Association Between Social Determinants of Health and Patient Portal Utilization in the United States

Abstract

(1) Background: Differences in health outcomes across populations are due to disparities in access to the social determinants of health (SDoH), such as educational level, household income, and internet access. With several positive outcomes reported with patient portal use, examining the associated social determinants of health is imperative. Objective: This study analyzed the association between social determinants of health—education, health insurance, household income, rurality, and internet access—and patient portal use among adults in the United States before and after the COVID-19 pandemic. (2) Methods: The research used a quantitative, retrospective study design and secondary data from the combined cycles 1 to 4 of the Health Information National Trends Survey 5 (N = 14,103) and 6 (N = 5958). Descriptive statistics and logistic regression were conducted to examine the association between the variables operationalizing SDoH and the use of patient portals. (3) Results: Forty-percent (40%) of respondents reported using a patient portal before the pandemic, and this increased to 61% in 2022. The multivariable logistic regression showed higher odds of patient portal utilization by women compared to men (AOR = 1.56; CI, 1.32–1.83), those with at least a college degree compared to less than high school education (AOR = 2.23; CI, 1.29–3.83), and annual family income of USD 75,000 and above compared to those <USD 20,000 (AOR = 1.59; CI, 1.18–2.15). Those with access to the internet and health insurance also had significantly higher odds of using their patient portals. However, those who identified as Hispanic and non-Hispanic Black and residing in a rural area rather than urban (AOR = 0.72; CI, 0.54–0.95) had significantly lower odds of using their patient portals even after the pandemic. (4) Conclusions: The social determinants of health included in this study showed significant influence on patient portal utilization, which has implications for policymakers and public health stakeholders tasked with promoting patient portal utilization and its benefits.

1. Introduction

Clinical care is not solely responsible for an individual’s or community’s health outcomes, and there are factors outside the health system that impact health outcomes. Social determinants of health (SDoH), the conditions in which people are born, live, and work, influence their daily activities and determine up to 50% of their health outcomes [1,2]. These non-medical factors, grouped into healthcare access and quality, education access and quality, social and community context, neighborhood and built environment, and economic stability, significantly shape health outcomes [1,3]. A secure online website that gives patients convenient, 24-hour access to personal health information from anywhere with an internet connection is a patient portal, and these portals are managed by healthcare organizations that provide this access to their patients [4,5]. Therefore, in the wake of the shift toward digital health solutions, healthcare systems and patients in the U.S. increasingly rely on patient portals [6]. In 2018, about 31% of individuals in the United States reported using their patient portals [7]. Patient portals differ from personal health records in that while the health organization maintains the patient portal, the patient maintains the personal health record [8]. The patient portal is connected to their provider’s electronic health record and provides real-time information about their health data [8].

Despite its existence since the 1990s, patient portals became famous in 2006, with most healthcare organizations in the United States providing interested patients access to the portal [9,10,11]. The 2009 HITECH Act contributed to the popularity and advancement of patient portal use due to the USD 30 million incentive allocated to health systems whose portals met the meaningful use criteria [5,9,12]. As of 2015, more than 90% of health systems now allow patients to create a portal [11]. The requirements for earning the incentives were met when applications such as this improved patient engagement by providing a patient portal that enables secure communication, such as messaging between providers and patients, a summary of the patient encounter after every clinic visit, and the ability of patients to view and download their data [5,9,12]. Several positive outcomes have been associated with patient portals, including improved patient engagement and service satisfaction, the prevention of medical mistakes, and better communication between patients and providers [5,11,13,14]. Patient portal use has also been associated with improved patient outcomes at the individual level, reducing readmission rates and hospital stays and reducing the probability of hospital and emergency room visits for chronic disease patients [15].

