Patient factors associated with telehealth quality and experience among adults with chronic conditions

Abstract Objective To evaluate patient-reported experiences of telehealth and disparities in access, use, and satisfaction with telehealth during the COVID-19 pandemic. Materials and methods We examined data from the fifth wave of the COVID-19 & Chronic Conditions (C3) study conducted between December 2020 and March 2021. Results Of the 718 participants, 342 (47.6%) reported having a telehealth visit within the past 4 months. Participants who had a recent telehealth visit were younger, reported worse overall health and chronic illness burden, and living below poverty level. Among participants who had a telehealth visit, 66.7% reported telephone visits and most participants (57.6%) rated telehealth quality as better-or-equal-to in-person visits. Inadequate health literacy was associated with lower likelihood of reporting telehealth quality and usefulness. In multivariable analyses, lower patient activation (adjusted odds ratio (AOR) 0.19, 95% CI, 0.05-0.59) and limited English proficiency (AOR 0.12, 95% CI, 0.03-0.47) were less likely to report telehealth as being better than in-person visits; lower patient activation (AOR 0.06, 95% CI, 0.003-0.41) and income below poverty level (AOR 0.36, 95% CI, 0.13-0.98) were associated with difficulty remembering telehealth visit information. Discussion Most participants reported usefulness and ease of navigating telehealth. Lower socioeconomic status, limited English proficiency, inadequate health literacy, lower educational attainment, and low patient activation are risks for poorer quality telehealth. Conclusion The COVID pandemic has accelerated the adoption of telehealth, however, disparities in access and self-reported visit quality persist. Since telemedicine is here to stay, we identify vulnerable populations and discuss potential solutions to reduce healthcare disparities in telehealth use.


Background and significance
The COVID-19 pandemic accelerated the pace of telehealth adoption in the United States through regulatory waivers and loosening of restrictions on interstate licensure or site of telehealth delivery. 1A growing body of research has investigated how telemedicine affects patient care and ways to better optimize these services.[4][5][6][7] Despite the potential clinical benefits of telemedicine, there are several barriers that may limit utility.][10][11] Differences in use of patient portals have also been documented prior to the pandemic, with men, lower educational attainment, LEP, preference for in-person communication, and not having a primary care clinician being more likely among portal nonusers. 12,135][16] With the expiration of the public health emergency in May 2023, long-term telehealth reimbursement remains uncertain while little data exist regarding its clinical effectiveness, optimal clinical use, and the patient experience.][19][20] Patient-reported experience measures (PREMs) have typically focused on overall satisfaction of care, with usefulness, easeof-use, and reliability as primary measurements, with limited considerations of perceived effectiveness of visit quality.
To date, limited research links patient experiences of telehealth to sociodemographic and psychosocial characteristics, yet to optimize telehealth visit quality, understanding potential barriers is important for clinicians and policymakers to optimize telehealth use.To fill these gaps, we examined: sociodemographic factors associated with PREMs in a diverse sample of middle-aged and older adults with at least one chronic condition who received care at academic practices and federally qualified health centers (FQHCs) in Chicago, IL.

Methods
The COVID-19 & Chronic Conditions (C3; R01AG030611-S1) study is an ongoing, longitudinal, telephone-based survey of participants who are currently active in 1 of 5 ongoing, primary care based, National Institutes of Health (NIH)funded studies (R01AG030611; R01AG046352; R01DK110172; R01HL126508; R01NR015444; See Appendix Table SA).The survey objective was to track the experiences of middle-aged and older adults, with underlying health conditions that placed them at higher risk for infection and adverse outcomes from COVID-19, through the course of the pandemic.The parent studies were chosen due to enrollment of participants that would have greater risk for COVID-19 (eg, largely middle-aged or older adult participants, with 1 or more chronic conditions).

