Advice for Improving the Experience of Web-Based Patient Portals: Qualitative Interviews With Caregiver-Adolescent Dyads

Introduction

Web-based patient portals are resources that allow patients to access their electronic health information (EHI) and communicate with their clinicians through portal messages. Availability of portals has been bolstered by the 21st Century Cures Act, which was signed into US law on December 13, 2016. This law prohibits US health care systems and electronic health record (EHR) vendors from information blocking, “a practice [by a health care clinician or health IT developer, exchange, or network] that except as required by law or covered by an exception ... is likely to interfere with access, exchange, or use of EHI” [1]. This portion that expanded transparency in health care and access to EHI went into effect on April 5, 2021. Additionally, portal use surged during the COVID-19 pandemic when electronic communication, including video visits, increased, spurring uptake on a national scale [2].

In pediatrics, caregivers can access their child’s EHI through a proxy version of the portal. When children reach the age of adolescence (ranging from 10 to 16 years, depending on the US state) [3], most health care systems also provide adolescents with the option to register for their own portal account. However, each US state has laws that dictate which adolescent health information must be protected, usually related to sexual and reproductive health, mental health, and substance use disorders [4]. Each pediatric health care system must determine how to comply with these laws, and caregiver access may be limited to varying degrees when a child becomes an adolescent. As such, the usefulness of the caregiver proxy portal is often in tension, with restrictions required to ensure confidentiality of the adolescent’s sensitive, protected information.

Providing portal access can benefit both adolescents and their caregivers by improving understanding of one’s health, empowering adolescent self-management and autonomy, enhancing communication with clinicians, and supporting caregiver influence [5]. These benefits of improved medical competency and autonomy are especially critical for adolescents who are beginning to transition to adulthood and self-management. However, many adolescents demonstrate limited interest in using their portal [5], and their use has been low with studies citing variable rates of uptake between 15% and 64% [6-10]. In those who do use the portal, there is a high attrition rate, with one study finding that over 50% of portal accounts were actively used for less than 1 year [8].

Health care systems and administrators must consider devising strategies to simultaneously increase the usefulness of adolescent and caregiver proxy portals, support adolescent interest in using the portal, and maintain confidentiality of sensitive information. Input from adolescents and caregivers will be essential to inform these strategies, but few studies have explored these perspectives in depth. Most of this literature has focused on protecting confidentiality and increasing enrollment [11-16]. In this analysis, we aimed to address this gap by identifying adolescent and caregiver advice for medical informatics administrators and clinicians directed at improving the web-based patient portal experience.

Methods

We report this study following the COREQ (Consolidated Criteria for Reporting Qualitative Research) guidelines [17].

Participants and Recruitment

We conducted semistructured interviews with caregiver and adolescent dyads who received care in St Louis. We stratified sampling for adolescents with and without chronic illnesses and purposively sampled for racial and ethnic minorities and fathers. Adolescents with chronic illness were recruited from specialty clinics at a major academic medical center, including sickle cell disease (SCD), inflammatory bowel disease (IBD), and diabetes mellitus (DM). We chose these 3 chronic illnesses because they create lifelong medical complications and represent diverse racial and ethnic populations. Adolescents without chronic illness were recruited from 3 pediatric clinics and 1 adolescent medicine clinic associated with the academic medical center.

Patient Portal Policies

All clinics used the same patient portal (called MyChart) with the same requirements for creating portal accounts. This portal allows participants to view scheduled appointments, clinician notes, and diagnostic results, communicate with their providers or clinics, and request medication refills, among other functions. Children aged 11 years and younger cannot have their own MyChart account, and caregivers may obtain full access to a proxy account by completing a proxy access form. When a child turns 12 years, caregivers lose proxy access by default. Adolescents can create their own account if they call the help desk during business hours. To provide proxy access for parents of adolescents, caregivers need to obtain signed forms of approval by the adolescent and the patient’s primary doctor. Every child in a family requires a separate proxy portal account with a unique password. There was no ability to link accounts of family members or to link accounts across institutions that used a different patient portal. There is no phone app available, though individuals can log in through an internet browser on their phone or computer.

Data Collection

As part of a larger qualitative study [5], we performed semistructured interviews, in which we asked adolescents and caregivers to provide advice for administrators and clinicians regarding (1) providing adolescent and caregiver portal access, (2) how doctors should discuss the portal with families and write their notes, and (3) what general advice they had to improve portal access for families (Multimedia Appendix 1). This guide was informed by literature review, our prior work, and a stakeholder advisory board. BAS, CB, and another team member performed interviews between May and December 2022 via telephone or Zoom videoconferencing software (Zoom Video Communications) based on participant preference. Caregivers and adolescents were interviewed separately and privately. Interviews were audio-recorded and professionally transcribed. We collected demographics via a Qualtrics survey that was sent via email to caregiver participants.

