Table 3.
Summary of the findings on patient characteristics.
| First author, year | Country | Health information technology | Study design | Sample size | Main findings |
| Abel, 2018 [31] | United States | My HealtheVet (MHV) and Clinical Video Telehealth (CVT) | Descriptive, quantitative | 2,171,325 patients | African American and Latino patients were significantly less likely to engage in use compared with White patients. Low-income patients with free care were significantly less likely to be users. Patients with schizophrenia or schizoaffective disorder were significantly less likely to be users than those with other mental health diagnoses (ORa 0.50, CI 0.47-0.53 and OR 0.75, CI 0.69-0.80, respectively). Although rural patients had 17% lower odds of My HealtheVet adoption compared with urban patients (OR 0.83, 95% CI 0.80-0.87), they were more likely to engage in Clinical Video Telehealth and dual use (OR 2.45, 95% CI 1.95-3.09 for Clinical Video Telehealth and OR 2.11, 95% CI 1.81-2.47 for dual use). |
| Ancker, 2019 [34] | United States | Blood glucose flow sheet (EpicCare and Weill Cornell Connect portal) | Observational, hypothesis testing | 53 patients | Pregnant patients: Uploaders were similar to the comparison group in terms of race, ethnicity, age, and socioeconomic status. Uploaders had more clinical encounters and portal logins before initial data upload, earlier establishment of patient portal accounts, and worse baseline blood pressure. 30 chronic disease patients: Uploaders were more likely to be Asian-American and younger, but the groups did not have other significant demographic differences. |
| Chan, 2018 [18] | United States | MyChart (EpicCare, Verona, WI) | Descriptive, quantitative | 17,699 patients | Positive associations between portal use and being aged 50-74 years, White, privately insured, English-speaking, and living outside San Francisco. |
| Fiks, 2015 [60] | United States | MyAsthma (Clinical interface in MyChart) | RCTb | 60 families of children | Parents of children with moderate or severe asthma used the portal more frequently (75% were frequent users vs 47% were parents whose child had mild persistent asthma). |
| Fiks, 2016 [32] | United States | MyAsthma | Descriptive, mixed methods | 237 families | Users were more likely to have children aged 6-9 years (P=.009), be White (P<.001), be privately insured (P<.001), have mild persistent or moderate or severe persistent asthma (P=.002), be on an asthma controller medication (P<.001), and be receiving a greater number of asthma medications at baseline on average than nonusers (P<.001). Those with persistent asthma were twice as likely to use the portal versus those with intermittent asthma (2.37% vs 1.25% at hospital practices where these data were available; P<.001). Sustained portal users were more likely than one-time users to be Hispanic (P=.02), have private insurance (P=.02), and be from the Northeast (P=.001). Sustained use parents had higher educational levels (P=.002). Positively associated with portal adoption in multivariable logistic regression: receipt of a controller medication at baseline (OR 2.0, 95% CI 1.5-2.7), private insurance (OR 2.0, 95% CI 1.3-3.1), lower child age (OR 1.4, 95% CI 1.1-1.9), and greater asthma severity (OR 1.9, 95% CI 1.2-3.0 for mild and OR 1.9, 95% CI 1.0-3.5 for moderate or severe persistent vs intermittent). |
| Foster, 2019 [43] | United States | Epic MyChart | Observational, hypothesis testing | 208,635 tests | Females (5546/13,149, 42.18%) were significantly more likely to activate the portal than males (3897/12,212, 31.91%; P<.001). Activation rates were highest for Asian (262/451, 58.1%) and White individuals (8155/20,637, 39.52%) and lower for African American/Black (491/2254, 21.78%; P<.001 compared with White), Hispanic/Latino (333/1257, 26.49%; P<.001 compared with White), and other individuals (241/762, 31.6%; P<.001 compared with White). The activation rate for patients aged 18-70 years was 41.61% (7593/18,246). The overall pattern of radiologic image viewing with respect to age and gender showed similar trends to those described for laboratory testing. |
| Gordon, 2016 [20] | United States | Kaiser Permanente Northern California patient portal | Observational, hypothesis testing/descriptive, quantitative | 231,082 patients/4980 patients | Older seniors (aged 70-74 and 75-79 years) were significantly less likely than those aged 65-69 years to have registered, and to have used the patient portal to send a secure message, view laboratory test results online, or order prescription refills at least once by the end of the year. Slightly over 70% had been diagnosed with a chronic cardiovascular condition, and 90% reported taking at least one prescription medication for a chronic condition. |