Procedure
Participants in parent studies were recruited from multiple academic and community health center performance sites in Chicago, IL: (1) Northwestern Memorial Healthcare (NMHC), consisting of multiple, large academic practices; (2) Access Community Health Network (ACCESS), a Public Health Service 330-funded network of FQHCs; and (3) Erie Family Health Center, a network of FQHCs comprised 12 large health centers affiliated with the AllianceChicago Electronic Health Record (EHR) user community.The expanded C3 cohort for timepoint 3, 5, and 7 (ie, T3, T5, and T7 respectively) included participants from (4) Mount Sinai Icahn School of Medicine academic practice in New York, New York.All sites serve diverse patient populations and have centralized EHR systems (Epic and GE Centricity [Alli-anceChicago only]).
Eligible participants were identified via EHR queries, sending letters to eligible participants describing the study, telephoning eligible patients who did not opt out of being contacted, screening patients for eligibility, scheduling baseline interviews.Common exclusion criteria for all studies were having severe or uncorrectable cognitive, visual, or hearing impairment.The C3 study targeted study participants who had an interview completed between 2018 and March 2020, which helped to ensure the data collected previously in the parent study was most current.Detailed information on the C3 study procedures has been previously published. 21

Data collection
The C3 cohort was recruited from a subset of patients enrolled in the parent studies, which had uniform data collected on a range of patient-reported outcomes within 1 year prior to COVID-19, as well as EHRs and pharmacy records.This study used data from T5 (December 2020) time point (Figure 1).

Measurement
The C3 study collected self-reported information on patient psychosocial characteristics, COVID-19 related beliefs and actions, health and lifestyle behaviors, health services use, and mental and physical health (Appendix Table SB).

Sociodemographic and psychosocial characteristics
All parent studies had uniform collection of patient information, including demographics (age, sex, race, ethnicity), socioeconomic status (SES) (household income, number in household, educational attainment, employment status, and health insurance), self-reported chronic conditions, and a 1item, self-reported overall health (excellent, very good, good, fair, or poor).Measures of SES are routinely collected during C3 waves to note any changes to employment or income.In addition, the C3 survey included measures of other factors, including:

Social support
Participants' perceived adequacy of tangible social support was evaluated using a 2-item social support scale as part of the parent studies and the C3 battery. 22lf-efficacy The Self-Efficacy for Managing Chronic Disease 6-item Scale covers multiple domains of chronic disease self-management, including symptom control, role function, emotional functioning and communicating with physicians, and inquires about patient confidence in completing each activity.23

Health literacy
All studies also included health literacy, measured by the Newest Vital Sign (NVS).The NVS is a reliable screening tool used to determine risk for limited health literacy by asking questions about a nutrition label.Patients are given a copy of a nutrition label and asked 6 questions about how they would interpret and act on the information contained on the label.The number of correct responses is summed to produce a health literacy score ranging from 0 to 6. Scores are classified in terms of likelihood of limited literacy (0-1: likely limited; 2-3: possibly limited; 4-6: adequate). 24

Patient activation
The Consumer Health Activation Index (CHAI) was used to assess patients' degree of "activation" or motivation to participate in healthcare decisions and actions. 25

Telehealth PREMs
The C3 telephone survey at T5 used an array of self-report questions evaluating participants' telehealth experiences, including satisfaction and preferences for future clinical visits, developed by Polinski et al. 26 These survey questions, response choices, and operationalization of PREMs are provided in Appendix Table SC.To evaluate self-reported telehealth use, we first asked C3 participants if they had telemedicine appointments in the past 4 months and how many telemedicine appointments they had, as well as whether the visits were via video or telephone.Additional outcome measures included self-reported difficulty and satisfaction with telehealth services.Participants were asked how telehealth visits compared to in-person visits, to comment on difficulty remembering visit instructions, and perceived usefulness of telehealth once the COVID-19 pandemic is over.

Analysis plan
Statistical analysis was conducted using RStudio version 3.6.1.Appropriate descriptive statistics (eg, percentage, frequency, median) were performed on all patient variables.Univariable analysis was conducted to determine if there were any statistically significant demographic disparities between the following telehealth experience outcomes: (1) participants who had versus did not have a telehealth visit in the past 4 months, (2) who reported higher versus lower satisfaction of telehealth, and (3) who reported higher versus lower difficulty of remembering information discussed during telehealth services, (4) higher versus lower perceived usefulness of telehealth visits once COVID is over.
For categorical variables, data were analyzed using chisquare tests, or Fishers-exact test when sample size was low.Telehealth experience outcomes with Likert scale responses were collapsed into binary measures.Quality of telehealth compared to in-person visits (better, just as good, worse, not sure) was dichotomized as better/just as good and worse/not sure.Difficulty of remembering information discussed during telehealth services (very easy, somewhat easy, somewhat difficult, very difficult) was dichotomized as very/somewhat easy and very/somewhat difficult.Perceived usefulness of telehealth visits once COVID is over (very useful, somewhat useful, neutral, not useful) was dichotomized as very/somewhat useful and neutral/not useful.Data were dichotomized for ease of interpretation.As all outcomes were dichotomous, logistic regression models were performed for telehealth experience outcomes.All models were adjusted for any covariates associated with the outcomes in univariable analysis at P<.05, and unadjusted and adjusted odds ratios (ORs and AORs, respectively) were reported.

