Acceptability and Feasibility of a Disease-Specific Patient Portal in Adolescents with Sickle Cell Disease

Abstract

Objective:

To examine the acceptability, feasibility, and efficacy of a healthcare portal.

Observations:

Adolescents and young adults (AYA) with sickle cell disease (SCD) were taught how to use an SCD-conditions page in MyChart^® and completed questionnaires at baseline, post-intervention (T2, 6 weeks after baseline) and after 3 months (T3). Forty-four participants (M age = 18.82, SD = 2.72) viewed an average of 58.07 pages from T1 to T2. The portal was highly accepted (90.32%). Efficacy data indicated that portal use was associated with improved patient-provider communication.

Conclusions:

Electronic portals are promising tools for improving medical self-management.

Tethered patient portals connected with the electronic medical record (EMR) allow patients to access certain aspects of their health information such as lists of prescriptions and laboratory results and to securely communicate with their medical team. A number of portals have been developed for specific diseases including diabetes,[1,2] heart disease,[3] and cancer[4] with the aim of engaging patients in their healthcare to improve health outcomes. In addition to providing medications and lab results, disease-specific portals may provide additional disease-specific information and educational materials.

Although research with adult samples has demonstrated the effectiveness of patient portals for disease self-management,[5] there is a dearth of research on self-management portals designed for AYA. Research is needed with AYA for several reasons. First, adolescence is a critically important time to intervene to improve self-management as this is often the time when healthcare responsibility transfers from parents to youth.[6] Second, in addition to needing more research on portals for AYA in general, there is a need to develop a portal specifically for sickle cell disease (SCD). Because the medical demands and clinical complications of SCD increase starting in adolescence,[7] novel approaches for engaging AYA in medical self-management are critically needed. Thus, the current study examined the efficacy and acceptability of a web-based portal designed for AYA with SCD.

This study had three aims. The first aim was to examine the acceptability of a disease-specific patient portal tethered to the EMR. We hypothesized that a web-based portal would have high acceptability and high ease of use among AYA. The second aim was to examine the feasibility of the portal. The specific feasibility objectives of this study were to determine if AYA would use the portal, to describe their patterns of use, to assess retention, and to describe characteristics of who would use the portal. We were also interested in the feasibility of conducting the intervention itself (i.e., able to integrate data successfully). The third aim was to examine initial efficacy of the intervention. We hypothesized that portal use would significantly predict improvements in medical decision making, patient-provider communication, and patient-provider relationships.

Materials and Methods

Participants

In this multi-center study, participants were recruited from pediatric and adult medical clinics serving AYA with SCD in the Midwest through letters, phone calls, clinic appointments, or SCD-related events. AYA were eligible if they had been diagnosed with SCD and were between the ages of 13–24 years (61 approached). Eligible participants who agreed to participate in the study signed consent/assent forms at the start of the study; parental consent was obtained for those under age 18 years.

Procedures

This study included the following three timepoints: baseline (T1), post-intervention (T2, 6 weeks after baseline), and a 3-month post-intervention follow-up (T3). At T1, participants were taught how to use an SCD-specific version of MyChart^® connected with the Epic^® EMR and asked to find a specific test result from a recent visit and determine if it were in the normal range. To encourage AYA to use the portal, AYA were given three additional homework assignments to complete from T1 to T2 that required logging into the portal. The three homework assignments were to 1) find their medication list; 2) find a past appointment summary; 3) graph test results over time. It was anticipated that each homework assignment would take about 5 minutes to complete. Homework was turned in via mail, email, or at a clinic appointment (see Figure 1 for depiction of assignments). One reminder call occurred for each assignment. Research assistants were available to answer any questions.

Figure 1.

Depiction of the enrollment task and three assignments

To assess the efficacy of the brief intervention (teaching how to use the portal and assigning homework), AYA completed questionnaires at T1, T2, and T3. Participants were compensated for their time in completing questionnaires at T1 ($35), T2 ($35), and T3 ($40), as well as completing the enrollment task and homework assignments ($25 per task).