Data Analysis

BAS, CB, and JMG developed a codebook following validated phases of thematic analysis [18]. Coding involved multiple iterative steps: (1) read transcripts to familiarize themselves, (2) descriptively coded 5 transcripts to formulate preliminary codes, (3) grouped codes into categories and collapsed categories into representative themes, and (4) refined definitions for themes through 3 cycles of independent coding and consensus meetings. After reviewing 20 adolescent transcripts and 19 caregiver transcripts, we reached saturation for representative themes.

CB and REG independently applied this codebook to all transcripts using Dedoose qualitative software, reviewed each other’s application of codes, marked disagreements, and resolved disagreements via discussion. Subsequently, REG reviewed transcripts for each dyad, identifying overlaps and contrasts between caregivers and adolescents, as well as overlaps and contrasts for adolescents with a chronic medical condition and their caregivers compared to adolescent-caregiver dyads who do not have a chronic medical condition. Quantitative data on patient-reported demographics were analyzed using SPSS Statistics (version 29.0.2.0; IBM Corp).

Ethical Considerations

The institutional review board at Washington University approved this study (202203147). We obtained informed consent from caregivers and assent from adolescents. All transcripts were deidentified prior to analysis. Participants were provided with a US $40 electronic gift card for their participation in the interviews.

Positionality Statement

BAS is a pediatric oncologist and communication researcher with training in qualitative research. REG is a fellow in pediatric hematology/oncology with a master’s degree in public health and prior training in qualitative research. CB is a research coordinator with a master’s degree in social work and prior qualitative experience. Another team member who performed interviews has a master’s degree in public health and qualitative methods experience. JMG was a medical student with training in qualitative research. All team members have prior experience in qualitative research.

Results

Demographics

There were 102 separate interviews performed with 51 dyads, averaging 25 (SD 11) minutes. The majority of adolescents and their caregivers were White (adolescents: n=28, 55% and caregivers: n=28, 55%) or Black or African American (adolescents: n=21, 41% and caregivers: n=21, 41%), non-Hispanic (adolescents: n=46, 90% and caregivers: n=47, 92%), and female (adolescents: n=30, 59% and caregivers: n=50, 98%). Many caregivers had accessed their child’s portal (n=33, 65%), but fewer adolescents had accessed their portal (n=17, 33%; Table 1).

Encouraging and Supporting Portal Use

Improving Accessibility

Caregivers reported significant challenges with the enrollment processes for the proxy portal. These caregivers advised that the entire procedure should be easier, stating that the experience was “confusing” and requesting that more information be provided. Both caregivers and adolescents reported that the login process could be easier, stating the requirements for adolescents to have their own email made accessing the portal difficult. In addition, the sign-up process for the adolescent proxy portal required the adolescent to call the help desk during business hours, which conflicted with school and sports schedules. Further, institutions had rules about required characteristics of passwords, such as numbers and symbols, that frustrated some participants. Adolescents and caregivers thought it would be beneficial to have increased resources to assist in the sign-up and usability of the portal platform, including a website, pamphlet, or an individual that could guide families through the process.

Clinician’s Role in Educating and Promoting Use

Caregivers and adolescents acknowledged that the portal should be “brought up more” (Adolescent 26, no chronic medical condition) by the clinician to encourage use, with one caregiver stating that it should be “a part of the visit every time” (Caregiver 26, no chronic medical condition). Only one respondent stated that they did not know “if that’s really a doctor’s responsibility to worry about the portal” (Caregiver 13, IBD). Some felt it would be useful for clinicians to “be more informational” (Caregiver 46, SCD) about the portal components and sign-up process, with one person stating that they felt the doctor should have experience using the portal to better understand the patient perspective. Others recommended that clinicians outline acceptable uses of the platform, including types of portal messages that are acceptable and how to best communicate with them.

Recognize the Importance of the Portal for Chronic Illness

Nearly all caregivers and adolescents acknowledged the importance of respecting and protecting adolescent confidentiality. However, many caregivers of children with chronic illness described the importance of retaining access to nonsensitive information in the portal, such as notes, appointments, laboratory results, and medication lists. These caregivers worried that limiting parental access to this actionable information could be “very detrimental” (Caregiver 51, SCD), especially because adolescents are often not mature enough to navigate health care on their own (Table 2).

