OBJECTIVES

To determine adolescent characteristics associated with patient portal secure messaging use within a health system.

METHODS

This study analyzed monthly data from individuals aged 13 to 17 who met study eligibility criteria from 2019 to 2021. The primary outcome was any secure messages sent from an adolescent’s account during each observed month. Unadjusted and adjusted associations between adolescent characteristics and secure messaging use were assessed using generalized estimating equations with log link and binomial variance.

RESULTS

Of 667 678 observed months, 50.8% occurred among males who were not transgender, 51.5% among those identifying as non-Hispanic white, and 83.3% among the privately insured. The adjusted relative risks of secure messaging use were significantly higher for individuals with female sex and transgender identities (female sex, not transgender: adjusted relative risk [aRR] 1.41, 95% confidence interval [CI] 1.31–1.52; male sex, transgender: aRR 2.39, CI 1.98–2.90, female sex, transgender: aRR 3.01, 95% CI 2.63–3.46; referent male sex, not transgender), those with prior portal use (aRR 22.06, 95% CI 20.48–23.77; referent no use) and those with a recent preventive care visit (aRR 1.09, 95% CI 1.02–1.16; referent no recent visits). The adjusted relative risks of portal secure messaging use were significantly lower among those with public insurance (aRR 0.58, 95% CI 0.50–0.67; referent private).

CONCLUSIONS

Adolescents who sent patient portal secure messages differed from those who did not. Interventions to encourage secure messaging use may require tailoring based on patient characteristics.

What’s Known on This Subject:

Health systems increasingly offer the option for adolescents to establish and manage their own patient portal accounts. Little is known about characteristics of adolescents who use a patient portal for secure messaging with their health care team.

What This Study Adds:

An analysis of electronic health record data in 1 health system from 2019 to 2021 found that adolescents who used a patient portal for secure messaging significantly differed from those who did not.

The landscape of health care communication has shifted dramatically over the past several decades with the emergence of secure messaging via online patient portals. Patient engagement enabled by secure messaging was further catalyzed by the coronavirus disease 2019 (COVID-19) pandemic when health systems rapidly scaled up more widespread access to the patient portal to facilitate telehealth access and virtual care communication.1,2  Secure messaging is a key indicator of meaningful online patient portal use in health care, although online patient portals can have a much broader range of functionalities, including scheduling appointments, viewing notes and care summaries, and requesting medication refills.3 

Patient-provider communication via secure messaging has been shown to improve patient satisfaction and may improve treatment adherence and health outcomes among adult populations.4  Although some evidence supports the feasibility and acceptability of secure messaging use in pediatric care,5  little is understood about how portal use by parent users of pediatric patients or by adolescent patients themselves impacts health outcomes. Limited qualitative research indicates that teens are enthusiastic about using secure messaging to communicate with providers, seek health advice, and make appointments.6  However, lack of experience as independent health care consumers and concerns about confidentiality are known barriers to teen digital health communication, and teens do not use patient portal accounts as frequently as adults.7 

As more digital tools for care communication become available to adolescents as part of mainstream health care delivery, it is critical to understand which adolescents are most likely to use secure messaging to address their health care needs and how their needs differ from adults. This will allow health systems to begin understanding disparities in adolescent secure messaging use and opportunities for engagement via the patient portal as adolescents progress toward independently managing their health care needs. Designing better secure messaging experiences for adolescents, their families, and care teams has the potential to offer a convenient, trusted source of care for adolescents and address their unique unmet health care needs.8 

The aim of the current study was to determine patient characteristics associated with secure messaging in a large sample of adolescents receiving care in an integrated primary care-based health system. Based on prior studies largely conducted among adults,4  we hypothesized that secure messaging was more likely to increase over time and occur among users with older age, female gender, white, non-Hispanic identity, private insurance, and having a parent or guardian with a proxy portal account. In this study, we aimed to identify independent predictors of patient portal use for secure messaging that could inform future research on adolescent health care utilization impacted by patient portal use. We also hoped to recognize populations of adolescents less likely to use the portal, gaining insights about systemic barriers to adolescent portal use for secure messaging that would inform interventions to promote equity in digital care communication.

This study describes adolescent patient characteristics associated with online patient portal use for secure messaging at Kaiser Permanente Washington (KPWA), a not-for-profit, integrated health care delivery system serving over 300 000 members in Washington State. Members are insured through employer-based and individual plans, Medicare, and Medicaid. Adolescents receive care in 32 KPWA-owned facilities staffed by over 1000 KPWA primary care and specialty clinicians. This study was the product of a research-health system partnership between KPWA and KPWA Health Research Institute and was approved by the KPWA Institutional Review Board.