Despite the 2009 HITECH Act, there are still disparities in patient portal access, and not everyone can enjoy the potential benefits, leading to inequitable health outcomes. Variations in exposure to the social determinants of health may be responsible for disparities in health outcomes and may also affect patient portal usage. Studies that used hospital datasets identified some social determinants of health associated with patient portal utilization, such as income level, healthcare access, internet availability, rurality, and education [6,12,15,16]. While individuals with access to the internet and a regular healthcare provider were more likely to use a patient portal, those who resided in rural areas and had lower education and income levels were less likely to use the portal [6,12]. The use of patient portals is still low among certain groups of people, such as racial minorities, people with disabilities, older adults, and the low-income population [11]. Patient portals may serve as a mechanism through which the social determinants of health impact and improve health outcomes [11,17].

Studies have been conducted using hospital-derived datasets on the sociodemographic factors associated with patient portal utilization [18,19]. While Gordon and Hornbrook (2016) focused on the use of patient portals among the older population, Wallace et al. (2016) assessed the pattern of portal use among the underserved population [18,19]. Researchers have also used earlier iterations of this dataset to determine how race and socioeconomic characteristics affected patient portal use [16,20,21] and the factors associated with the non-use of a patient portal [22]. This study differs from past studies because it used the complete iterations of a nationally representative dataset collected majorly during the pre-pandemic and early post-pandemic periods. Also, while past researchers used a four-race category [12,20], the current study included the non-Hispanic Asian racial identity as a stand-alone category.

The current study aims to complement and update findings from past studies on the association between the social determinants of health and patient portals during the pre- and post-pandemic period. The results will provide a baseline that will be useful in monitoring the trend in the post-pandemic period when more people have had to use some health information technology. This study investigates the role of social determinants of health in patient portal use among adults in the United States.

2. Materials and Methods

2.1. Data Source

The current study was based on a quantitative retrospective study design using a nationally representative sample of 14,103 respondents in the HINTS 5 cycles 1 to 4 conducted majorly in the pre-COVID-19 era and 5958 respondents in the HINTS 6 conducted in 2022 who answered the question about patient portal utilization. The respondents were non-institutionalized civilian adults in the United States, and a two-stage sampling design was used to collect the data. Data for this study were obtained from the Health Information National Trends Survey (HINTS) 5 cycles 1 to 4 from 2017 to 2020 and HINTS 6 in 2022, which was collected so that households were not sampled twice. The National Cancer Institute conducts this survey using a cross-sectional retrospective study design [23]. The HINTS data are nationally representative and were collected from 2017 to 2020 and 2022 [23]. The sample size for this study was restricted to those who responded to the patient portal utilization question, and each survey iteration had a response rate of at least 30%.

2.2. Measures

2.2.1. Dependent Variable

The dependent variable of interest was patient portal usage. The survey question that measured patient portal usage was “How many times did you access your online medical record in the last 12 months?” The original response categories of 0 times, 1–2 times, 3–5 times, 6–9 times, and 10 or more times were dichotomized into No or Yes, with those that responded 0 times being No and those that responded 1 or more times being Yes. The 2022 survey response was updated to include those who did not have access to an online medical record. For this study, respondents who answered this question were coded as 0 times.

2.2.2. Independent Variables

The variables that operationalized the social determinants of health were the educational level (less than high school/high school graduate/some college/college graduate or more), annual household income (<USD 20,000; USD 20,000–USD 35,000; USD 35,000–USD 50,000; USD 50,000–USD 75,000; >USD 75,000), health insurance (Yes/No), rurality (Yes/No), and internet access (Yes/No). Sociodemographic characteristics such as age, sex, and race were used as control variables.

2.3. Analytical Methods

The data analysis was restricted to those who answered the survey question that asked if respondents had used a patient portal in the 12 months before the survey. Descriptive statistics were computed for both the dependent and independent variables. After controlling for other covariates in the model, we conducted a multivariable logistic regression analysis to show the association of each independent variable with the patient portal utilization. Our analyses used the survey sampling weights available in the HINTS dataset to improve the generalizability of our results to the survey population, which is adults in the U.S. The statistical software used for all analyses was STATA version 16 by STATA Corp. Georgia Southern University Institutional Review approved this study under protocol H23189.

