Improving Electronic Patient Portal Use in Groups with Low Utilization at a Family Medicine Practice

Citation:
Saif, N., Blackstone, S., Sebring, A & Compton, R., (2022). Improving electronic portal use in groups with low utilization at a family medicine practice.  Online Journal of Nursing Informatics (OJNI), 26 (2), https://www.himss.org/resources/online-journal-nursing-informatics

Abstract

Background: Electronic patient portals benefit health outcomes; yet, they are underutilized in certain patient sub-populations including adolescents, adults over age 65, racial/ethnic minorities, and non-English-speaking patients.

Purpose: This study describes a multi-faceted intervention to increase patient portal use and assesses the program’s effectiveness in these patient subgroups.

Method: An intervention was developed with three areas of focus: 1) patient education, 2) provider engagement and education, and 3) health information technology engagement. Baseline data were collected for the entire patient population and segmented by sub-populations of interest. The percentage of patients with an active patient portal were monitored at 12 months and 18 months.

Analysis: Logistic regression models examined the change in portal activation over time within sub-populations. Portalactivation increased significantly in all sub-populations examined (ranging from 5-12%), and changes in activation rate within sub-populations were consistent with other patient groups that were higher utilizers. Despite the significant increases, the disparities in portal activation remained.

Results: This study suggests that, while disparities were not reduced, a multifaceted intervention involving patient education, provider engagement and education, health information technology increased activation in patient groups that are low utilizers. This increased activation is an important first step toward reducing disparities in patient portal use, which can impact patients’ experiences and outcomes.

Patient portals are internet-based services that allow patients to view health records and appointment information and to submit requests for appointments and medication refills. The effective use of online patient portals through which patients can access their health information and communicate with their clinical team members has become increasingly important. The Centers for Medicaid and Medicare Services’ (n.d.) Medicare and Medicaid Promoting Interoperability Program Basics outline requirements for the use of electronic health records, a component of which is provider-to-patient exchange including patient electronic access to personal health information. Further, in the COVID-19 pandemic era during which telemedicine increased across many primary care practices, patient portals are instrumental in bridging care gaps. Patient portals are a potentially under-utilized resource with many actual and perceived barriers to its adoption ( Miller et al., 2016).

Patient portal use benefits include improved patient satisfaction, patient-provider communication, patient self-management, and quality of care (Miller et al., 2016; Osborn et al., 2010; Zhong et al., 2020). Improved chronic disease monitoring and overall outcomes associated with diabetes, cholesterol and blood pressure control have been demonstrated among patients using electronic portals (Sorondo et al., 2016; Zhong et al., 2020). Furthermore, patient portal use can increase the utilization of health services resources. Zhong et al., (2020), demonstrated that patients who used a portal for messaging and laboratory functions had lower rates of appointment no-shows compared to non-users. Another study that examined healthcare utilization outcomes following a patient portal education intervention demonstrated a 21% decrease in emergency department admissions (per 1,000) and a 38% decrease in hospital admissions (per 1,000) after seven months (Goel et al., 2011).

Despite the benefits of patient portal use, several disparities in activation and engagement are well-documented (elderly, racial and ethnic minorities, low health literacy, chronic illness, insurance status). Compared to non-Hispanic White patients, Latino and African-American patients are less likely to use a patient portal (Sarkar et al., 2011). The Health Information National Trends Survey suggested Hispanic and Black non-Hispanic patients are less likely to use patient portals and were less likely to be offered access to their patient portal (Clarke et al., 2021). Patients with limited English proficiency (LEP) also lagged behind in patient portal utilization (Ancker et al., 2011).

Older adults (aged 65+) are less likely to be willing or able to enroll in and use patient portals (Goel et al., 2011). Adults age 70+ are especially less likely to register for patient portals, own a digital device, use the Internet, or be willing to use an electronic platform for health management. Among older adults who do not use patient portals, common concerns are security of information, lack of personalization, and limited understanding of portal utility (Price-Haywood et al., 2017). Patient stakeholders have recommended strategies to increase portal use in older adults, including explaining to them what patients can do on portals, how to use portals, and why increasing portal use in older adults is useful (Price-Haywood et al., 2017). Another group with notable underutilization of patient portals is adolescents (Goldzweig et al., 2013; Olphert & Damodaran, 2013; Riippa et al., 2014). Despite adolescents being early adopters of technology, uptake of and engagement with patient portals has been challenging. In a feasibility assessment of a portal uptake intervention in adolescents, Ramsey et al., (2018), reported that most intervention participants felt there was a high need for access to the electronic health record through patient portals. Another study in a California detention center suggested 90% of 13-18 year olds were interested in accessing their health information (Irizarry et al., 2015). Despite this interest, rates of patient portal activation in youth are low.