Results
Of the 718 participants enrolled in the C3 study at T5, 342 (47.6%) reported having a telehealth visit within the past 4 months.Patient characteristics of this subsample are presented in Table 1.
The median age was 65.8 years (range 23-91), 35.0%(251/718) were male, and 46.5% (334/718) were African American.There were no significant differences between participants who had or did not have a telehealth visit within the past 4 months by gender, race, education, health literacy, primary care setting, English proficiency, marital status, or employment status.Participants who did not report a recent telehealth visit were older (mean 66.8 (SD 11.1) vs 64.6 (SD 11.9) years old; P ¼ .03),more likely to have private insurance (58.5% vs 40.8%;P < .01),more likely to report better overall health (66.2% vs 31.2%;P < .01),more likely to have a lower chronic illness burden (46.4% had �3 chronic conditions vs 53.3%; P < .01),and more likely to be living above poverty level (54.7% vs 44.9%; P < .01).
Among participants with a recent telehealth visit, the average number of visits was 2.61 (range of 1-22), and 66.7% (228/342) reported that their most recent visit was via telephone.More than half (57.6%, 197/342) reported that the quality of telehealth was better or just as good as in-person visits.Most participants also reported ease of navigating telehealth visits: 88.9% (304/342) reported it was very or somewhat easy to describe their current health during a telehealth appointment, and 84.8% (290/342) reported it was very or somewhat easy to remember information discussed during their telehealth appointment.Most of the participants perceived telehealth as useful during the pandemic (80.4%, 275/342) and once the pandemic is over (75.1%, 257/342).
Individuals were less likely to have a video telehealth visit if they did not have access to the internet at home  a Indicates statistically significant differences between groups (P-value�.05).
(OR 0.27, 95% CI, 0.08-0.91),such as Wi-Fi or data plan through a phone, or access to any video-enabled device (OR 0.03, 95% CI, 0.002-0.49),this included their own or a family member's device.

Telehealth PREMs
Factors associated with telehealth PREMs in univariable analyses and multivariable analyses are shown in Table 3. .17 Bold indicated P < .05.
a Indicates statistically significant differences between groups (P-value�.05).
Ease of remembering information discussed during telehealth visit A total of 84.8% participants reported that it was very or somewhat easy to remember information discussed during their telehealth visit.In univariable analyses, participants reporting difficulty remembering information discussed during the telehealth visit were more likely to have inadequate health literacy (OR 0.

Usefulness of telehealth visits once COVID-19 pandemic is over
About 25% of individuals reported that telehealth visits would not be useful postpandemic.In univariable analyses, those who endorsed telehealth to be not useful postpandemic were more likely to have inadequate health literacy (OR 0.47, 95% CI, 0.26-0.84)or reported race as Other (OR 0.26, 95% CI, 0.07-0.90).Technology-related access was also a significant factor, with no access to video-enabled devices (OR 0.37, 95% CI, 0.17-0.80),no access to internet (OR 2.79, 95% CI, 1.20-6.49),and telephone versus video visit (OR 0.49, 95% CI, 0.28-0.87)associated with reporting telehealth visits being not useful once COVID is over.In multivariable models, only having 3 or more chronic conditions remained significantly associated with a higher likelihood to report usefulness of telehealth visits post-COVID (AOR 2.11, 95% CI, 0.99-4.62).

