Abstract
Background: Patient portals provide patients with access to their health information, including medical history and conditions, lab results, and upcoming appointments and ability to message their care team. However, not all patients have access to their portal.
Methods: A cross-sectional survey of 233 adult family medicine patients from 4 community health clinics and a local community organization was conducted. The survey assessed patient demographics, internet usage, device ownership, and engagement with patient portals.
Key Results: Nearly half (47%) of patients had never accessed any portal, whether their own, a friend’s, or family member's. Among patients without home internet, 66% had never used a portal. 36% of laptop owners, 37% of desktop owners, and 34% of tablet owners had never used a portal. Across different self-identified identities, 46% of White patients, 41% of Black patients, and 53% of Hispanic/Latino patients had never used a portal. Among English speakers, 42% had never used a portal, compared with 59% of Spanish speakers. Roughly one-third (32%) of patients were interested in portal training, with more than half preferring individualized training.
Conclusions: Significant disparities in patient portal access exist in safety net practices. In addition to limited access to digital technology, lack of awareness of portal benefits and lack of clinician discussion and assistance likely contribute to inequities. Many patients desire individual portal training.
- Digital Divide
- Doctor-Patient Relations
- Health Care Disparities
- Health Information Technology
- Internet Access
- Limited English Proficiency
- Patient Engagement
- Patient Portals
- Rochester
- Social Determinants of Health
- Surveys and Questionnaires
- Vulnerable Populations
Introduction
Patient portals are vital in providing patients with access to their electronic health record (EHR). Over 90% of health care settings offered patient portals in 2018, representing an increase from 43% in 2013.1,2 Patient portals enhance health information access and enable secure, bidirectional messaging between patients and clinicians.3 This minimizes the need for long waits on phone trees by patients seeking to set up appointments, pay bills, request refills, or ask their clinician simple questions.4
The digital access divide has created a digital health divide.5–8 Studies show significant disparities among people who are members of minoritized groups,9–14 the disabled community,15 low socioeconomic status,13,16 and non-English speaking populations.13,16,17 Even though clinicians may offer portal use to their patients, one national study shows that only 15% of hospital and 30% of office patients will access their health records through the portal.18 Another national survey showed racial disparities in being offered a portal.9
Clinicians can facilitate patients’ access to federal assistance programs. Research suggests that collaborative efforts between health care settings and community-based organizations could yield benefits to patients beyond what each could achieve alone.19 This collaborative model could be extended to patient portal utilization. The extent to which patients enrolled in such programs are currently engaged in their patient portals is unknown.
This study aimed to characterize patient portal usage in patients from Rochester-area health systems. We characterized the associations of patient portal use, interest and engagement, and patient characteristics. Secondarily, we explored the associations between patients’ participation in federal assistance programs and patient portal engagement.
Methods
Survey Design
We designed the survey by adapting questions from the Health Information National Trends Survey,20 piloted and adjusted based on feedback from 6 community members. Changes included the removal of potentially stigmatizing language such as “low-income.” We included questions on patient demographics, internet usage, federal program participation, device ownership, patient portal usage and skills, as well as awareness and usage of the Lifeline and Affordable Connectivity Programs. A skip pattern ensured the survey was relevant to individual participants by excluding training questions for patients already using the portal. The denominator of the survey was not recorded, so no formal response rate was calculated.
Recruitment
Survey respondents were recruited using convenience sampling through their connection to federally qualified community health centers and community-based organizations serving economically disadvantaged patients.
Setting
We conducted convenience surveys of 233 participants at community health centers and community events in the city of Rochester. Three sites were part of a large Federally Qualified Health Center (FQHC), one was affiliated with a university, and the fifth was a community-based organization. The clinics served patients who were members of minoritized groups and typically insured through Medicaid.21 Surveys were conducted primarily in English and Spanish. We used Google Translate or an on-site translator for other languages.
Patient Eligibility and Survey Administration
To qualify for the survey, participants were 18 or older. Patients at the FQHCs were approached prior/after their office visits and invited to participate. Depending on patient preference, surveys were read to the patients or completed by patients. Patients received a $15 gift card for survey completion. Care managers surveyed participants from the community-based organization over the phone, using the same survey instrument as the research assistants, and participants either had the card mailed to them or came in to pick it up in person from the care manager. Data from the article surveys was entered into RedCap.22 Following the survey, the interviewer offered to help the patients with enrolling and navigating their health care portals and provided other resources on digital health care literacy. This project was undertaken as a Quality Improvement (QI) initiative designed to inform QI strategies As a community-driven quality improvement effort, not as a research study, according to the University of Rochester’s guidelines, it did not count as human subjects research as defined by the US Health & Human Services Common Rule 45 CFR 46 and was not submitted to the Institutional Review Board (IRB) and or assigned a nonexempt IRB protocol number.23
Statistical Analysis
We conducted descriptive statistics on all patient variables and used χ2 tests to examine factors associated with portal use and awareness. Analyses were conducted using STATA/S.E. version 18 (StataCorp).