KPWA and its predecessor, Group Health, have offered an online patient portal (MyChart) to members since 2003 – initially available for those over age 18 and parents for children aged 12 and younger. In 2017, 2 features were released for parents of teens aged 13 to 17 to allow for messaging with providers and viewing immunization records. Adolescent members aged 13 to 17 were not eligible to initiate and manage their own MyChart accounts until January 2019 after KPWA implemented systems to ensure confidentiality of services protected by state minor consent laws. At that time, KPWA members aged 13 to 17 and their parents and caregivers became eligible for “teen accounts” and “teen proxy accounts,” respectively – separate accounts with limited functionality to send and receive messages from select health care providers (including their primary care team and select specialists), request refills for medications not deemed confidential by the patient, and view immunization records (excluding human papillomavirus vaccine to limit proxy view of any confidential administrations). To avoid breaches of confidentiality, teen and teen proxy account users cannot view and initiate messages to their current providers in Mental Health and Wellness, Obstetrics and Gynecology, and Adolescent Medicine departments, but can reply to messages sent at the discretion of the provider in these departments.

Teen members must actively register for a MyChart account. This process has evolved iteratively since 2019 to reduce barriers to registration. When first available, teens were required to complete a brief registration online, then verify their identity in person at any clinical facility to receive a password or elect to have a temporary password mailed to their primary household address before the account could be activated. After the onset of the COVID-19 pandemic, a phone-based activation process was made available as an alternative to in-person or mailed password. In 2020, teen and proxy MyChart accounts also added the functionality to join video visits directly from MyChart.

The study population met the following eligibility criteria: age 13 to 17 years, with at least 3 months of continuous enrollment in a KPWA or Medicaid plan (with 31 days allowable administrative gap) and having either had at least 1 clinical visit at KPWA in primary or specialty care in the past 36 months or being paneled with a KPWA provider during the month of observation. We excluded individuals likely to have severe intellectual disability limiting their ability to communicate independently with their health care team. These individuals were identified using a modification of the Pediatric Medical Complexity Algorithm with at least 1 body system involved being neurologic.9  The study population was designed to represent all other adolescents receiving care within our health system.

An analytic dataset was created using deidentified data from electronic health records assessed on the first day of each month from January 1, 2019 to December 31, 2021. Analyses were conducted using month of individual patient characteristics as the unit of observation. Monthly observations allow for more specificity of age at time of measurement within the dynamic adolescent period and also allow values for selected variables to vary with time, such as age, year, insurance status, parent proxy, and having had a recent preventive health visit.10 

Characteristics of each unit of observation include calendar month, year, patient age during the month, most recent administratively recorded sex, gender identity, self-reported racial and ethnic identity, insurance status during the person-month (Kaiser Permanente Health Maintenance Organization [HMO], preferred provider organization [PPO], or Medicaid), presence of 1 or more active parent and proxy accounts linked to teen record, whether a preventive care visit had occurred in the prior 18 months, and counts of secure messages sent via teen MyChart account during the month. Based on our sample definition, unique adolescents meeting inclusion criteria could potentially contribute a maximum of 36 months of observation if they were continuously enrolled in the 3 months before January 1, 2019 and remained enrolled with KPWA until December 31, 2021.

Our primary outcome of interest was portal use for secure messaging, a binary indicator of no secure messages versus any messages sent from the adolescent patient account during each month of observation. We chose to use sent messages rather than sent and received messages to measure the most proactive measure of patient engagement. We did not include messages sent from any parent proxy account.

We included the following independent variables in our analysis (categorical unless noted): calendar year (2019, 2020, 2021), age in years (continuous), gender defined using administrative sex and an optional gender identity field (female sex, not transgender or gender diverse [TGD]; male sex, not TGD; female sex, TGD; male sex, TGD); racial and ethnic identity (nonmutually exclusive options in administrative data: Asian, Black or African American, Pacific Islander or Native Hawaiian, American Indian or Alaska Native, Hispanic, more than 1 race selected, white non-Hispanic only, other (not Hispanic), or no race or ethnicity recorded); insurance status (private HMO/PPO or Medicaid); parent proxy account (active or inactive/absent); online portal use for secure messaging before index month (prior use or no prior use); and preventive care visit in past 18 months. We chose to not collapse race and ethnicity variables into mutually exclusive categories because the data collected from patients does not allow them to self-select into mutually exclusive categories, and we did not want to incorrectly assign multiracial individuals.11,12 

Descriptive Characteristics

We summarized characteristics in the overall sample and within subgroups defined by outcome status of secure messaging. We calculated descriptive P values quantifying the univariate association between baseline characteristics and the outcome using χ-square tests.