| Gossec, 2017 [59] | France | Sanoia | RCT | 320 patients | In multivariate analyses, the only factor related to connecting more than twice to the platform was being a member of a patient association (OR 1.44, 95% CI 1.17-1.77; P<.001). In the groups with high and low numbers of connections, the percentages of patient association members were 24.7% and 6.5%, respectively. |
| Griffin, 2016 [44] | United States | My UNC Chart | Observational, hypothesis testing | 2975 patients | Active users had a higher proportion of Caucasian patients, higher Charlson Comorbidity scores, and a higher proportion of patients admitted to an academic medical center than light users. |
| Huang, 2019 [45] | United States | myPennMedicine (branded version of Epic MyChart) | Observational, hypothesis testing | 10,000 patients | Users were more likely to be younger (63.46 years [users] vs 66.08 years [nonusers]; P<.001) and have higher income (US $74,172 [users] vs US $62,940 [nonusers]; P<.001) than nonusers. The percentage of White race was substantially higher among users (72.77% [4317/5932] [users] vs 52.58% [2139/4068] [nonusers]; P<.001). For users, the percentage of payments by commercial insurance was higher (60.99% [3618/5932] [users] vs 40.12% [1632/4068] [nonusers]; P<.001) and the percentage of payments by Medicare or Medicaid was lower (Medicare: 34.91% [2071/5932] [users] vs 48.72% [1982/4068] [nonusers]; P<.001; and Medicaid: 3.49% [207/5932] [users] vs 10.08% [410/4068] [nonusers]; P<.001). The difference in sex between users and nonusers was not statistically significant. No significant difference was found in any provider-level characteristic between the two groups. |
| Jhamb, 2015 [46] | United States | Free patient portal tethered to an ambulatory EHRc | Observational, hypothesis testing | 1098 patients | Users were younger and more likely to be non-Black, be married, have private insurance, and have higher neighborhood median household income. Users were less likely to have diabetes, coronary artery disease, or congestive heart failure, but were more likely to have had a kidney transplant. Older age, Black race, unmarried status, Medicaid or Medicare insurance (vs private), and lower neighborhood median household income were associated with not using the portal. |
| Kipping, 2016 [35] | Canada | Ontario Shores HealthCheck Patient Portal | Observational, hypothesis testing | 91 patients | A similar proportion of patients (1756/3158, 55.6%) and portal users (266/432, 61.6%) were female. Age distribution was relatively similar. The majority of users were between 25 and 34 years. |
| Krist, 2014 [33] | United States | AllscriptsTouchworks EHR | Descriptive, mixed methods | 112,893 patients | Older patients were more likely to create a PHRd account as they had chronic conditions. |
| Laranjo, 2017 [22] | Australia | Portuguese National patient portal | Descriptive, quantitative | 109,619 participants | Geographic analysis revealed higher proportions of PHR adoption in urban centers when compared with rural noncoastal districts. |
| Lau, 2014 [36] | Canada | BCDiabetes.ca | Observational, hypothesis testing | 1957 patients | Users tended to be younger (mean difference of 4.28 years; P=.06), have lower baseline HbA1c (mean difference of 0.89%; P<.01), and have higher baseline weight (mean difference of 7.53 kg; P=.06) than nonusers. There was no difference in gender or total follow-up time. Follow-up HbA1c levels tended to be lower in users than nonusers (mean difference of 0.75%; P<.01), and users were significantly more likely to have HbA1c of 7% at their last follow-up visit (P=.01). No significant differences in LDLe and SBPf were observed between users and nonusers at initial visits and follow-up visits. |
| Manard, 2016 [37] | United States | Online patient portal | Observational, hypothesis testing | 1571 patients | Users were significantly younger (P<.001), more often White (P<.001), and more often married (P<.001) than nonusers. Users were significantly from upper-middle to the highest socioeconomic status compared with nonusers (P<.001). Portal use was more common among FMg patients than GIMh patients (P<.001), and users were more often high health care utilizers (P<.02). Portal use was less common among current smokers (P<.001). Users were more likely to have depression (P<.01) and lower comorbidity scores (P<.001). |