Results
Of 233 respondents, ages ranged from 18 to 80, with a mean age of 48 (Table 1). Most self-identified as Black/African-American and 37% as Hispanic/Latino. English was the primary language spoken by 71% of patients, with Spanish, Ukrainian, Mai Mai, Karen, American Sign Language, and Khmer also represented. One-in-four patients had less than a high school education, while 39% had a high school diploma or equivalent. Most (72%) patients relied on Medicaid for health insurance.
Patient Demographics of Survey Respondents (n = 233)
Only 34% of respondents recalled being told about the portal’s benefits by the medical office, but more than half (55%) received assistance in accessing their portal (Table 2). Portal expertise was evenly distributed among the patient population. Portal usage within the last year varied, with the most patients (29%) using the portal 1 to 2 times. Most portal users accessed the portal to view test results (81%) and accessed their portals using their smartphones (85%). Most users reported portal navigation to be neutral, easy, or very easy (87%).
Awareness, Access Patterns, and User Experience with the Patient Portal Among Survey Respondents
Nearly half (47%) of patients had never used any patient portal (Table 3). Of those without home internet, two-thirds had never accessed a portal. About one-third of those who owned large devices had never used a portal. Portal use showed slight variation by race/ethnicity: 46% of White, 41% of Black, and 53% of Hispanic/Latino patients had never used a portal. Of those who identified their primary language as English, only 42% had never used a portal, compared with 59% of Spanish speakers.
Demographics and Technology Access of Patients Who Never Used Any Type of Portal (n = 109)
Among respondents asked about training on portal navigation (n = 131), one-third (32%) were interested (Table 4). The most common reason for not wanting portal training was a preference for in-person communication as opposed to portal messages, followed by privacy concerns, lack of a device, and discomfort with technology. Half (52%) of those interested desired one-on-one training. We observed no association between interest in portal training and primary language, education, zip code, or the primary care provider's endorsement of the health care portal.
Interest in Patient Portal Training and Format Preferences Among Eligible Participants (n = 131)
Discussion
Among a sample of predominantly low-income and minority patients, nearly half of the patients surveyed had never accessed their patient portal. This lack of engagement was notably high among patients without home internet. Many portal users reported receiving assistance in accessing their portal. Among nonusers, 32% were interested in assistance, mainly one-on-one training.
National surveys yield similar findings, with lower portal usage among patients with less education and limited internet access.24 While other studies show much higher rates of portal usage among White and English-speaking patients,25–27 our study does not follow the same trend regarding racial disparities found in portal usage, presumably because we sampled adults from safety net health systems. The data regarding English and Spanish speakers’ portal usage remains consistent between our study and previous research. English speakers are much more likely to have used a portal. This could be attributed to a lack of awareness or the limitations of English-only portals.
Our findings, in the context of existing research, suggest opportunities for improving digital health access in primary care clinics. First, our finding of high smartphone ownership suggests that most low-income patients possess the technology and internet connectivity required for portal access. For individuals without such access, several national initiatives aim to bridge these gaps. Federal programs include the Lifeline and the now-terminated Affordable Connectivity Program, which provide subsidies for internet access to low-income individuals.28,29 Primary care clinics can connect patients with information and support to access these programs. Second, clinics may consider raising patient awareness of the benefits of portals: accessing clinic note summaries, viewing laboratory results, and secure messaging with their clinicians.30 Lastly, clinics and/or community-based organizations may consider assisting patients with patient portals.31 This may include helping patients create an e-mail account, downloading the portal app to their phone, and setting up their accounts. Assisting patients in accessing their portal has the strongest evidence base for improving portal access.31 Our findings are consistent with national samples, documenting that the most frequent reason for declining is that they want to discuss it with their clinician.31,32 Most importantly, primary care clinicians can reduce inequities by recommending that patients enroll in portals and offering to assist them. In our experience, this often requires no more than 1 to 3 minutes of time. As a result of this work, our findings have directly informed our train-the-trainer model for community health workers in the region.
Study limitations include the use of a modest-sized, convenience sample. In addition, there could be potential for data entry errors, as article-based survey responses were manually entered into REDCap rather than using direct electronic entry with programmed logic constraints. However, our findings are supported by national data and other data.9–14
In conclusion, we observed that patients from safety net health systems and those served by community-based organizations have suboptimal use of patient portals. Our findings suggest that the provision of individualized training might help more patients access their portals.
Acknowledgments
We sincerely thank Dr. Daniel Lee for his invaluable guidance and support throughout this project. We thank the Ibero-American Action League and the Anthony Jordan Health Center for their support with survey collection and collaboration. Their contributions were vital to this project. We also deeply appreciate all participants for their time and insights.
Notes
This article was externally peer reviewed.
Funding: This study was internally funded by the University of Rochester Health Equity Program Services Office (HEPSO).
Conflict of interest: The authors declare no conflicting or competing interests.
- Received for publication April 1, 2025.
- Revision received July 27, 2025.
- Accepted for publication August 4, 2025.
References
Jump to section
Related Articles
Cited By...
- No citing articles found.