In terms of study feasibility, we assessed the success of each step in the study protocol. Steps that were expected to result in more difficulties were the following: 1) researchers had to turn in signed MyChart^® agreements for each participant to hospital staff in the medical records department in order to activate accounts; 2) data from MyChart^® had to be downloaded and integrated with survey data and homework completion data. There were no technical difficulties in completing any of these steps. We were able to assist participants in their first MyChart^® login, easily access data from MyChart^®, and merge data from multiple sources successfully. Results indicated that it is feasible to implement the intervention and to examine outcomes.

SCD-Specific Portal

MyChart^® for SCD is a disease-specific website designed specifically for patients who have been diagnosed with SCD. Through this portal, patients can view upcoming appointments, past appointments and the instructions that were provided, current medications and the instructions for them, results for tests and labs, current diagnoses, allergies, immunizations, information on preventative care, family medical history, hospital admissions with discharge instructions, and medical information on SCD. After the EMR is updated following a clinic visit or with lab/test results, that information is automatically available in MyChart^®. Patients can also request a medical appointment or message their providers. After enrollment, 18 participants sent messages to providers through the portal (M=.80, SD=1.46, range=0–7). The portal was created with the hope that if patients knew more about their health, then they could be a part of conversations about the treatment plan and make more informed decisions.

Measures

Objective portal use. The number of pages that participants viewed was extracted from the portal’s analytics at each timepoint (i.e., from T1-T2 and from T2-T3). Next, participants were statistically divided into low, average, and high users, and a subset were interviewed about their use of the portal to further elucidate patterns of use (see Table 2).

Table 2.

Interview responses from a randomly selected subset of low, average, and high users of the portal