3. Results

3.1. Characteristics of Respondents

The respondents had an average age of approximately 49 years. In the pre-pandemic era, a significant proportion, about sixty percent (60%), did not utilize their patient portal in the 12 months leading up to the survey. However, in the post-pandemic era, the proportion of users increased to 61%. Among the respondents, 52% were women, and the majority identified as Non-Hispanic White (66%). Individuals in the 50–64 age range constituted 29% of the respondents, with 70% holding at least some college degree. Notably, 42% reported an annual household income of USD 75,000 or more, and 13% lived in rural areas. The majority had health insurance coverage (92%), and internet access (86%), as detailed in

Table 1. Descriptive statistics of the patient portal utilization and the independent variables.

	2017–2020		2022
Variables	N (Unweighted) = 14,103	% (Weighted)	N (Unweighted) = 5958	% (Weighted)
Dependent Variables
Patient Portal Utilization
None	8337	60%	2304	39%
1–2 times	2538	19%	1196	20%
3–5 times	1844	12%	1132	19%
6 times or more	1384	9%	1326	22%
Patient Portal Usage
No	8337	60%	2304	39%
Yes	5766	40%	3654	61%
Independent Variables
Sex
Male	5402	48%	2261	49%
Female	7584	52%	3442	51%
Race/Ethnicity
Non-Hispanic White	8114	66%	3156	61%
Non-Hispanic Black or African American	1736	10%	869	11%
Hispanic	1900	16%	972	17%
Non-Hispanic Asian	581	5%	283	6%
Non-Hispanic Other	453	2%	178	5%
Age Group
18–34 years	1725	23%	892	26%
35–49 years	2692	28%	1186	25%
50–64 years	4416	29%	1711	28%
65–74 years	3064	12%	1307	13%
75+ years	1810	7%	786	8%
Educational Level
Less than High School	855	8%	353	6%
High School Graduate	2448	22%	1025	21%
Some College	4066	36%	1635	39%
College Graduate or More	6370	34%	2694	33%
Annual Household Income
<USD 20,000	2210	15%	919	14%
USD 20,000–<USD 35,000	1627	11%	697	11%
USD 35,000–<USD 50,000	1641	13%	715	12%
USD 50,000–<USD 75,000	2265	19%	927	18%
USD 75,000 or more	4902	42%	2148	45%
Rural Residence
No	12,455	87%	5185	88%
Yes	1648	13%	773	12%
Health Insurance
No	684	8%	493	11%
Yes	13,278	92%	5429	89%
Internet Access
No	2564	14%	974	14%
Yes	11,495	86%	4982	86%

Abbreviations: N = number of observations; the counts may not add up to the total number of observations due to missing data.

. The proportion of respondents’ characteristics in the pre-and post-pandemic era datasets was largely similar.

3.2. Logistic Regression of Patient Portal Use and the Social Determinants of Health

3.2.1. Pre-Pandemic Results

Table 2. Multivariable logistic regression of patient portal utilization from HINTS 5 cycles 1–4 (2017–2020) and HINTS 6 (2022).

