicting it in a way that makes it easily detectible and actionable, using bar graphs to show directionality, including icons or coloring schemes in addition to PRO labels that enhance the readability of the report, and including summary data in percentages or raw numbers to show the patient’s progress over time.
Applying the findings from steps 1 and 2, the study team reduced the number of PRO categories to 4. Diabetes-related fatigue (symptom category) was removed from the list because providers viewed it as too nebulous and not actionable, whereas patients felt sleep quality was a more meaningful PRO for their diabetes management. In addition, physical activity was added to the lifestyle category because many patients felt that physical inactivity was a major contributor to weight gain and poor diabetes control.
We completed 2 rounds of user testing with patients: 7 patients completed the first round of testing (1 Spanish-speaking), and 3 patients completed the second round.
In the second round of user testing, the message protocol was modified to address the suboptimal percentage of missed responses. For example, to address the wide range of response times seen in the first round of testing (range 0-661.6 min), we restricted the patients’ ability to respond to the morning and evening PRO questions to a 1-hour window (based on the median response time). Overall, the i-Matter platform sent 222 messages and received 188 responses (84.6%) from patients. The most frequently missed message was sleep quality (77/188, 40.9% of missed messages). The most common reason for an invalid message was the patient responding to a question outside the 1-hour response window.
In qualitative interviews, patients in both rounds of user testing described the program as easy to use, not intrusive to their daily life, and helpful for managing their T2D. Similar findings were seen in the TAM3 survey responses (
Patient text messaging use behavior during user testing.
User behaviors | User testing round 1 (n=7) | User testing round 2 (n=3) |
Time-on-task | 44 min (range 0-661.6) | 20 min (range 0.08-30) |
Task success (messages), n (%) | 232 (90.6) | 175 (93.1) |
Missed responses, n (%) | 100 (39.2) | 28 (15.0) |
Late responses, n (%) | 49 (19.3) | 14 (7.5) |
Invalid responses, n (%) | 24 (9.4) | 13 (6.9) |
Response to technology acceptance model version 3 survey questions.
Questions | Proportion of patients agreeing with statement, n (%) | |
|
||
|
I would definitely use the i-Matter program in the future | 5 (71) |
|
The PRO questions are very helpful for managing T2Db | 6 (86) |
|
Receiving and responding to PRO questions was easy | 7 (100) |
|
I responded to the PRO questions all the time | 5 (71) |
|
I would recommend i-Matter to friends and family | 7 (100) |
|
My provider would be more effective managing T2D with my PRO data | 5 (71) |
|
Overall, the i-Matter program is great or excellent | 6 (86) |
|
||
|
I would definitely use the personalized report in the future | 8 (89) |
|
The personalized report is very helpful for managing T2D | 7 (78) |
|
The personalized report is easy to use | 5 (56) |
|
I would share the personalized report with friends or family | 5 (56) |
|
Showing my provider the personalized journal would help make clinic visits more effective | 7 (78) |
|
The charts and images are great | 6 (67) |
|
Overall, the personalized report is great | 6 (67) |
aPRO: patient-reported outcome.
bT2D: type 2 diabetes.
As shown in
Recommended changes to patient-reported outcome text messages from user testing sessions.
PROa categories and original messages | Original timing | Revised message | Revised timing | |
|
||||
|
Have you taken all of your diabetes medications as prescribed today? | Daily at 7 AM | Retain as is | Allow patients to decide if they want the message in the AM or PM, or both (11 AM and 9 PM) |
|
||||
|
Reply with 1-4 to track ONE healthy living goal: 1=Lose weight 2=Eat more fruits and vegetables 3=Eat less sweets and carbohydrates 4=Have better portion control |
Daily at 7 AM | Retain for all patients. Separate less carbs and sweets to 2 separate goals | Changes so patients choose healthy goal at baseline visit (with option to change goal every 3 months) |
|
How successful were you in achieving your goal to (custom text healthy goal) yesterday? Response: 0 (not at all) to 10 (very successful) | Daily at 2 PM | How successful were you in achieving your goal to (custom text healthy goal) this |
Change timing to weekly at 2 PM |
|
In general, how healthy your overall diet was today? | Daily at 7 AM | Retain message but change timing to assessing overall diet |
Change to daily at 10 AM |
|
Rate your sleep quality last night. Think how easily you fell asleep, how often you woke up and if sleep was refreshing. Response: 0 (poor) to 10 (excellent) | Daily at 7 AM | Reply with the number that best describes how well you slept last night | Change to daily at 9 AM |
|
How many days in the past week did you do any physical activities like brisk walking where you breathed harder than normal? | Weekly at noon | Other than your regular job, did you do any physical activities like brisk walking for at least 30 min today? | Change to daily at 8 PM |
|
||||
|
Reply with the number that best describes how much control you felt you had over your diabetes over the past 2 weeks | Biweekly at noon | Retain as is change timing to weekly | Change to weekly at 3 PM |
|
||||
|
Reply with the number that best describes how irritable or moody you felt today owing to your diabetes | Daily at 7 PM | Remove | N/Ac |
aPRO: patient-reported outcome.
bText in italics show the changes made to the PRO timing across user testing sessions.
cN/A: not applicable.
