Impact of Social Determinants and Digital Literacy on Telehealth Acceptance for Pediatric Cardiology Care Delivery during the Early Phase of the COVID-19 Pandemic

Abstract

Objective(s):

To determine if telehealth acceptance by parents of children with heart disease is predicted by sociodemographic and/or by parental digital literacy; and to assess parental perceptions of TH usability and reliability.

Study design:

We conducted a single center study comparing TH acceptance versus visit cancellation/rescheduling for pediatric cardiology visits during the early phase of the COVID19 pandemic.

All parent/guardians who consented to survey completion received a validated survey assessing their DL. Consenting parents who accepted TH received an additional validated survey assessing their perceptions of TH usability and reliability.

Results:

849 patients originally were scheduled for in-person visits between March 30 and May 8, 2020. TH acceptance was highest among younger, publically insured, Hispanic patients with primary diagnoses of arrhythmia/palpitations, chest pain, dysautonomia, dyslipidemia and acquired heart disease.

Among parents who completed surveys, a determinant of TH acceptance was DL. TH was determined to be a usable and reliable means for health care delivery.

Conclusion:

Although the potential for inequitable selection of TH due to sociodemographic factors exists, we found that such factors were not a major determinant for pediatric cardiology care within a large, diverse, free-standing pediatric hospital.

Infection due to COVID-19 was declared a global pandemic by the World Health Organization (WHO) in March 2020 (1). Across the country, outpatient health care visits were rapidly transitioned from in person visits to virtual visits, using a telehealth platform. The parent/guardians of pediatric and adolescent cardiology patients were offered TH visits in place of traditional in-person visits during the initial phase of the pandemic (2) (3). Other families chose to cancel and/or reschedule their child’s outpatient visit.

Use of a virtual-digital health platform for the delivery of care, particularly for pediatric cardiology patients, is not novel. What has been novel during the COVID-19 pandemic has been the relaxation of rules regarding provision of care across state lines and insurance reimbursements allowing for more children to be evaluated via TH visits and the offering of TH visits for children with complex congenital heart disease (CHD) (3) (4) (5). During the COVID19 pandemic, most centers used TH to decrease the burden of spread of the virus. However, acceptance of a virtual health platform model by all patients may not be equitable; potentially disadvantaging persons from under-resourced communities and/or families with language barriers. Inequitable acceptance of TH may also be influenced by parental -DL. The objectives of our study were to determine if TH acceptance by parent/guardians of children with heart disease is predicted by sociodemographic factors and/or by parental DL; and for the delivery of pediatric cardiology care, to assess parental perceptions of TH usability and reliability.

METHODS:

This study was approved by the Nemours Institutional Review Board prior to initiation of study procedures. During the early phase of the COVID-19 pandemic, between March 30 and May 8, 2020, children and adolescents were originally scheduled for in-person pediatric cardiology visits at Nemours but were transitioned to either TH or rescheduled/cancelled for visit dates after May 8, 2020. Consenting parent/guardians of these pediatric patients were the study survey participants.

Sociodemographic data, including age (categorized as <4, 4–12 and 13 years of age and above), sex, primary language, self-identified race, religion, state, county, and zip code were extracted from the electronic health record (EHR) of all patients originally scheduled for an in-person pediatric cardiology visit. Age at the time of scheduled visit (eg, date of TH visit or date of original visit between March 30th and May 8^th) was used. Persons of “Latino”, “Mexican”, “Puerto-Rican” description were classified as Hispanic ethnicity. Poverty status was classified utilizing zip code data. Poverty status =1 for persons residing within a community with ≥ 40% of families living below the federal poverty level and poverty status=0 for persons residing within communities with <40% of families living below the federal poverty level (6). Insurance type was relabeled as payer type (public, private, self-pay, missing/other). Insurance assigned to public payer type included: “Govt/Public”, “Medicaid”, “DE MCAID MGD”, “TRICARE”, and “MEDICARE”. Private payer type included “BCBS”, “COMMERCIAL”, “HUMANA MILITARY”, and “IBC”. Self-pay payer type included ”SELF-PAY”. Missing/other included: “missing”, ”NEMOURSFUNDING”, “MULTIPLAN”, and “OTHER”. Religion categories were simplified into Christian and non-Christian faith.