Patient Portal Utilization
	2017–2020				2022
	AOR	95% CI		p-Level	AOR	95% CI		p-Level
		LL	UL			LL	UL
Sex
Male	(Ref. Category)
Female	1.59	1.35	1.87	<0.001	2.08	1.62	2.67	<0.001
Race
Non-Hispanic White	(Ref. Category)
Non-Hispanic Black or African American	0.93	0.73	1.18	0.53	0.73	0.53	0.99	0.04
Hispanic	0.66	0.52	0.84	0.001	0.64	0.45	0.9	0.01
Non-Hispanic Asian	1.1	0.73	1.65	0.65	0.94	0.45	1.96	0.86
Non-Hispanic Other	0.95	0.56	1.6	0.83	1	0.5	2.01	0.99
Age Group
18–34 years	(Ref. Category)
35–49 years	1.33	1.02	1.75	0.04	1.05	0.7	1.58	0.8
50–64 years	1.24	0.99	1.58	0.06	1.48	0.96	2.28	0.07
65–74 years	1.21	0.91	1.62	0.19	0.84	0.56	1.26	0.4
75+ years	1.06	0.73	1.54	0.74	0.68	0.39	1.19	0.17
Educational Level
Less than High School	(Ref. Category)
High School Graduate	1.08	0.64	1.82	0.76	1.23	0.54	2.82	0.62
Some College	1.45	0.86	2.44	0.16	1.54	0.72	3.28	0.26
College Graduate or More	2.11	1.29	3.45	0.004	2.67	1.21	5.87	0.02
Annual Household Income
<USD 20,000	(Ref. Category)
USD 20,000–<USD 35,000	0.8	0.53	1.21	0.28	1.22	0.88	1.7	0.22
USD 35,000–<USD 50,000	1.29	0.86	1.93	0.21	2.12	1.42	3.18	<0.001
USD 50,000–<USD 75,000	1.17	0.84	1.61	0.35	1.99	1.26	3.12	0.004
USD 75,000 or more	1.69	1.2	2.37	0.003	2.83	1.86	4.31	<0.001
Rural Residence
No	(Ref. Category)
Yes	0.74	0.57	0.98	0.03	0.71	0.51	0.98	0.04
Health Insurance
No	(Ref. Category)
Yes	3.84	2.04	7.23	<0.001	2.12	1.34	3.35	0.002
Internet Access
No	(Ref. Category)
Yes	3.67	2.65	5.09	<0.001	5.41	4.23	6.93	<0.001

Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; LL, lower limit; UL, upper limit; RC, reference category. Note: bolded AORs indicate that adjusted odds of the outcome variable are significantly different at p ≤ 0.05 for an attribute of a variable compared to the reference category.

shows the logistic regression of patient portal use with the variables representing social health determinants. In the pre-pandemic era, after adjusting the effect of other variables in the model, women had higher odds (AOR, 1.56; CI, 1.32–1.83) of accessing their patient portals than men. Hispanics had lower odds (AOR, 0.75; CI, 0.58–0.96) of using the patient portals compared to non-Hispanic Whites. U.S. adults with a college degree or higher education had higher odds (AOR, 2.23; CI, 1.29–3.83) of accessing their patient portals than those with less than a high school education. Study participants with an annual family income of USD 75,000 and above were significantly more likely (AOR, 1.59; CI, 1.18–2.15) than those with a household income of less than USD 20,000. The odds of patient portal utilization were also significantly higher for those with healthcare insurance vs. those with no healthcare insurance (AOR, 2.81; CI, 1.46–5.41) and those with access to the internet rather than no access (AOR, 3.54; CI, 2.55–4.93). Compared to those residing in urban areas, rural residents had lower odds (AOR, 0.72; CI, 0.54–0.95) of using their patient portal. Our study showed that patient age was not significantly associated with using the patient portal after adjusting for the impact of other variables in the model.

3.2.2. Post-Pandemic Results

The logistic regression analysis showed largely the same results in the post-pandemic period with the following variables: female, Hispanics, non-Hispanic Blacks, and college graduate education level. However, the new findings showed that both Non-Hispanic Blacks (AOR, 0.73; CI, 0.53–0.99) and Hispanics (AOR, 0.64; CI, 0.45–0.90) were less likely to use their patient portal compared to Non-Hispanic Whites. Compared to respondents with an annual household income of less than USD 20,000, respondents with an annual family income of USD 35,000–<USD 50,000 (AOR, 2.12; CI, 1.42–3.18), USD 50,000–<USD 75,000 (AOR, 1.99; CI, 1.26–3.12), and USD 75,000 and more (AOR, 2.83; CI, 1.86–4.31) had higher odds of using their patient portal.