A total of 9 patients provided feedback on the 1-month and 3-month versions of the personalized report: 4 of these patients participated in the user testing (of which 2 were recruited from the focus groups), whereas 5 were naive to the program. Overall, the majority of patients (8/9, 89%) felt the report was easy to read, eye-catching, and comprehensive. There was a strong preference for the 1-month version of the report owing to the larger font size. Patients also felt that receiving the report more frequently would help sustain motivation. Patients preferred layouts that used darker fonts and lighter background colors to help make the text easier to read. All patients viewed the color-coded schema favorably because it helped draw attention to the most important aspects of the report and made the data easy to interpret.
Several patients had difficulty reading the bar graphs of PROs that were collected biweekly (eg, quality of life) and recommended changing the items to weekly measures to be consistent with other PROs. Finally, email was the preferred delivery method, and most patients would share the report with their family and friends (
Benefits of using the personalized report for diabetes self-management included being able to analyze how well one is adhering to recommended diabetes behaviors (“being honest with yourself”)
Overall, all (n=6) providers thought the report was a good tool to help patients manage their T2D. Similar to patients, they felt that the insight messages were helpful for interpreting the data and prompting behavioral changes. When reviewing the PRO content, providers cautioned that before starting the program, patients would need to be educated on the recommended dietary and physical activity guidelines for diabetic patients and the medications they are currently taking for their T2D to ensure they are reliably answering the questions. On the basis of this feedback, at the baseline visit, trained study staff provide a brief overview of evidence-based guidelines for healthy eating and physical activity for T2D using low-literacy and language-congruent patient education handouts from the American Diabetes Association and review the patient’s current diabetes medication regimen.
To integrate the report into clinical practice, providers preferred having a separate tab built into the EHR, which included a summary of the personalized report and highlighted key trends in patients’ PRO data over the past 3 months. All providers found value in discussing the report with patients during the clinic visit because the data complemented the questions that they had already asked about diabetes self-management. Finally, although they found value in the longitudinal trends displayed in the graphs, owing to time constraints, they felt that patients should bring up anything important that stood out in the detailed view. On the basis of this feedback, the study team is working in collaboration with NYULH MCIT to integrate the personalized report into Epic. This includes the development of security protocols that will link patients’ encrypted research ID to their medical record number and integrate the report image into an Epic web integration record. Web integration records are used to visually integrate external apps with Epic. Providers will be able to access the i-Matter report via a button located within the patient’s chart at the top of the Office Visit toolbar (
Screenshot of i-Matter Epic integration.
Although achieving glycemic control is of clinical importance, it is the daily experiences of living with T2D that drive patients’ decisions to adhere to treatment recommendations and become engaged in their care [
This paper describes the design and refinement of i-Matter through an iterative user-centered approach that actively involved patients and providers throughout the process. Active involvement of end users in the development of the intervention can help to address the difficulties with protocol compliance, lack of clinical integration in the EHR, and provider skepticism about the utility of PROs in practice, which are hallmarks of previous trials, thus increasing the likelihood of developing a sustainable approach [
Although there are many strengths of our intervention approach, we note limitations that can be considered for future research. First, although our intervention is designed to target patients with T2D, it is more common for patients to have 2 or more chronic diseases (ie, multimorbidity) than 1 disease in isolation (89.3% vs 8.5%, respectively) [
Finally, 2 (out of the 6) providers interviewed during the development of i-Matter indicated that they found less value in PROs that were not immediately actionable in primary care practice (eg, depression and quality of life). A key strength of the i-Matter study is the full EHR integration of the PROs with the health care team. Many previous PRO initiatives share the patient PRO data back with the providers in a workflow disruptive manner—asking providers to change their normal activities and make a special effort to review the PRO data. i-Matter overcomes these challenges by delivering the patient PRO data directly into the patient’s chart in the EHR—presented as just another commonly viewed data visualization by the provider such as patients’ lab and test results. Thus, our intervention will test the hypothesis that if actionable diabetes PRO data are delivered in the right context, it will influence patient-provider interactions. Early adopters of our intervention will also help to provide important data on the potential effectiveness and (time) efficiency of using PROs in clinical care. Sharing the outcomes of this work could provide providers who are hesitant to adopt such innovations with much needed information about the benefits of using these tools.
Example motivational and feedback messages.
capability, opportunity, and motivation model of behavior
electronic health record
end-stage renal disease
hemoglobin A1c
investigating an mHealth texting tool for embedding patient-reported data into diabetes management
Medical Center Information Technology
mobile health
New York University Langone Health
patient-reported outcome
research assistant
randomized controlled trial
type 2 diabetes
technology acceptance model
user-centered design
The authors would like to thank Aditya Verma and Sara Chokshi, DrPH, for their assistance with this project. This work was supported by a grant from Merck & Co, Inc, (principal investigator: AS) and the Agency for Healthcare Research and Quality, R01HS026522 (principal investigators: AS and DM).
AS analyzed, interpreted data, and drafted the manuscript. JC, LP, KL and MR acquired data and critically reviewed the manuscript. CJ and JG oversaw the EHR integration and critically reviewed the manuscript. MP, SP, and EL developed the mHealth platform and contributed to all critical revisions of the manuscript. DM interpreted data and critically reviewed the manuscript.
DM, JC, LP, MR, KL, CJ, and JG have no competing interests or financial disclosures to declare. AS is a consultant for Rip Road, Inc. MP, SP, and EL were paid as consultants to develop the mHealth intervention for this project.