Unadjusted and Adjusted Analyses

To examine the association between baseline characteristics and portal use for secure messaging, we used generalized estimating equations (GEE) with a log link and binomial variance to estimate unadjusted and adjusted marginal relative risks (RRs) and 95% confidence intervals (95% CI).13  The adjusted GEE model included all baseline characteristics of interest. RRs were reported for all covariates. Estimated RRs were interpreted as the association between the independent variable of interest (e.g., age) and the dependent variable (secure messaging during the month) adjusted for all other baseline characteristics. The GEE model adjusted for correlation of monthly observations within an individual via robust sandwich variance estimates (to ensure robustness to any type of correlation structure) obtained by assuming a working independence correlation structure.

The study sample included 37 818 unique individuals who contributed 667 678 months of individual observation (Fig 1). Table 1 displays descriptive characteristics of all months of individual observation included in the analysis and χ-square comparisons by patient portal secure messaging use. The total number of months were evenly distributed across years and age categories. The majority were months occurring among adolescents with male sex, not TGD (50.8%), non-Hispanic white racial identity (51.5%), and private insurance (83.3%). Fewer than 1% of observed months in 2019 showed any secure messaging use. Use increased to 6% of observed months in 2020 and over 8% in 2021 (data not shown). Figure 2 shows the proportion of months of observation by quarter with any secure messaging use, indicating marked increases over each quarter in 2020 and a leveling off in 2021.

FIGURE 1

Flow diagram to develop analytic dataset. OM, observed months; PCP, primary care provider.

FIGURE 1

Flow diagram to develop analytic dataset. OM, observed months; PCP, primary care provider.

Close modal
TABLE 1

Descriptive Characteristics of Patients and Observed Months Included in Our Analysis With Number, Percentage, and χ-Square Comparisons by Patient Portal Secure Messaging Use

Unique PatientsaTotal Months Observed During Study PeriodNo Patient Portal Secure Messaging Use During MonthAny Patient Portal Secure Messaging Use During MonthP*
NN%N%N%
Calendar year        <.001 
 2019 24 151 216 187 32.38 215 821 32.71 366 4.70  
 2020 24 883 225 521 33.78 222 570 33.73 2951 37.86  
 2021 24 715 225 970 33.84 221 492 33.57 4478 57.45  
Age, y        <.001 
 13 16 265 133 407 19.98 132 848 20.13 559 7.17  
 14 16 354 134 610 20.16 133 486 20.23 1124 14.42  
 15 16 595 134 990 20.22 133 506 20.23 1484 19.04  
 16 16 819 137 658 20.62 135 531 20.54 2127 27.29  
 17 16 074 127 013 19.02 124 512 18.87 2501 32.08  
Genderb        <.001 
 Male sex, not TGD 19 206 339 340 50.82 336 771 51.03 2569 32.96  
 Female sex, not TGD 18 081 317 852 47.61 313 405 47.49 4447 57.05  
 Male sex, TGD 113 2277 0.34 2132 0.32 145 1.86  
 Female sex, TGD 418 8209 1.23 7575 1.15 634 8.13  
Race and ethnicityc         
 Asian 5552 105 385 15.78 104 153 15.78 1232 15.81 .96 
 Black or African American 3817 71 105 10.65 70 545 10.69 560 7.18 <.001 
 Pacific Islander and Native Hawaiian 958 17 856 2.67 17 733 2.69 123 1.58 <.001 