Visit types included new visit, follow-up, and pre-operative visit. Pre-operative visits (N=16, 1.9%) were excluded from multivariable analysis due to the small size. Primary diagnosis was assigned based upon manual review of the problem list for each patient scheduled for an in-person visit. The visit diagnosis is the reason for the visit as obtained from the EHR. A code was assigned (0–11) based upon the primary reason for the visit. (0) palpitations, supraventricular tachycardia, history of abnormal ECG (e.g., “first degree heart block”), arrhythmia (e.g., unspecified tachycardia, sick sinus syndrome, bradycardia, paroxysmal atrial fibrillation, ventricular tachycardia, long QT syndrome, pre-excitation syndrome) or palpitations; (1) congenital mitral insufficiency, malformation of coronary vessels, atrial septal defect, ventricular septal defect, double aortic arch, pulmonary valve stenosis, atrioventricular septal defect, patent ductus arteriosus, hypoplastic left heart syndrome, Ebstein anomaly; (2) unspecified cardiac murmur and innocent murmur; (3) chest pain and precordial pain; (4) cardiomyopathy, dilated cardiomyopathy, and pulmonary hypertension; (5) syncope and collapse, dizziness, orthostatic hypotension, postural orthostatic tachycardia syndrome; (6) essential hypertension and elevated blood pressure reading; (7) hyperlipidemia, pure hypercholesterolemia, familial hypercholesterolemia, hypertriglyceridemia; (8) myocarditis, Kawasaki disease, nonrheumatic aortic valve stenosis, mitral valve prolapse or mitral valve insufficiency; (9) chromosomal abnormality, Pompe disease, Hurler syndrome, screening for cardiovascular disorders, Rett syndrome, Mucopolysaccharidoses, Turner syndrome; (10) Marfan syndrome, aortic root dilation, possible Ehlers Danlos syndrome, bicuspid aortic valve/aortic root dilation; and (11) cough and shortness of breath. For simplicity of analysis, only a single diagnosis code was included in the analysis. Primary diagnoses were coded and formatted for ease of description (Figure 1).

FIGURE I: Social Determinants of Health and Digital Literacy.

849 patients were initially scheduled for in-person visits during the early phase of the COVID-19 pandemic, between March 30 and May 8, 2020. N=441 (52%) parents accepted TH. N=407 (48%) of parents/guardians chose to reschedule or cancel their child’s visit. Demographic and social factors associated with TH acceptance were identified. **Indicates variables predictive of TH acceptance by multivariable logistic regression (p <0.05). Not all parents accepted participation in the survey portion of the study. N=218 (26%) parents completed surveys to assess digital literacy (DL) and perceptions of TH usability.

An email was generated from REDCap and an invitation to participate in the survey portion of the study was distributed electronically to all parent/guardians of children and adolescents originally scheduled for at least 1-in-person visit in Pediatric Cardiology between March 30 and May 8, 2020. The invitation was sent in English or Spanish, based upon primary language listed in the EHR, and included a link to the electronic consent form. Parents/guardians were given the option to speak with a member of the study team regarding any questions/concerns prior to signing the electronic consent form. All parent/guardians who consented to participate were automatically directed within REDCap to the study survey(s) and received a $20 gift card upon survey completion.

DL was assessed via the Technological Ease and Computer-based Habits Inventory (TECHI) survey. The TECHI is a self-reported assessment of the frequency and facility with which one uses technology in their daily life, as well as the extent to which they perceive themselves as equipped to attend to technology-based matters. The questionnaire includes 20 items that are rated on a 0–5 scale with 0 being strongly disagree and 5 being strongly agree. Elevated scores on the TECHI reflect regular use of technology in one’s daily life, a high degree of perceived DL, and great overall comfort with technology. In contrast, low scores on the TECHI reflect minimal use of technology in one’s daily life, a low degree of perceived technological literacy, and limited overall comfort with technology. The TECHI survey was distributed to all consenting parents, including those who accepted TH visits as well as those who chose to cancel/reschedule previously scheduled visits.