User Type	1. Why do you use MyChart®?	1a. If you don’t use MyChart®, or haven’t used it very much, why not? What keeps you from using it?	2. What could we change/add that would make you more likely to use MyChart®?	2a. If we posted new info 1–2x/month, would you log-in to read it? Why or why not?	2b. If we had videos on MyChart® about how to use different features, would you watch them?	3. What did you think MyChart® would have or do that is not there or it does not do?	4. What do you think are the 3 most important pieces of information for you to know about your Sickle Cell?
Low	I don’t use it much.	I do not have internet at home. Plus the clinic sends appointment reminders to my house, so I do not really need MyChart®.	Nothing.	No. My mom researches SCD really well, so I know everything that I need to know. There is nothing new that would be helpful.	Maybe	Nothing	1. Medications including instructions and dose, 2. Hemoglobin levels because the doctor told me it was important to know my levels
Low	So that I can see what is going on with my health	N/A	I would like to use video-chat to talk to my doctor.	Yes, because I could learn new information and see if my health is changing.	Yes	I don’t know.	1. How to manage pain, 2. Which medicines to take, 3. Appointments.
Low	To find out information about my health issues and to contact my doctor.	N/A	Problems should be alerted in red. Emails to doctors should be red alerts, so that doctors can reply right away.	Yes, because I want to know any new information about SCD.	No, I read and log out unless I have spare time. The length of the video would make a difference.	It does not provide steps on how to overcome sickle cell disease. It only tells you the conditions you have. I want the doctor to put in specific steps to follow.	1. Issues, 2. Diagnosis, 3. Ways to overcome symptoms of SCD
Low	N/A	I only use it at the hospital. I do not need it at home.	Nothing. I am not going to use it.	I probably would, because I like to learn.	No, because I know how to use MyChart® already.	N/A	1. How not to pass SCD onto children
Average	I like to browse to see information and to keep updated on sickle cell.	I only need to check it about once a month to find the updated information.	I would like more pictures. The explanations are great.	Yes, because knowledge helps to live a long and healthy life, and I only go to the doctor so often.	Yes, as long as the videos are not long.	Nothing, the website is fine the way it is.	1. How to manage pain, 2. What medication does, 3. Nutrition recommendations
Average	I wanted to see what the doctors put in their medical notes and to check my lab values.	I do not always remember my login and password.	I liked that there was not too much detailed information. It would be nice if you could click on big words to get definitions. You could add temperature alerts with friendly reminders (e.g., “Bundle up! It is cold outside.”).	Yes, I would login more to learn and see what information was added.	Yes, but only if MyChart® were more difficult to use.	Nothing	1.What each medication does, 2. How temperature/weather affects SCD, 3. Why it is important to stay hydrated
Average	N/A	I only use it to see appointments. There is no other reason to use it.	A feature to add would be that calls could be made to a nurse or care manager from the website.	Yes, to learn something new	No	Nothing	1. How to manage pain better
Average	To use the messaging system.	N/A	I would use it more if it were more interactive (e.g., a live chat system like instant message with the nurses or doctor for quick answers to questions).	Yes, because it might be useful.	Yes	Nothing	1. What I have to do for treatment for the rest of my life, 2. Why I do my treatments regularly, 3. Why it is important to take medications
High	I like to check appointments and medications. It helps to talk to doctors and learn information about sickle cell.	N/A	Nothing to add. It is easy and simple.	Yes, I want to learn new health information.	Yes	Nothing	1. Medications, 2. What to do during pain crisis, 3. Information about SCD.
High	I use it to check test results every two months.	I do not think I need to use it.	No, there is nothing wrong with it. I still have to come to the hospital to get results though.	Yes	Yes	Research about SCD (results of studies) to help learn about my disease	1. What does SCD mean and how I to explain SCD to other people, 2. How to avoid a pain crisis such as not staying outside without hydrating 3. Medications (e.g., requesting refills).
High	I like to check when my next appointments are. It is helpful to see hemoglobin levels. I tell my mom the results about once a week.	N/A	I like it how it is.	Yes, the doctor says we need to learn more tips about how to deal with pain crises since I do not like to take medications.	No, it would probably help other people, but I do not need them.	Nothing. I honestly thought the website would be pointless, but it actually has been helpful.	1. Why I need to take medications and what is important about each medication, 2. How to manage pain without using medications, 3. How to explain to other people about SCD.

Portal survey. Participants answered questions about ease of use and usefulness on a 5-point Likert scale ranging from Strongly Disagree to Strongly Agree from the System Usability Scale.[8] They rated the quality of the portal using a 5-point Likert scale ranging from Poor to Excellent.

Medical decision making. Participants answered the following question on a 5-point Likert scale ranging from Strongly Disagree to Strongly Agree: “My use of the chronic disease portal has improved how I make health care decisions about my condition.” This question was adapted from published studies of portal use.[5]

Patient-provider communication and relationship. To assess communication, the Talking with Providers subscale of the Transition Readiness Assessment Questionnaire (TRAQ-5), a valid and reliable measure,[9] was included. The Talking with Providers subscale includes the sum of two items which assess the patient’s willingness to answer questions and to share their feelings with providers rated on a 5-point Likert scale.

To assess perceptions of the patient-provider relationship, participants answered the following question on a 5-point Likert scale ranging from Strongly Disagree to Strongly Agree: “My use of the chronic disease portal has improved my relationship with my provider.” This question was adapted from published studies of portal use.[5]

Analysis Plan

For Aims 1 and 2, acceptability and patterns of use were described by examining univariate statistics (means, frequencies). For Aim 3, separate multivariate path analyses were conducted in Mplus version 8 using maximum likelihood (ML) estimation. All response variables on a 1–5 ordinal scale were log-transformed prior to analyses, because research has demonstrated that log transformation with ML estimation preserves overall relationships well.[10]

Results

Demographics.

Forty-four African Americans with SCD (M age = 18.82, SD = 2.72, range 15–24 years; 54.50% female) participated in the study. See Table 1 for demographics.