 American Indian and Alaska Native 735 12 999 1.95 12 841 1.95 158 2.03 .61 
 Hispanic 3256 58 771 8.80 58 162 8.81 609 7.81 .002 
 White non-Hispanic only 18 621 343 707 51.48 338 846 51.35 4861 62.36 <.001 
 More than 1 raced 3483 65 753 9.85 64 887 9.83 866 11.11 <.001 
 Other (not Hispanic) 894 16 481 2.47 16 249 2.46 232 2.98 .004 
 No race or ethnicity recorded 5370 66 615 9.98 66 297 10.05 318 4.08 <.001 
Insurance status        <.001 
 Private (HMO or PPO) 31 771 556 220 83.31 548 909 83.18 7311 93.79  
 Medicaid 6456 111 458 16.69 110 974 16.82 484 6.21  
Preventive care visit in past 18 mo        <.001 
 No 30 690 344 812 51.64 341 211 51.71 3601 46.20  
 Yes 23 484 322 866 48.36 318 672 48.29 4194 53.80  
Prior portal use for secure messaging        <.001 
 No 37 818 641 417 96.07 638 358 96.74 3059 39.24  
 Yes 2850 26 261 3.93 21 525 3.26 4736 60.76  
Parent proxy account active        <.001 
 No proxy account, or not active 27 229 429 124 64.27 424 554 64.34 4570 58.63  
 Yes 13 160 238 554 35.73 235 329 35.66 3225 41.37  
Total 37 818 667 678  659 883  7795   
Unique PatientsaTotal Months Observed During Study PeriodNo Patient Portal Secure Messaging Use During MonthAny Patient Portal Secure Messaging Use During MonthP*
NN%N%N%
Calendar year        <.001 
 2019 24 151 216 187 32.38 215 821 32.71 366 4.70  
 2020 24 883 225 521 33.78 222 570 33.73 2951 37.86  
 2021 24 715 225 970 33.84 221 492 33.57 4478 57.45  
Age, y        <.001 
 13 16 265 133 407 19.98 132 848 20.13 559 7.17  
 14 16 354 134 610 20.16 133 486 20.23 1124 14.42  
 15 16 595 134 990 20.22 133 506 20.23 1484 19.04  
 16 16 819 137 658 20.62 135 531 20.54 2127 27.29  
 17 16 074 127 013 19.02 124 512 18.87 2501 32.08  
Genderb        <.001 
 Male sex, not TGD 19 206 339 340 50.82 336 771 51.03 2569 32.96  
 Female sex, not TGD 18 081 317 852 47.61 313 405 47.49 4447 57.05  
 Male sex, TGD 113 2277 0.34 2132 0.32 145 1.86  
 Female sex, TGD 418 8209 1.23 7575 1.15 634 8.13  
Race and ethnicityc         
 Asian 5552 105 385 15.78 104 153 15.78 1232 15.81 .96 
 Black or African American 3817 71 105 10.65 70 545 10.69 560 7.18 <.001 
 Pacific Islander and Native Hawaiian 958 17 856 2.67 17 733 2.69 123 1.58 <.001 
 American Indian and Alaska Native 735 12 999 1.95 12 841 1.95 158 2.03 .61 
 Hispanic 3256 58 771 8.80 58 162 8.81 609 7.81 .002 
 White non-Hispanic only 18 621 343 707 51.48 338 846 51.35 4861 62.36 <.001 
 More than 1 raced 3483 65 753 9.85 64 887 9.83 866 11.11 <.001 
 Other (not Hispanic) 894 16 481 2.47 16 249 2.46 232 2.98 .004 
 No race or ethnicity recorded 5370 66 615 9.98 66 297 10.05 318 4.08 <.001 
Insurance status        <.001 
 Private (HMO or PPO) 31 771 556 220 83.31 548 909 83.18 7311 93.79  
 Medicaid 6456 111 458 16.69 110 974 16.82 484 6.21  
Preventive care visit in past 18 mo        <.001 
 No 30 690 344 812 51.64 341 211 51.71 3601 46.20  
 Yes 23 484 322 866 48.36 318 672 48.29 4194 53.80  
Prior portal use for secure messaging        <.001 
 No 37 818 641 417 96.07 638 358 96.74 3059 39.24  
 Yes 2850 26 261 3.93 21 525 3.26 4736 60.76  
Parent proxy account active        <.001 
 No proxy account, or not active 27 229 429 124 64.27 424 554 64.34 4570 58.63  
 Yes 13 160 238 554 35.73 235 329 35.66 3225 41.37  
Total 37 818 667 678  659 883  7795   
a