4. Discussion

Using a quantitative retrospective study design, this study examined the association between the social determinants of health and adult patient portal utilization in the United States. We used data from multiple waves of a national survey, i.e., HINTS 5 cycles 1–4 conducted from 2017 to 2020 and HINTS 6 conducted in 2022. Knowing the influence of social determinants of health on patient portal utilization by U.S. adults is critical for addressing barriers to patient engagement, good health outcomes, and care coordination efficiency [24]. Addressing those barriers is essential because patient portal use can increase patients’ understanding of their health conditions, improve patient safety, reduce caregiver burden, increase medication adherence through ease of refills through the patient portal, and improve patient communications with providers for both preventive and follow-up care [24,25].

Although less than 50% of adults accessed their patient portal in the 12 months preceding the pre-pandemic survey, this study found an increase in portal use from 28% in 2017 [11] to 41%. The 2022 post-pandemic survey showed a further increase in patient portal use to 61%. The relatively increased usage in the post-pandemic period indicates an increased patient portal uptake in the general population. It is safe to assume that the pandemic forced people to explore technology that will connect them to their healthcare provider, and patient portals played a major role in telemedicine access. More efforts, including patient incentives, should be explored to ensure more people have access to their portals.

In addition to the existing findings from past studies, the pre-pandemic study results showed that Hispanics had lower odds of accessing their patient portal; however, in the post-pandemic era, the results showed that both Hispanics and Non-Hispanic Blacks were less likely to access their patient portal. While Goel et al. and El-Touky found no racial differences in patient portal use, other researchers found that compared to non-Hispanic Whites, non-Hispanic Blacks and Hispanics were less likely to use their patient portal [12,20,26]. Owolo et al. (2003) found that being non-White was a barrier to patient portal utilization [24]. This study’s results show that racial disparities and inequity in patient portal use persist. A past research study cited privacy concerns about why Blacks and Hispanics were not using their portals [22]; another study suggests that the disparity in use originated from these demographic groups not being offered patient portal access by healthcare providers. There might be low interest in accessing this platform due to limited knowledge of the associated benefits. However, it is safe to assume that the limited language options may make it difficult for people whose first language is not English. Language preferences, mobile-friendly interface, and conversational ability in the added languages must be incorporated into patient portal development to overcome the language barrier and racial inequities associated with patient portal use. The patient should have the option of choosing their preferred language so the portal is more accessible to a more diverse population.

Our study showed that, compared to men, women were more likely to use the patient portals. Our findings are similar to previous studies that found that females are more likely to use patient portals [12,24,25,26,27]. This could be because women are the primary caregivers in many families [28] and, as a result, must log into the portals to monitor and manage health trends. Also, women use more healthcare services than men due to menopause and other chronic conditions, which could contribute to their higher likelihood of using patient portals [29]. Although women are more likely to use patient portals than men, their role as care seekers for the family necessitates their ongoing support in connecting them to patient portals through better access to the internet and other logistics needed for patient portal utilization.

Our study results showed that U.S. adults with a college degree or higher education were more likely to use their patient portal. Also, respondents with an annual household income of USD 75,000 or more and those with an annual household income of at least USD 35,000 had greater odds of accessing their patient portals in the pre-and post-pandemic periods, respectively. Our finding is consistent with a study by Owolo et al. (2023) and other studies that found that higher education and income levels were associated with increased patient portal utilization [7,12,23]. The level of education appears to be a significant predictor of patient portal use. Although Turner et al. (2019) found that at least a high school education increased the likelihood of patient portal use, the authors found that those with a college degree were more likely to use their patient portals [12]. El-Toukhy et al. (2020) also found that those with less than a college degree were less likely to be patient portal users [20]. It is safe to assume that formal education and digital literacy are required to access and use the patient portal. Hence, individuals with higher levels of education are likely to be more health literate and understand the benefits associated with patient portal use. However, the level of education required to use a portal is concerning, and there is a need to make patient portals user-friendly so that individuals with lower levels of education can benefit from the use. Also, educated people usually have higher income levels. They may be more interested in saving time that would have been lost if they had to obtain information about their care in person from the providers. The reduced threshold level for annual household income associated with patient portal use may be a move in the right direction, indicating more access to patient portals.