Table 1.

Description of demographic variables.

Variables	N	%	Mean	SD
Age	44		18.82	2.72
Adult-status	44
<18 years old	18	40.91
Adult	26	59.10
Sex	44
Male	20	45.5
Female	24	54.5
Race/Ethnicity - African-American	44	100

Aim 1 - Acceptability.

In terms of acceptability, 90.32% rated the portal a high-quality website and 77.42% found the portal very easy to use. Further, 80.70% indicated that they either agreed or strongly agreed that the portal was useful. Refer to Table 2 for interview responses describing specific aspects of the portal that participants liked and disliked.

Aim 2 - Feasibility.

Because participants were taught how to use the portal as part of the intervention, everyone logged into the portal at least once, with a mean of 58.07 pages viewed from T1 to T2 and 25.73 pages viewed from T2 to T3. The median number of pages viewed from T1 to T2 was 38.50, and the median number of pages viewed from T2 to T3 was 0 (T1 to T2 ranged 12 – 360 pages viewed; T2 to T3 ranged 0 – 440 pages viewed). When examining retention, after 3 months, 45.5% continued to log into the portal. The intervention included assigned homework that required logging into the portal to access information. In terms of homework completion, 82.76% of participants turned in homework.

To examine the characteristics of those who may have been more likely to use the portal, age, gender, and hemoglobin type were included as predictors of portal use in a multivariate path analysis. Neither age, gender, nor hemoglobin type significantly predicted portal use at T2 or T3 (all p’s > .05).

Aim 3 - Efficacy.

In the first path analysis with portal use from T1 to T2 included as the predictor and medical decision making at both T2 and T3 included as the outcomes, portal use did not significantly predict improved medical decision making, [T2: b=.001, SE=.001, ß=.17, t=.86, p=.392; T3: b=.002, SE=.001, ß=.39, t=1.60, p=.109].

In the second multivariate path analysis, portal use from T1 to T2 was included as the predictor, and both patient-provider communication and patient-provider relationships at T2 and T3 were included as the outcomes. Although portal use did not significantly predict improved relationships [T2: b=.001, SE=.001, ß=.01, t=.10, p=.922; T3: b=.002, SE=.001, ß=.32, t=1.57, p=.117], portal use did significantly predict improved communication with providers at T2 but not at T3 [T2: b=.002, SE=.001, ß=.31, t=2.56, p=.010; T3: b=.001, SE=.002, ß=.10, t=.43, p=.666].

Discussion

This study examined a brief intervention designed for AYA with SCD to provide in-person assistance for the first logon of a tethered patient portal (MyChart^®) specific to SCD and to provide accountability and structure for using the portal through homework assignments. Results indicated high acceptability and feasibility of the portal. Overall, youth rated the portal as useful, easy to use, and high quality. Use of the portal was associated with improvements in patient-provider communication (but not decision making or relationships). Although patients reported that they communicated with their providers more, they did not indicate that the portal helped improve patient-provider relationships. A possible explanation for this null finding is that patients may already have had strong relationships with providers at baseline. Alternatively, using a web-based portal may feel disconnected from the human interaction that occurs during clinic visits.

Implications

Although there are many benefits of portals (e.g., making important health information easily accessible), it was unknown if AYA would use a portal. We anticipated that there would be several barriers preventing youth from using portals based on previous research. First, a systematic review by Goldzweig and colleagues[5] indicated that racial and ethnic minorities in the US are less likely to use web-based portals, which is particularly relevant for SCD, because SCD disproportionately affects African Americans and Latinos.[5] Second, research by Crosby and colleagues on youth with SCD identified that although the majority of AYA with SCD have personal smartphones or computers, access the internet multiple times per day, and use the internet for health information; they also suffer from frequent internet service disruption.[11] Thus, in order to help AYA overcome barriers to accessing the portal, this study included an intervention to teach AYA with SCD how to use the portal and encourage use through assignments.