Recognizing the Emotional Experience of Using the Portal

Mitigate Stress Through Discussion and Resources

To decrease stress, caregivers advised that clinicians discuss with families the maturity level of the adolescent prior to granting access to ensure that the adolescent is ready for access. Caregivers also recommended that adolescents should be walked through the portal at their first time of access. A few adolescents worried about portal access being problematic for parents who were “helicopter parents” or overbearing. These adolescents advised that clinicians discuss the types of information that might be released through the portal, prepare them for potentially stressful results, and encourage them not to become fixated on every abnormal result. One caregiver and one adolescent noted the importance of maintaining the option for verbal communication with clinicians outside of the portal, especially when worrisome test results are released.

Modify Note Content

Adolescents and caregivers advised that notes should be tailored to remove and protect confidential discussions that occurred between the adolescent and their clinician. Some adolescents and caregivers advised a collaborative discussion between clinicians and adolescents about what specifically should be excluded from the notes to ensure confidentiality tailored to the comfort level of that patient and the patient-caregiver dynamic. Another caregiver suggested the clinician inform the adolescent that the confidential portion would not be included in the note to build trust.

Caregivers and adolescents also had recommendations for the content and language that were included in the portal. Many caregivers and adolescents stated the importance of considering adolescents’ emotions when writing notes, which involved using professional language and “stay[ing] positive and to stay just on facts. Maybe not put personal thoughts” (Caregiver 4, DM). A few participants discussed concerns that comments about the adolescent’s weight could have a negative emotional effect. These participants advised clinicians to be “sensitive and mindful” (Caregiver 42, no chronic medical condition) and use “very gentle phrasing” (Adolescent 23, no chronic medical condition). One adolescent reflected that they had “a doctor make jokes about me in my medical notes before ... It just made me not trust the doctor really, and withhold information from them because I didn’t wanna be judged about it” (Adolescent 10, DM) (Table 3).

Improving Portal Usability, Understandability, and Individualization

Improving Interpretation of Laboratory Results and Notes

For interpreting laboratory results, adolescents and their caregivers noted that having a dictionary, glossary of medical terms, or additional information such as reference values would be beneficial. Some specifically stated that without these resources, they resort to web-based searches to interpret results. Adolescents without chronic medical conditions did not have advice regarding improving laboratory understanding for individuals accessing the portal, presumably because they had few or no laboratory values in their medical records. Regarding clinician notes, many adolescents and caregivers described a “language barrier” (Adolescent 2, DM) due to the medical terminology and complexity of notes. They advised that notes should be understandable to the average person, with “everyday language instead of medical terms” (Adolescent 13, IBD). One caregiver specifically advised how clinicians’ notes could be a “bridge from the labs or the records” (Caregiver 8, IBD), with notes providing an explanation for difficult-to-understand laboratory results. Others advised that the notes did not need to change but that clinicians should speak with adolescents and their caregivers about the readability of these notes. A few adolescents and caregivers had suggestions regarding the content of notes, stating that they should contain next steps, address the concerns of the caregiver and patient, and be transparent and detailed.

Improving Portal Design

Some adolescents and caregivers reflected that the patient portal should be overall “user-friendly” (Caregiver 30, SCD) and “making sure that everything is findable, that needs to be found” (Caregiver 28, no chronic medical condition) with an organized layout. Caregivers of children with a chronic medical condition noted the potential benefits of having the office and doctor’s contact information “in case we need to get ahold of them quickly” (Caregiver 9, IBD). Several adolescents noted that “it would be more acceptable if it was an app” (Adolescent 46, no chronic medical condition).

Individualizing the Interface

Caregivers and adolescents recommended improving the level of customization of portal contents for adolescents and their caregivers. Suggestions included having a simpler interface for adolescents, having a more summarized note personalized to the adolescent, or personalizing the interface with users selecting for desired or undesired features.

Linking Accounts

Caregivers and adolescents recommended a feature by which portal accounts could be linked, such that caregivers could have one sign-in for all children or family members. Participants noted that this would ease usability and access, which can be hindered by requiring multiple logins and passwords. Caregivers also recommended an additional feature where records could be linked and shared across institutions, which would enable the streamlining of medical care (Table 4).

Discussion

Principal Findings

Using the technology acceptance model framework, we consider the 2 key factors of perceived ease of use and perceived usefulness that influence the attitude, behavioral intention, and actual portal system use [19]. Adolescents and caregivers provided advice on how to improve the perceived ease of use, including improving the accessibility of portal enrollment and access, the understandability of its contents, and mitigation methods for potential emotional distress that could result from portal use. Regarding its perceived usefulness, while some recognized the usefulness and importance of the portal, many stressed that the perceived usefulness could be bolstered by clinician promotion of the portal and by improving the understandability and customization of its contents. There were three overarching themes that permeated the adolescent and caregiver recommendations: (1) encouraging and supporting portal use, (2) recognizing the emotional experience of portal use, and (3) improving portal usability, understandability, and individualization.