We found that individuals with health insurance were more likely to use the patient portals before and after the pandemic. Health insurance coverage is associated with patient portal utilization, perhaps due to insured persons’ better access to a regular provider [12]. Health insurance companies can increase patient portal users by providing incentives to those within their coverage.

This study showed that rural residents had lower odds of using the patient portals. Rural residents have long experienced health disparities in the United States. Our finding of reduced portal utilization among rural residents is like another study conducted by Bhavsar et al. (2022), which compared differences in portal use between rural and urban residents and found a disparity in the use of patient portals [30]. Individuals living in rural areas have reported poorer health outcomes for several reasons, including reduced access to healthcare providers [31,32]. Despite the potential benefits of increasing access to health services, including virtual care for rural residents, they may be having challenges accessing their portals due to a lack of internet access and low awareness. Rural health workers should increase awareness of the benefits of using a patient portal, and health systems in these localities should provide access to their patients. Interestingly, this association between residing in a rural area and reduced likelihood of using patient portals has not changed even after the pandemic. There is a need to explore if low interest and unwillingness to access the patient portal are contributing causes.

We found that individuals with internet access were more likely to use a patient portal. Studies have found internet access to be associated with the use of patient portals, and this is consistent with our results [6,12]. Broadband internet access is needed to use the patient portal, which is especially important for rural residents who still have trouble accessing it. Government policies that provide free or subsidized internet access to low-income individuals and rural residents will increase patient portal use.

Our study found that patient portals were not significantly associated with age after adjusting for other variables in the model. However, a study by Saif et al. (2022) and Owolo et al. (2023) found that older adults (aged 65+) were less likely to enroll in and use patient portals compared to younger adults [24,33]. Our findings suggest that the age barrier that prevented portal use among the elderly in the past has been overcome.

Strengths and Limitations

This study used the complete iterations of a nationally representative dataset to provide an updated view of the relationship between the social determinants of health and patient portal use in the United States in the pre-and post-pandemic periods. The pooled and weighted datasets provided a large sample size, making our findings generalizable to the entire U.S. population of adults. The study responses are liable to recall bias because the dependent variable-patient portal use is self-reported. The cross-sectional study design of the survey is a limitation because causality cannot be established between the social determinants of health and patient portal use. Also, frequent and one-time patient portal users were categorized as using patient portals. Another limitation is that not all dimensions of SDoH were assessed due to the use of secondary data.

5. Conclusions

Given that patient portals have the potential to improve health outcomes and social determinants of health are fundamental to accessing healthcare and information, addressing the disparities in SDoH access is essential in promoting patient portal utilization. Some of the associated correlates, such as internet use, can be addressed at the policy level by policymakers, and this is especially important for improving healthcare access for rural residents. To assure health equity, ensuring better access to patient portals is imperative for minority-race Americans. Additionally, addressing the inequitable distribution of social determinants of health, such as family incomes, formal education, and health insurance, will ensure equitable patient portal utilization. While portal developers should endeavor to ensure that, regardless of race or educational status, patient portals are user-friendly and more language-accessible, health systems have more responsibility in using their patients’ demographic characteristics to select portal functionalities that complement their patients’ needs. More context on patients’ perspectives about portals, especially vulnerable populations, is needed to understand how patient portals can be optimized for maximal benefits. Beyond access, stakeholders should explore how patient portal use influences and helps the patient make an informed decision regarding their healthcare.