Specific Aims

The transition to telemedicine for a large portion of patient care at the start of the COVID-19 pandemic placed an urgent emphasis on patient portal use and its benefits. Patient portals are helpful tools to assist patient-provider and patient-nursing communication, appointment scheduling, and the flow of telemedicine visits through delivery of secure links via messaging. The specific aim of this study was to assess the preliminary effectiveness of a practice-level, multi-faceted intervention – engaging patients, providers and information technology – on increasing the activation of patient portals among demographics known for underutilization (Black/African American, Hispanic, adults age 65+, adolescents aged 13-18, and non-English-speaking patients).  For patients under age 18, parents could activate the portal as a proxy and use the portal on the child’s behalf. Youth aged 13 and above were allowed confidential access to their patient portal and could grant parents proxy rights. The intervention was developed with the intention to benefit all patients at the practice; however, this study presents the assessment of effectiveness in underutilizers, comparing changes in portal activation status between these patients and the larger patient population. Understanding how practice-wide interventions impact underutilizers can provide beneficial information about whether targeted interventions are needed for sub-populations in the future.

Methods

Setting

The study location was a patient-centered medical home within a university health system. This clinic serves as the primary care home for all recently resettled refugees in the surrounding area through the International Rescue Committee (IRC). The clinic thus serves a diverse patient population with a large percentage of refugee patients whose first language is not English and whose primary source of coverage is financial assistance, charity care, or Medicaid. A patient’s portal status (active vs. inactive) can be obtained from a panel report generated in the electronic medical record (EMR).

Intervention Development

An intervention was developed with three areas of focus: 1) patient education, 2) provider engagement and education, and 3) health information technology engagement.

Patient Education

Given findings from the initial state data, the largest share of our efforts was dedicated to patient education and awareness about the portal.In the clinic exam rooms, patients were exposed to flyers providing an overview of portal benefits, a QR code linked to the online sign-up page, and a brief statement about proxy enrollment for parents and caregivers of patients. The team created the following educational videos, in collaboration with the pediatrics department portal committee and clinical information technology oversight committee: 1) overview of the portal and the various options for signing up, and 2) overview of proxy access to the portal and how to enroll. These videos were posted on the clinic’s patient resources webpage and incorporated into the slide show displayed in our waiting room. Both contained QR codes to the relevant online forms. Similar to the suggestions from Price-Hayward et al., (2017), to increase portal use in older adults, the team provided clear, real-life examples of how portal use is beneficial and demonstrated that it is easy to use. To encourage parent proxy registration, an educational video outlining the process for registration was included. Materials were also translated into Spanish.

Provider Engagement and Education

Baseline and follow-up data and project updates were presented to faculty, residents and nursing staff through family medicine department meetings and resident conferences. The goal was to increase commitment and dedication to this important issue among department staff, as well as to provide education on the technical processes of portal administration (accessing the portal administration page in the EMR to view patient’s status, assist with enrollment, and enroll in proxy access). Providers were regularly encouraged to discuss patient portal use with patients who were not active. Providers were easily able to see a patient’s portal status through an icon in their chart.

Engaging Information Technology

The team leveraged a feature of the EMR (smart phrases) as part of the intervention. A smart phrase is a tool in the EMR that allows providers to complete patient documentation or populate patient instructions faster (also known as dot phrases). Smart phrases allow commonly used chunks of text to be inserted in patient notes, communications and instructions via a short line of text (phrase) so the provider does not have to retype the message for each patient. Two smart phrases were created to help enroll patients in portal. One detailed the process of portal proxy access that would be used for patient instructions for well child visits. The second provided instructions for adult patients to enroll in the portal. These smart phrases were reviewed at department meetings and resident conferences, along with strategies to overcome language barriers to portal use by using browser translation services as a potential tool for some patients (i.e., Google Translate).