These values indicate the total number of unique patients who contributed patient-time to each category.

b

Gender defined using administrative sex and an optional gender identity field in the electronic health record.

c

Race and ethnicity categories are not mutually exclusive.

d

More than 1 race and ethnicity category was reported. These observations are also counted in individual race and ethnicity categories, so percentages will add up to more than 100.

*

χ-Square P value comparing characteristics across observed months with and without use.

FIGURE 2

Percentage of eligible teen users with patient portal use for secure messaging, by quarter, 2019 to 2021. 95% confidence interval error bars displayed for each percentage. Bars above each year label represent quarters 1 (January–March), 2 (April–June), 3 (July–September), and 4 (October–December) of each respective year.

FIGURE 2

Percentage of eligible teen users with patient portal use for secure messaging, by quarter, 2019 to 2021. 95% confidence interval error bars displayed for each percentage. Bars above each year label represent quarters 1 (January–March), 2 (April–June), 3 (July–September), and 4 (October–December) of each respective year.

Close modal

During observed study months, a higher proportion of the portal secure message users were older, female sex, TGD, and privately insured compared with nonusers (Table 1). Higher proportions of secure message users had prior online portal use for secure messaging and an active parent proxy account compared with nonusers (Table 1).

The unadjusted and adjusted regression analyses are presented in Table 2. For each additional year of age, adolescents had a 1.14 higher adjusted relative risk of secure messaging use (95% CI 1.11–1.17). Individuals who were female sex and TGD-identifying had higher adjusted relative risks of portal secure messaging use compared with those who were male sex, not TGD-identifying (aRR 3.01, 95% CI 2.63–3.46 for female sex, TGD; aRR 2.39, 95% CI 1.98–2.90 for male sex, TGD; aRR 1.41, 95% CI 1.31–1.52 for female sex, not TGD). Those who identified as Asian, Black or African American, Pacific Islander and Native Hawaiian, those of Hispanic ethnicity, and those with no race or ethnicity recorded had lower adjusted relative risks of portal secure messaging use compared with those identifying as white, non-Hispanic (Asian: aRR 0.88, 95% CI 0.79–0.98; Black or African American: aRR 0.75, 95% CI 0.65–0.87; Pacific Islander and Native Hawaiian: aRR 0.75, 95% CI 0.56–1.00; Hispanic: aRR 0.88, 95% CI 0.77–1.00). Individuals who identified as having more than 1 race had a higher adjusted relative risk of portal secure messaging use compared with those identifying as white, non-Hispanic (aRR 1.16, 95% CI 1.01–1.33).

TABLE 2

Unadjusted and Adjusted Regression Analyses of Patient Characteristics Associated With Secure Messaging

UnadjustedAdjusteda
RR (95% CI)RR (95% CI)
Calendar year   
 2019 1.00 (ref) 1.00 (ref) 
 2020 7.73 (6.62–9.03)*** 4.56 (3.89–5.36)*** 
 2021 11.71 (9.95–13.77)*** 3.96 (3.35–4.68)*** 
Age, continuous, years 1.41 (1.36–1.45)*** 1.14 (1.11–1.17)*** 
Gendera,b   
 Male sex, not TGD 1.00 (ref) 1.00 (ref) 
 Female sex, not TGD 1.85 (1.67–2.04)*** 1.41 (1.31–1.52)*** 
 Male sex, TGD 8.41 (6.00–11.79)*** 2.39 (1.98–2.90)*** 
 Female sex, TGD 10.20 (8.27–12.58)*** 3.01 (2.63–3.46)*** 
Race and ethnicityb   
 Asian 0.76 (0.65–0.88)*** 0.88 (0.79–0.98)* 
 Black or African American 0.49 (0.40–0.60)*** 0.75 (0.65–0.87)*** 
 Pacific Islander and Native Hawaiian 0.44 (0.30–0.64)*** 0.75 (0.56–1.00)* 
 American Indian and Alaska Native 0.76 (0.55–1.03) 0.86 (0.67–1.09) 
 Hispanic 0.77 (0.64–0.92)** 0.88 (0.77–1.00)* 
 White non-Hispanic only 1.00 (ref) 1.00 (ref) 
 More than 1 racec 1.62 (1.34–1.97)*** 1.16 (1.01–1.33)* 
 Other (not Hispanic) 0.97 (0.73–1.29) 1.11 (0.91–1.35) 
 No race or ethnicity recorded 0.34 (0.28–0.43)*** 0.57 (0.48–0.67)*** 
Insurance status   
 Private (HMO or PPO) 1.00 (ref) 1.00 (ref) 
 Medicaid 0.33 (0.28–0.40)*** 0.58 (0.50–0.67)*** 
Preventive care visit in past 18 mo   
 No 1.00 (ref) 1.00 (ref) 
 Yes 1.24 (1.15–1.35)*** 1.09 (1.02–1.16)** 
Prior portal use for secure messaging   
 No 1.00 (ref) 1.00 (ref) 
 Yes 37.81 (35.62–40.15)*** 22.06 (20.48–23.77)*** 
Parent proxy account active   
 No proxy account or not active 1.00 (ref) 1.00 (ref) 
 Yes 1.27 (1.16–1.39)*** 0.91 (0.85–0.97)** 
UnadjustedAdjusteda
RR (95% CI)RR (95% CI)
Calendar year   
 2019 1.00 (ref) 1.00 (ref) 
 2020 7.73 (6.62–9.03)*** 4.56 (3.89–5.36)*** 
 2021 11.71 (9.95–13.77)*** 3.96 (3.35–4.68)*** 
Age, continuous, years 1.41 (1.36–1.45)*** 1.14 (1.11–1.17)*** 
Gendera,b   
 Male sex, not TGD 1.00 (ref) 1.00 (ref) 
 Female sex, not TGD 1.85 (1.67–2.04)*** 1.41 (1.31–1.52)*** 
 Male sex, TGD 8.41 (6.00–11.79)*** 2.39 (1.98–2.90)*** 
 Female sex, TGD 10.20 (8.27–12.58)*** 3.01 (2.63–3.46)*** 
Race and ethnicityb   
 Asian 0.76 (0.65–0.88)*** 0.88 (0.79–0.98)* 
 Black or African American 0.49 (0.40–0.60)*** 0.75 (0.65–0.87)*** 
 Pacific Islander and Native Hawaiian 0.44 (0.30–0.64)*** 0.75 (0.56–1.00)* 
 American Indian and Alaska Native 0.76 (0.55–1.03) 0.86 (0.67–1.09) 
 Hispanic 0.77 (0.64–0.92)** 0.88 (0.77–1.00)* 
 White non-Hispanic only 1.00 (ref) 1.00 (ref) 
 More than 1 racec 1.62 (1.34–1.97)*** 1.16 (1.01–1.33)* 
 Other (not Hispanic) 0.97 (0.73–1.29) 1.11 (0.91–1.35) 
 No race or ethnicity recorded 0.34 (0.28–0.43)*** 0.57 (0.48–0.67)*** 
Insurance status   
 Private (HMO or PPO) 1.00 (ref) 1.00 (ref) 
 Medicaid 0.33 (0.28–0.40)*** 0.58 (0.50–0.67)*** 
Preventive care visit in past 18 mo   
 No 1.00 (ref) 1.00 (ref) 
 Yes 1.24 (1.15–1.35)*** 1.09 (1.02–1.16)** 
Prior portal use for secure messaging   
 No 1.00 (ref) 1.00 (ref) 
 Yes 37.81 (35.62–40.15)*** 22.06 (20.48–23.77)*** 
Parent proxy account active   
 No proxy account or not active 1.00 (ref) 1.00 (ref) 
 Yes 1.27 (1.16–1.39)*** 0.91 (0.85–0.97)** 
a