| Mishra, 2019 [23] | United States | OpenNotes within the HealtheLife patient portal | Descriptive, quantitative | 1487 patients | Overall, 90% (n=784) were above 30 years, with 8% between 18 and 29 years; 40% were above 60 years; and 50% were between 30 and 59 years. One participant stated accessing the portal for a minor and another as a family surrogate. Moreover, 92% (n=797) had a college degree or greater, 24% (n=205) had a graduate degree, and 1% (n=12) had less than 12th grade education. Individuals in the 18-29 and >60 years groups were more likely to find the notes helpful. Greater note comprehension was correlated with greater education. Noncollege participants were more likely to access notes “many times” than college participants (P=.02). |
| North, 2014 [38] | United States | Mayo Clinic Health System | Observational, hypothesis testing | 2357 primary care patients | The majority of patients sending messages were female, were White, lived locally, and were employed by the Mayo Clinic. |
| Plate, 2019 [39] | United States | MyChart; Epic Systems Corporation | Observational, hypothesis testing | 6426 patients | Overall, 4623 people registered on MyChart logged into the patient portal at least once within 1 year from surgery, and 1803 (28%) patients were not registered users. Active users were significantly more likely to be young, have a healthy ASAi score (ASA 1 or 2), be Caucasian, be married, be employed, be privately insured, and be discharged to home. Patients not using MyChart had a higher ASA score (ASA 3 or 4) and were more likely to be African American, unmarried, and unemployed. Patients without MyChart were more likely to have Medicare or Medicaid insurance and be discharged to a skilled nursing facility. |
| Portz, 2019 [52] | United States | My Health Manager (Kaiser Permanente Colorado patient portal) | Qualitative descriptive study | 24 patients | The mean age was 78 years. Patients were primarily White (12, 80%) and women (12, 80%). Education: high school graduate, 1 (7%); some college graduate, 7 (47%); and college graduate, 7 (47%). Income: <US $30,000, 2 (13%); US $30,000-49,999, 7 (47%); US $50,000-74,999, 2 (13%); >US $75,000, 2 (13%); chose not to answer, 2 (13%). |
| Powell, 2018 [40] | United States | FollowMyHealth portal | Observational, hypothesis testing | 500 patients | No significant relationship between the number of logins and any of the demographic variables; however, when those with zero logins were removed from the model, age, distance separating the patient from his or her provider, and having a diagnosis of heart failure were all significant predictors of portal use (P<.05). |
| Price-Haywood, 2017 [25] | United States | MyOchsner patient portals (Epic System), wearable technology, smartphone mobile apps | Descriptive, quantitative | 247 patients | Portal users had higher levels of education, lower rates of inadequate health literacy, and higher rates of using the internet and having an interest in websites or smartphone apps for tracking health. The odds of portal use increased with total eHEALS scores (health literacy scale) and decreased among Black patients. |
| Riippa, 2014 [56] | Finland | No specific portal | Interventional, other than RCT | 876 patients | Patients with a severe diagnosis during the intervention showed the greatest positive change in patient activation (mean change 5.4, SD 8.4). Patients diagnosed 1-2 years ago (mean change 2.3, SD 15.7) and patients with no severe diagnoses (mean change 1.6, SD 13.1) showed a positive change in patient activation. |
| Ronda, 2014 [26] | Netherlands | Digitaal Logboek | Descriptive, quantitative | 1390 patients | Multivariable analysis showed that increasing age and smoking were associated with not using the portal. A higher educational level, treatment by an internist, using insulin, polypharmacy, better diabetes knowledge, and more hyperglycemic episodes were less likely to be associated with not using the portal. |
| Smith, 2015 [27] | United States | EpicCare | Descriptive, quantitative | 534 patients | Significant predictors of registering were as follows: gender (male 65.3% vs female 55.1%), race (White 71.7% vs African American 27.7% vs “other” races 41.7%), education (more educated people were more likely to register), number of chronic conditions (70.9% with zero conditions, 63.2% with one condition, and 50.0% with two or more conditions), health literacy (adequate 72.7% vs marginal 46.4% vs limited health literacy 21.7%). |