Discussion
We report on telehealth access and patient experience in a unique diverse cohort with chronic medical conditions and detailed sociodemographic and behavioral data in the Chicago area.Telehealth use was less common among participants of older age, fewer chronic conditions, and living above poverty level.Two-thirds of participants with a telehealth visit had their appointment via telephone versus video.Access to technology, such as video-enabled devices or internet, were barriers to participants having a video visit, as were inadequate health literacy, LEP, and lower SES (eg, less than high school education, income below poverty level).
Participants with a recent telehealth visit typically reported >2 visits within 4 months and overall positive experiences regarding quality (57.6%) and ease of navigating visits (88.9%), similar to results found in a 2020 systematic review which illustrated high patient and provider satisfaction with telehealth during the pandemic. 27Furthermore, 80% reported telehealth to be useful during the pandemic, and 75% thought it would be useful postpandemic.Most participants reported it was easy to describe their current health during a telehealth appointment and that it was easy to remember health information discussed during the appointment.Although this highlights the potential benefits and success of continued access to telehealth services, concerns around widening existing health disparities in healthcare access were also highlighted by our data.
Our study highlights continued health disparities across key sociodemographic characteristics (eg, English proficiency, ethnicity, SES, modality of telehealth) in healthcare access and patient experiences of telehealth visits.2][33][34][35] Challenges with technology access, use, and general readiness with telehealth have also been documented, with telehealth "unreadiness" more prevalent in patients who were older, men, not married, Black or Hispanic, had lower SES, and had poorer selfreported health. 36n sum, vulnerable populations, such as those with lower health literacy, LEP, low SES, Hispanic ethnicity, receiving care at a FQHC, or telephone visits may be more likely to report having less favorable telehealth experiences (eg, lower quality telehealth visits, less likely to remember telehealth visit information).This may indicate that patients who have greater difficulty understanding health information, whether because of lower health literacy or LEP, may be more likely to report difficulties remembering telehealth information.Thus, telehealth visits also may be less effective for these populations, given the increased likelihood of challenges in remembering telehealth information.Furthermore, perceptions of usefulness of telehealth post-COVID-19, were significantly associated with technology access and modality of telehealth visits (telephone).Patient barriers to technology and internet access, as well as ease of navigating technology, may be important considerations when evaluating reach and effectiveness of telehealth services.

Limitations
There were several limitations of this study.First, the C3 study surveyed patients who were participants with underlying health conditions actively enrolled in existing, NIH-sponsored studies or clinical trials in one, large US city, which limit the generalizability of findings.Second, this was a cross-sectional analysis of a longitudinal study.Further research is necessary to determine how these health disparities may have potentially changed over the continued pandemic.Third, the telehealth questionnaire did not inquire as to what device or video platform participants used for telehealth.Given that telehealth experiences may likely be influenced by what device and/or video platform a participant uses, this is a limitation of our analysis.
Despite these limitations, this study captured cognitive and psychological determinants of health (eg, CHAI) and their potential impact on patient experiences of telehealth access and use.Currently, limited research studies have been able to capture such detailed psychosocial determinants of health and telehealth PREMs.6][37][38][39] This makes it difficult to make concrete conclusions of effectiveness or on patients' experiences of telehealth services.
1][42][43][44] Clinicians should consider in-person visits, if feasible, for vulnerable patients who may be more at risk for appointment no-show and/or poorer quality visits when using telehealth (eg, low SES, LEP) or conducting patient visits in another language (ie, Spanish) for patients with LEP.Solutions might also benefit from optimizing patient follow-up, such as sending a summary of the provider's recommendations immediately after the visit by email, text, or letter; engaging home health or community health workers for follow-up; using home-based diagnostic equipment for follow-up; or connecting the patient with other ancillary professional care (eg, dietician, pharmacist).At a system or institutional-level consideration, also finding opportunities to evaluate a patients' technology/internet access and internet quality as part of the scheduling process and providing technology onboarding and support as needed. 45,46

Conclusion
COVID has accelerated and revolutionized telehealth, however, disparities in healthcare access and utilization have persisted, and perhaps been exacerbated with telehealth.In particular, patients with lower SES, LEP, and low health literacy may not have high-quality visits and have difficulties retaining instructions and information discussed during their telehealth visit.Health-systems and clinicians need to be aware of these vulnerabilities in order to build effective and equitable telehealth programs.
Northwestern University's Institutional Review Board serves as the single IRB of record, and all research staff maintain Human Subjects Training (Collaborative Institutional Training Initiative, ie, CITI).

Table 1 .
Characteristics of participants who had versus did not have a telehealth visit in past 4 months.

Table 2 .
Likelihood of most recent telehealth visit being video.

who had telehealth visit in past 4 months (n ¼ 342) Likelihood of most recent telehealth visit being video
a Indicates statistically significant differences between groups (P-value�.05).

Table 3 .
Bivariate and multivariable analyses of factors associated with telehealth PREMs.