Adjusted models included calendar year, age, gender, race and ethnicity, insurance status, preventive care visit in past 18 months, prior portal use for secure messaging and parent proxy account active.

b

Race and ethnicity categories are not mutually exclusive.

c

More than 1 race and ethnicity category was reported.

*

P < .05;

**

P < .01;

***

P < .001.

Compared with adolescents with private insurance, those with Medicaid had a lower adjusted relative risk of portal secure messaging use (aRR 0.58, 95% CI 0.50–0.67). Adolescents who had a history of prior secure messaging or a recent (within the last 18 months) preventive care visit had higher adjusted relative risks of portal secure messaging use during the observed months compared with those who had no prior portal use or no recent preventive visits (prior secure messaging: aRR 22.06, 95% CI 20.48–23.77; recent preventive care visit: aRR 1.09, 95% CI 1.02–1.16). Compared with adolescents without an active parent proxy account, those with an account had a lower risk of portal secure messaging use during the observed months (aRR 0.91, 95% CI 0.85–0.97).

This study examined a large population of adolescents in a single health system to understand characteristics associated with patient portal secure messaging use. Overall, a small proportion of eligible adolescents are using the patient portal for secure messaging, though rates increased significantly from 2019 to 2021. Consistent with our hypotheses, we found meaningful differences in adolescent secure messaging use based on several characteristics. Specifically, we found significant associations between portal secure messaging use and older adolescent age, female sex and TGD identities, and being privately insured. We also found that portal secure messaging use was associated with a history of prior portal use and recent preventive care utilization. Finally, we found that less secure messaging use was associated with minoritized racial and ethnic identities (except for multiracial youth) and having parent proxy accounts. Importantly, these differences can be used to inform future intervention development to encourage adolescent use of these digital platforms for confidential communication.

Reasons for low use may be attributed to low awareness of the patient portal, limited experience navigating health care independently, or concerns about confidentiality.14  A 2016 study in Florida found that only 3% of adolescents receiving care in 1 health system had ever used the portal during multiple years.15  A more recent 2021 study of more than 3000 eligible adolescent accounts across 3 institutions found that 64% to 75% of adolescent portal accounts with outbound messages were accessed by guardians.7  We were not able to distinguish whether parent and guardians were using adolescent patient portal accounts to send messages to the care team. Our finding that those with an active parent proxy account had a lower adjusted risk of portal secure messaging use suggests that parent use of adolescent accounts may be occurring if the parent does not have their own proxy account established. Alternatively, teens who know their parents have an account may be less confident that any communications between teens and providers will remain confidential.

Prior secure messaging use being strongly associated with portal use during the index month suggests that online patient portal use is a reinforcing behavior as adolescents gain more experience with the platform. Having had a preventive health visit in the past 18 months was also associated with portal use, a finding that is more difficult to interpret in the context of this study design. Patients may have been familiarized to the patient portal during a recent preventive health visit, or patients who demonstrate engagement via the portal may also be more likely to be up to date in their preventive care because of a common characteristic of overall health care engagement that we were not able to measure.

We identified significant differences in secure messaging use among adolescents of minoritized racial and ethnic identities that persisted after adjusting for factors known to contribute to disparities, such as insurance status and use of preventive care services. Interpreting these results first requires an acknowledgment that racial and ethnic identities are social constructs and likely serve as a proxy measure of lived experiences of racism and structural oppression that may impact an individual’s health communication, behaviors, and access to health services.16  Our finding that those individuals who identify as Asian, Black or African American, Hispanic, and Native Hawaiian or Pacific Islander demonstrate a lower risk of patient portal secure messaging use is consistent with prior studies among adults and pediatric patients.4,5,17  Taken together, these findings signal that unique, equity-informed research and interventions are essential to align health system digital health communication approaches with patient and community needs and remove barriers disproportionately faced by minoritized communities. Ignoring differences by racial and ethnic identities risks developing interventions that further exacerbate health disparities.