Measures

The patient population was defined as patients who had been seen in the last three years in clinic (active patients) and had a clinic provider listed as their primary care provider. Data were extracted from the EMR at three time points: November 2019, November 2020 and May 2021. The panel report of active patients included an indicator of each patient’s portal status. This was coded as “Active,” “Pending Activation,” and “Non-Active” for descriptive analyses, and dichotomized (Active vs. Non-Active) for the regression models. Analyses were conducted within each population of underutilizers to examine changes in patient portal activation status. Additional information extracted included patient sex (male or female), insurance status (Medicaid, Medicare, financial assistance, no insurance), language (English, non-English), refugee status (refugee, non-refugee), and month of visit.

The study period was from November 2019 to May 2021. The study team regularly ran EMR reports to obtain the proportion of portal active patients and their characteristics. We selected three time points (baseline, 12 months and 18 months) to conduct a time-series analysis to assess change in portal activation rate over the study period.

Statistical Analysis

Frequencies and proportions were used to describe the patient panel at all three time points. See Table 1 for a breakdown of the total sample at each time point. To evaluate the effect of the initiative on portal use, we conducted a logistic regression to examine the change in odds of having an active portal over time while controlling for covariates outlined in the measures section. A value of p < .05 was used to denote statistical significance. Analyses were conducted using R.4.1.

Results

Baseline Assessment

Table 1 shows the patient population breakdown at baseline and follow-up. Table 2 shows the proportion of patients with active portals. Patient sub-populations of specific focus are bolded in Table 2. Fourteen percent of 13-18 year olds and 45.4% of patients aged 65+ had an active portal. Both 13-18 year-olds (OR = 0.12, 95% CI, 0.11, 0.15) and patients 65+ (OR = 0.49, 95% CI, 0.45, 0.52) were less likely to have an active portal compared to 19-44 year-olds (reference group). Hispanic patients were less likely to have an active portal compared to non-Hispanic patients (OR = 0.32, 95% CI, 0.29, 0.34), with 27.1% having an activated portal. African Americans (40.9%) were less likely than White or Caucasian patients to have an active portal (OR = 0.35, 95% CI, 0.33, 0.37). Finally, 19.9% of non-English speakers had an active portal, significantly less than English speakers (OR = 0.17, 95% CI, 0.16, 0.18).

Post-intervention Assessment
Between baseline and follow-up portal activation increased by 11% in adolescents age 13-18 and 10.7% in adults 65+. This was slightly less than the activation increase of 12% among 19-44 year old patients. Non-English-speaking patients increased portal activation by 10%, consistent with the increase in English speakers (10%). Portal activation in African-American patients increased by 11%, consistent with the increase in White or Caucasian patients (11%). Hispanic patients’ portal activation increased by 4.8% compared to an increase in non-Hispanic patients by 6.4%. While portal activation increased at a similar rate, the underutilizer patient groups had lower rates of portal activation at follow-up.

Adjusted analyses demonstrated that in all patient sub-populations, the odds of having an active portal increased between baseline and 12 months (aORs ranging from 1.22 – 1.81) and between baseline and 18 months (aORs ranging from 1.44-2.25). See Table 3 for aOR and confidence intervals. In African-American and Hispanic patients, respectively, male (aOR = 0.44, 95% CI, 0.40, 0.49; aOR = 0.52, 95% CI, 0.42, 0.64) and refugee patients (aOR = 0.37, 0.26, 0.51; aOR = 0.19, 95% CI, 0.07, 0.44) had lower odds of having active portals. In addition, patients on financial assistance, Medicare, Medicaid or without insurance were significantly less likely to have an active portal compared to patients with private insurance.

Discussion

Overall, our results demonstrate that a multi-faceted intervention involving patients, providers and information technology was effective at increasing patient portal activation among groups of patients that are typically low utilizers. However, compared to a referent population in each respective group, rates of patient portal activation in teens, seniors, African Americans, Hispanics and non-English speakers remained substantially lower. Additional disparities noted across sub-populations indicated that males, refugee patients and individuals with financial assistance, Medicare/Medicaid or no insurance were less likely to have an active patient portal.