| Sun, 2019 [41] | United States | Epic’s personal health record system | Observational, hypothesis testing | 38,399 patients | Almost one-third of patients (n=12,615; 32.9%, 95% CI 32.38%-33.32%) had used the portal for a mean of 2.5 (SD 1.9) years prior to the study period. Portal use was higher on weekdays (P<.001). An increase in portal use was observed in response to email reminders. A nonlinear relationship between age and portal use was observed and depended on several other predictors (P<.05). Patients living in more rural areas with low income were at lower odds to use the portal (P=.02), and this finding also applied to non‐Whites with low income (P<.001). More chronic conditions and a higher initial HbA1c value were associated with portal use (P=.01). |
| Tsai, 2019 [28] | United States | Epic’s personal health record system | Descriptive, quantitative | 109,200 patients | Active portal users were on average older (49.45 vs 46.22 years) and frequently female (62.59% vs 54.91%). Both the differences in mean age (P=.008) and gender (P=.04) were significant. There was a bimodal peak in terms of active users, with active users more likely to be in their 30s and 60s. The difference among age groups was significant (P<.001). Differences in racial composition, insurance, and language were not significant. |
| Van der Vaart, 2014 [57] | Netherlands | Medisch Spectrum Twente | Interventional, other than RCT | 360 patients | Univariate analyses showed that age, marital status, education level, employment, health literacy, and internet-related characteristics were significantly related to portal use. Nonusers were more often older, single, lower educated, and unemployed. Respondents with higher health literacy were more inclined to login on the portal, and respondents who used the internet more often had more years of experience and perceived their own skills as better. |
| Wade-Vuturo, 2013 [8] | United States | MyHealthAtVanderbilt patient portal | Descriptive, quantitative | 54 patients | Participant age, gender, race, income, and education level were not associated with using SMj to send a message to a provider for any reason or using SM to schedule an appointment. |
| Wallace, 2016 [58] | United States | MyChart by Epic health record system | Interventional, other than RCT | 36,549 patients | Men, non-White patients, and Hispanic patients were significantly less likely to login once, 2 to 23 times, or 24 times than women, White patients, or non-Hispanic patients. Patients with public insurance were less likely to login than those with private insurance across all MyChart usage categories. Patients with income levels 100% of the FPLk were more likely to login one time than those below the FPL level. |
| Wedd, 2019 [30] | United States | Unspecified patient portal | Descriptive, quantitative | 710 patients | Black patients were less likely to use the portal vs White patients among both kidney (Black 57% vs White 74%) and liver (Black 28% vs White 55%) transplant recipients. In adjusted multivariable analyses, kidney transplant recipients were more likely to use the portal if they had higher education. Among liver recipients, patients who were White and had higher education were more likely to use the portal. |
| Zhong, 2018 [42] | United States | MyUFHealth (also known as MyChart by Epic) | Observational, hypothesis testing | 15,659 nonusers and 5494 users | The user group comprised 53.1% patients with more than four chronic problems (vs 40.2% of the matched nonuser group), and had more patients bearing 10 or more chronic problems (18.2% vs 12.2%). Individuals enrolled in the patient portal were mostly middle aged (31-64 years) and female. Married patients were more likely to adopt the portal. Medicare and Medicaid patients, and Black or African American patients were less likely to be adopters. Portal adoption was also associated with the baseline number of active medical problems (P<.05). |
aOR: odds ratio.
bRCT: randomized controlled trial.
cEHR: electronic health record.
dPHR: personal health record.
eLDL: low-density lipoprotein.
fSBP: systolic blood pressure.
gFM: family medicine.
hGIM: general internal medicine.
iASA: American Society of Anesthesiology.
jSM: secure messaging.
kFPL: federal poverty level.