Gender differences in patient portal use have been previously reported in adult studies.1820  However, our finding that individuals identifying as transgender had the highest relative risk of portal use is notable. Although we did not include type of care received as a covariate, patients receiving gender-affirming care in our health system may have more support and opportunities to engage in portal-based care communication regarding their gender healthcare. This phenomenon should be explored in future studies.

The increasing secure messaging use we observed over time (2019–2021) must be placed in the context of the COVID-19 pandemic. This not only impacted overall health care utilization among adolescents21,22  but also resulted in a prioritization of virtual care encounters to minimize exposure risks of in-person care. We included time as a covariate in our analysis to account for the evolving trends in MyChart utilization that may have been influenced by the pandemic’s impact on healthcare utilization. Given that we did not capture the reason for secure message use, we are unable to explore whether messaging about COVID-19 specifically impacted the findings. We can infer that decreases in pediatric in-person care during various stages of “stay at home” measures may have contributed to an increasing proportion of teens sending MyChart messaging across the first 3 quarters of 2020 as demonstrated in Fig 2. Following this, greater familiarity with the platform may account for ongoing levels of higher use among the population studied.

This study examined a large sample of adolescents receiving care in a nonacademic setting. Using months of individual patient characteristics as the unit of observation for analysis allowed for more specific understanding of dynamic characteristics like age and parent proxy account status over time. We adjusted for presence of an active proxy account to minimize measured teen online patient portal secure messaging use driven primarily by parent users of teen accounts, as recent studies have demonstrated a high proportion of teen accounts are accessed by parents.7  Based on the nature of this administrative data, we were not able to determine whether secure messages sent from an adolescent account originated only from the adolescent patient and not a parent or guardian accessing their account. Messaging rates may have been higher if teens were able to initiate secure messages with specialty providers in Mental Health and Wellness, Obstetrics and Gynecology, and Adolescent Medicine. Transgender and gender diverse identities are likely underrepresented because of how limited and variable practices are to record gender identity data in the electronic health record. Administratively recorded sex may be sex assigned at birth or reflect a changed sex marker on one’s birth certificate as part of gender affirmation. We note that relative risks of the smaller categories, such as American Indian and Alaska Native, had wider confidence intervals, raising the risk of inappropriately accepting a null hypothesis when there may be significantly different risk of use among this population. We also acknowledge that multiple comparisons in this analysis, particularly the nonmutually exclusive racial and ethnic groups, increase the risk of spurious findings.

This study design did not allow for an assessment of causality nor could it account for all potentially relevant but unmeasured confounders, such as medical complexity, use of prescription medications, primary care versus specialty care context, or primary language spoken. Future research is needed to elucidate the causal pathway of adolescent secure message use and to assess and adjust for additional confounders to clarify targets for intervention development.

Adolescents are engaging with the patient portal for secure messaging at increasing rates. Pediatric researchers and health systems must design interventions and workflows that respond to the dynamic and diverse nature of adolescent developmental capacity, care navigation skills, and parent involvement in care. Our findings highlight the importance of how differences in the uptake of teen-provider electronic communication for health care should be viewed in the context of all forms of care access. Designing care to match individual teen needs whether in person, via telehealth, or asynchronously on a patient portal is essential for equitable, effective care.

We thank health system leaders Drs. Gina Sucato and Caryn Avery who supported this effort as well as Ron Johnson for his programming support.

Drs Hoopes, Coley, Mangione-Smith, and Ralston conceptualized and designed the study and drafted the initial manuscript; Ms Cushing-Haugen conducted the analyses; Ms Fuller provided programmatic support to develop the analytic dataset; Dr White conceptualized the study; and all authors critically reviewed and revised the manuscript, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.

FUNDING: Dr Hoopes was supported by grant number K12HS026369 from the Agency for Healthcare Research and Quality. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Agency for Healthcare Research and Quality. The other authors received no additional funding.

CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no conflicts of interest relevant to this article to disclose.