Based on the interviews that were conducted with users who had low, average, and high levels of MyChart® use, there were several factors that influenced why patients with SCD did or did not use the portal. Barriers to using MyChart^® included lack of internet access at home, forgetting the password, and lack of perceived need. Adolescents who used the portal noted that they wanted to check their lab values, learn more about SCD, see when appointments were scheduled, and use the messaging system (see Table 2).

When Phelps and colleagues[12] examined patient use of portals for kidney disease across all hospitals using the system, they realized that two hospitals had administrators who provided support for the first login. They found that this initial support was associated with a 20% increase in the probability of continuing to use the portal after 2.5 years compared to patients within hospitals that simply provided written instructions on how to access the portal[12]. Consistent with Phelps’ serendipitous finding, the current study examined the intentional implementation of providing support for the first log in. The results provide initial support for this intervention, such that providers who use patient portals are encouraged to systematically provide in-person support for the first login.

Because past research has demonstrated that racial and ethnic minorities in the US are less likely to use medical portals [5], it is vital that healthcare professionals implement strategies to increase portal use among diverse groups. In this study, a simple intervention to introduce African American AYA with SCD to a portal demonstrated initial acceptability, feasibility, and efficacy. Further, about half of the participants continued to use the portal after the intervention period was finished. Thus, healthcare professionals are encouraged to allow time to teach patients how to log into the portal at medical appointments and to give simple assignments to increase the likelihood that AYA will use a portal. In terms of provider perspectives, a recent systematic review revealed generally positive attitudes from providers toward patient portals including time savings, improved communication, and decreased no-show rates to clinic appointments.[13]

Limitations and future directions

Several limitations of the current study must be noted. First, this study used single items for several of the outcome measures assessing satisfaction, quality of the portal, and medical decision making. Although these questions have been used in previous research studies, it is possible that significant changes were not noted for medical decision making, because using one item may not be the most valid way of assessing this construct.[5] Future researchers may consider using multi-dimensional outcome assessments to increase reliability and validity. Second, this pilot study was designed to test the feasibility and initial efficacy of an intervention to increase portal use among AYA with SCD and thus, the sample was modest (N=44). Future researchers may want to test the intervention using a larger, randomized-controlled design. Third, although the average age of AYA was 18 years, this study included adolescents as young as 15 years. While these AYA are likely involved in medical decisions, caregivers/parents may still play an important role in medical decision making. Future researchers would benefit from including parent data when examining how portals are associated with health decision making in younger adolescents. Fourth, if the information available through the portal is not engaging or AYA are not motivated to learn, the portal may not increase AYA’s knowledge and self-management skills. Future iterations of the portal would benefit from including strategies in the portal itself to increase engagement, so that the portal will be used by more AYA, and not just AYA who are already motivated to access their health information and learn about SCD.

In this study, the results examining patient-provider communication demonstrated that portals were related to improved communication. However, future researchers may examine other ways in which portals may improve patient outcomes in order to provide more detailed information about the potential impact of these portals on the outcomes for AYA with SCD.

Several lessons were learned from this pilot study that may help inform future research. First, although all participants used the portal initially, effects were not sustained after 3 months for all participants, suggesting a booster session could be evaluated. Perhaps a staff member could check in with adolescents who have disengaged to facilitate problem solving around barriers. Second, when examining efficacy, it appears that the portal may be too passive of an intervention to improve variables other than patient-provider communication. Researchers may need more active interventions to improve medical decision making and patient-provider relationships.

Conclusions

Study results demonstrate the acceptability, feasibility, and some initial efficacy of an EMR-based, disease-specific healthcare portal designed for AYA with SCD, and contribute to the limited research currently available on medical portals. Electronic patient portals may be an effective tool that medical providers can incorporate into patient care to improve medical self-management. The integration of clinical support tools has the potential to transform the delivery of care for patients with SCD, improving morbidity and mortality and decreasing health care costs.