There are many reasons why teens, seniors, African-American, Hispanic and non-English-speaking patients continue to lag in patient portal activation. While teens tend to be more amenable to using technology, many health systems struggle to increase adolescent enrollment. It is not clear whether this struggle is due to confidentiality concerns, the registration process, or lack of interest and access to internet. Ramsey et al., (2018), found that most of their teen participants in an urban, academic primary care setting were interested in accessing their patient portal and did not have access limitations; however, it is unclear if these findings are true in other settings. Adolescents may have concerns over parent proxy access to their medical records; the present intervention included patient education on portal proxy access for pediatric patients. However, proxy access may be a topic that is not well understood and requires additional attention. Older adults may be less inclined to use a patient portal if they believe process of registering for and using the portal is cumbersome or if there are concerns about data security (Price-Haywood et al., 2017). Furthermore, electronic health literacy is strongly related to patient portal use and is typically low in older adults (Price-Haywood et al., 2017).

Racial disparities in patient portal use are well documented (Ancker et al., 2011; Clarke et al., 2021) and may arise for a variety of reasons. Non-White patients may prefer speaking with their provider in person and see little benefit to using an electronic portal. In addition, disparities in digital literacy and education may contribute to challenges during the registration process (Goel et al., 2011). In our intervention, we tried to be mindful of varying literacy levels and provided support resources (see Figure 1 for examples); however, these resources may not have reached all patients. Finally, there may be greater distrust of web-based communication for health information among non-white patients (Clarke et al., 2021; Hillen et al., 2011); this distrust has been consistent over the last decade, limiting their willingness to activate the patient portal. Finally, consistent with other research (Ancker et al., 2011) our non-English-speaking patients had lower portal activation compared to patients whose primary language was English. While portal use in LEP patients lagged behind, utilization over the study period increased 10%, a similar amount to English-speaking patients. This suggests the LEP patients were benefiting from the intervention at similar rates to English-speaking patients. Continued patient and provider education may lead to further increase over time.

While we still saw disparities in portal activation, the significant increase in activation among low-utilization groups suggests multifaceted interventions applied in primary care clinics can improve patient portal use. Our study illustrates examples of several clinic- and organization-level interventions to increase patient portal use, in contrast to the body of current literature, which mostly focuses on individual-level interventions to increase portal use in vulnerable populations (Grossman et al., 2019). Our focus on provider education and engagement on this issue is also notable, as previous studies demonstrate the importance of provider-patient education on portal use in vulnerable groups (Kim et al., 2018).  An important next step will be to address potential barriers to access patient education through the QR code. This can be a challenge for patients with limited digital literacy or without a compatible device. Engaging patient navigators to help patients scan the QR code and providing devices in clinic for those without a smartphone are potential strategies to mitigate this concern.

A limitation of this study is that we used the presence of an active portal account as a proxy to represent portal use; this use may over-represent true patient portal use in our population (Lafata et al., 2018). In addition, we were unable to control for factors known to affect portal activation and use, such as education and digital literacy. However, these results still suggest that clinic-wide interventions can benefit underutilizers of patient portals similarly to other patient groups.

Conclusions

Universal patient portal enrollment should remain a consistent priority of our patient access staff. Future steps could include setting up tablets or computer stations for patients to enroll and log into the portal. These steps could be useful in overcoming technology barriers for patients. Another potential intervention is to train medical students or other willing groups to serve as “portal ambassadors” to assist with in-person portal education and sign-up assistance in our clinic waiting room. Other teams in the health system are also working on the technology access issue by investigating the utilization of public libraries to meet patient access needs. Continuing to engage with our EMR specialists to ensure future versions of the portal take into account accessibility factors highlighted in our study remains a priority. A forthcoming version of the portal will include a Spanish version, increasing language accessibility. Other factors that should be considered include improving readability for lower education levels and improving accessibility for elderly patients and those with less experience navigating the internet. Developing strategies to address these barriers to portal use may be helpful in reducing the disparities in patient portal use that are well documented in the literature and supported by this study.