95% CI

95% confidence interval

GEE

generalized estimating equations

HMO

Health Maintenance Organization

KPWA

Kaiser Permanente Washington

PPO

preferred provider organization

RR

relative risk

TGD

transgender or gender diverse

1
Keesara
S
,
Jonas
A
,
Schulman
K
.
Covid-19 and health care’s digital revolution
.
N Engl J Med
.
2020
;
382
(
23
):
e82
2
Wosik
J
,
Fudim
M
,
Cameron
B
, et al
.
Telehealth transformation: COVID-19 and the rise of virtual care
.
J Am Med Inform Assoc
.
2020
;
27
(
6
):
957
962
3
Centers for Medicare and Medicaid Services
.
2022 Medicare promoting interoperability program requirements
.
4
Antonio
MG
,
Petrovskaya
O
,
Lau
F
.
The state of evidence in patient portals: umbrella review
.
J Med Internet Res
.
2020
;
22
(
11
):
e23851
5
Bush
RA
,
Connelly
CD
,
Fuller
M
,
Pérez
A
.
Implementation of the integrated electronic patient portal in the pediatric population: a systematic review
.
Telemed J E Health
.
2016
;
22
(
2
):
144
152
6
Bergman
DA
,
Brown
NL
,
Wilson
S
.
Teen use of a patient portal: a qualitative study of parent and teen attitudes
.
Perspect Health Inf Manag
.
2008
;
5
:
13
7
Ip
W
,
Yang
S
,
Parker
J
, et al
.
Assessment of prevalence of adolescent patient portal account access by guardians
.
JAMA Netw Open
.
2021
;
4
(
9
):
e2124733
8
Hargreaves
DS
,
Elliott
MN
,
Viner
RM
,
Richmond
TK
,
Schuster
MA
.
Unmet health care need in US adolescents and adult health outcomes
.
Pediatrics
.
2015
;
136
(
3
):
513
520
9
Simon
TD
,
Cawthon
ML
,
Stanford
S
, et al;
Center of Excellence on Quality of Care Measures for Children with Complex Needs (COE4CCN) Medical Complexity Working Group
.
Pediatric medical complexity algorithm: a new method to stratify children by medical complexity
.
Pediatrics
.
2014
;
133
(
6
):
e1647
e1654
10
Wolcott
V
,
Agarwal
R
,
Nelson
DA
.
Is provider secure messaging associated with patient messaging behavior? Evidence from the US Army
.
J Med Internet Res
.
2017
;
19
(
4
):
e103
11
Flanagin
A
,
Frey
T
,
Christiansen
SL
.
Updated guidance on the reporting of race and ethnicity in medical and science journals
.
JAMA
.
2021
;
326
(
7
):
621
627
12
Cook
LA
,
Sachs
J
,
Weiskopf
NG
.
The quality of social determinants data in the electronic health record: a systematic review
.
J Am Med Inform Assoc
.
2021
;
29
(
1
):
187
196
13
Liang
K-Y
,
Zeger
SL
.
Longitudinal data analysis using generalized linear models
.
Biometrika
.
1986
;
73
(
1
):
13
22
14
Carlson
JL
,
Goldstein
R
,
Buhr
T
,
Buhr
N
.
Teenager, parent, and clinician perspectives on the electronic health record
.
Pediatrics
.
2020
;
145
(
3
):
e20190193
15
Thompson
LA
,
Martinko
T
,
Budd
P
,
Mercado
R
,
Schentrup
AM
.
Meaningful use of a confidential adolescent patient portal
.
J Adolesc Health
.
2016
;
58
(
2
):
134
140
16
Trent
M
,
Dooley
DG
,
Dougé
J
,
Section On Adolescent
H
;
Section on Adolescent Health
;
Council on Community Pediatrics
;
Committee on Adolescence
.
The impact of racism on child and adolescent health
.
Pediatrics
.
2019
;
144
(
2
):
e20191765
17
Miklin
DJ
,
Vangara
SS
,
Delamater
AM
,
Goodman
KW
.
Understanding of and barriers to electronic health record patient portal access in a culturally diverse pediatric population
.
JMIR Med Inform
.
2019
;
7
(
2
):
e11570
18
Mold
F
,
de Lusignan
S
,
Sheikh
A
, et al
.
Patients’ online access to their electronic health records and linked online services: a systematic review in primary care
.
Br J Gen Pract
.
2015
;
65
(
632
):
e141
e151
19
Kruse
CS
,
Bolton
K
,
Freriks
G
.
The effect of patient portals on quality outcomes and its implications to meaningful use: a systematic review
.
J Med Internet Res
.
2015
;
17
(
2
):
e44
20
Goldzweig
CL
,
Orshansky
G
,
Paige
NM
, et al
.
Electronic patient portals: evidence on health outcomes, satisfaction, efficiency, and attitudes
.
Annals of Internal Medicine
.
2013
;
159
(
10
):
677
687
21
Lebrun-Harris
LA
,
Sappenfield
OR
,
Warren
MD
.
Missed and delayed preventive health care visits among US children due to the COVID-19 pandemic
.
Public Health Rep
.
2022
;
137
(
2
):
336
343
22
Nguyen
KH
,
Nguyen
K
,
Lekshmi
D
,
Corlin
L
,
Niska
RW
.
Delays in children’s preventive health services during the COVID-19 pandemic
.
Fam Med
.
2022
;
54
(
5
):
350
361