Increasing resources for accessing, using, and understanding the portal was nearly universally recommended. Areas that were identified to benefit from increased resources were establishing portal access and initial use as well as improving the understandability of laboratory results and clinician notes. This is particularly salient when discussing portal design for the adolescent age range, as the target audience’s maturity level and reading skills evolve over time. Some resources could be standardized, such as handouts to understand the enrollment process, a tutorial to demonstrate display features, or a glossary that explains basic medical terms. Other aspects, such as interpreting radiology reports or understanding the potential impact of an abnormal laboratory result, may require a more nuanced and personalized approach. Improvement on these more nuanced aspects could involve efforts to improve communication with clinicians or the incorporation of advanced technologies, such as large language model (LLM)–powered chatbots. Prior work demonstrated that LLMs can improve the readability of radiology reports [20,21]. Another study found that LLMs were able to generate accurate and comprehensive responses with minimal risk of harm to radiation oncology patient care questions, though with a higher-than-recommended readability level [22]. Additionally, LLMs can provide inaccurate information or generate “hallucinations,” which “include a vast array of failures of faithfulness and factuality” of the LLM [23]. More research, particularly in its application to the pediatric and adolescent population, is required to determine the safety, acceptability, and usefulness of LLMs in supporting portal use.

This study further emphasizes and expands on the idea that clinicians should be discussing patient portals with families. While clinician endorsement can influence portal engagement [24,25], this study is novel in identifying other important roles of clinician communication for the portal experience, including setting up parameters for confidentiality, alleviating emotional distress, and outlining expectations and communication preferences. Many caregivers and adolescents had recommendations for how clinicians may improve communication by way of their notes. These recommendations included using sensitive language that recognizes the potential emotional impact, using layman terms, and removing confidential information [26]. This research underscores the importance of clinician communication in promoting appropriate use and enhancing the effectiveness of the patient portal. Furthermore, it raises difficult questions about the purpose and audience of clinical notes. Should these notes primarily be written for medical documentation and communication with other clinicians, for understandable communication with caregivers and patients, or for liability or billing purposes? The same note might not be able to fulfill these divergent purposes [27].

Many caregivers and adolescents provided recommendations on improving the layout, as well as how EHR vendors could improve portal engagement and use by focusing on a user-centered design, which seeks to adapt the design and functionality of the portal based on user feedback and goals. Research in adult populations has found that identifying user goals and interests and then tailoring portal design and function to meet these needs improves engagement in the portal [28,29]. Furthermore, as research in adolescent and adult populations shows that portal engagement is associated with better medical adherence and understanding, as well as a better perception of control over one’s health, this engagement could have significant downstream effects [30-33]. Given that many adolescents show limited interest in using the portal [5-8], engaging in user-centered design might increase the perceived usefulness and ease of use of portals, which are central drivers of uptake and acceptance of novel technologies [34].

Strengths and Limitations

This study had several limitations. First, while we purposively sampled for racial and ethnic minorities, as well as fathers, our study was limited by excluding non-English speakers and having few Hispanic participants and few fathers as participants. Therefore, these perspectives may not be appropriately reflective or generalizable to these populations. Second, adolescents with greater concerns about privacy and confidentiality might have been less inclined to participate in this study, which may result in a decreased emphasis on this aspect compared to a more generalized adolescent population. Third, as patients were recruited from a single health care system in St Louis, MO, using a single EHR platform, the results may not be generalizable to other regions of the United States or to other hospital systems using a different EHR platform.

This study also had several strengths. First, the unique perspective of this qualitative research evaluates not just the barriers and benefits of portal engagement for adolescents and their caregivers, but uniquely identifies potential solutions. Second, while the study was limited in having few fathers, few Hispanic participants, and excluding non-English speakers, it still represents a diverse population by including both adolescents and caregivers, those with and without chronic medical illnesses, and purposively sampling for racial and ethnic minorities. This improves generalizability by providing advice for clinicians and administrators that serves a broader group of patients and families, particularly when considering those with versus those without a chronic medical condition, who may have different priorities when using the portal.

Conclusions

Caregivers and adolescents had specific recommendations for administrators and clinicians to improve the usefulness of the patient portal. Adolescents and caregivers universally recommended more resources for accessing and using the portal, improved clinician communication regarding the portal, and enhanced user-centered design of the portal. Clinicians can take steps to better understand and inform families about portal access and functions. However, the onus lies with administrators and EHR vendors to make structural changes to the portal user interface and functionalities that can improve the usefulness and usability of the portal. The advice provided in this study should guide these improvement efforts.