Telehealth usability and reliability was assessed via the Telehealth Usability Questionnaire (TUQ) (7). The TUQ is a validated questionnaire used to assess perception of usefulness, ease of use and learnability, interface quality, interaction quality, reliability and satisfaction and future use of the TH platform. This survey was distributed to all families who chose and completed TH visits.

At the beginning of the pandemic, patient and families were contacted by schedulers and nursing staff at Nemours. A set script was utilized to inform patients and families that due to the COVID-19 pandemic it was recommended that appointments either occur virtually (via TH) or be rescheduled to a later date. Parent/guardians were contacted by phone. Families were assessed for accessibility to a smartphone, tablet, computer with camera, microphone, and speaker. They were asked if they had access to reliable WiFi or Cell Service. Email addresses were confirmed and if not correct, a new email address was obtained (Figure 4; available at www.jpeds.com). Families were then informed that they would receive an email from Nemours with subject line: “Nemours Video Visit-Action Required for Account Set Up.” Families were given detailed instruction on how to create a password and how to download “the free Nemours CareConnect app.” Families were informed to follow a series of prompts to enroll. Families were advised to download the app in advance “to get set up for your appointment. I would like to encourage you to do this as soon as we are off the phone to avoid any potential technical issues when your appointment time arrives.” Families were also provided with detailed instructions on what to expect on the day of the appointment: “you will see your appointment-click “start visit” and it will bring you into a virtual waiting room.” Families were encouraged to start the TH visit “5–10 minutes before your appointment time and stand by until the doctor connects with you.” Parent/guardians were also asked to have their child with them at the time that the TH visit started. Requests for new patient visits were provided a similar option: TH or to reschedule for a later date when in-person visits were available. Translation services were used to communicate with families using their preferred primary language.

Figure 4.

Caption: Decision tree for CARECONNECT NP.

Nemours utilizes Amwell Telemedicine and Technology Solutions for virtual care delivery via the CARECONNECT platform. Using this platform, providers and patients could enter a shared, secure videoconferencing room.

Univariate analyses of continuous data were carried out via Mann Whitney rank test for comparison of medians. Univariate analysis of categorical data was carried out using Chi Square test or Fisher exact (if the expected counts ≤5 for 25% or more of the m x n table). Multivariable logistic regression modeling the dichotomous outcome variable (TH vs cancelled/rescheduled visit) was carried out via a stepwise variable selection method. All statistical analyses were carried out using SAS 9.4 (Chapel Hill, North Carolina). Significance level of p < 0.05.

RESULTS:

The 849 patients originally scheduled for in-person pediatric cardiology were of median age of 10.2 years (IQR: 3.0, 15.3 years), predominantly male (52%), English speakers (95%), of white-race (59.6%), non-Hispanic ethnicity (89%), Christian faith (59%), resided in communities with fewer than 40% of households below the poverty line (99%), privately insured (52%), and resided in-state (58%). Poverty status was highest among Blacks and Hispanics (data not shown), however, only 6 (0.7%) participants met criteria for poverty. The largest percentage of patients had congenital heart disease (N=247, 30%), followed by palpitations/abnormal ECG (N=155, 19%), innocent murmur/murmur (N=102, 12%), dysautonomia/POTS/orthostatic hypotension (N=64, 7.7%), chest pain (N=59, 7.1%), dyslipidemia (N=34, 4.1%), cardiomyopathy/post heart transplant (N=24, 2.9%), acquired heart disease/Kawasaki disease (N=23, 2.8%), elevated blood/systemic hypertension (N=16, 1.9%), or connective tissue disease (N=16, 1.9%) (Data not shown).

Parent/guardians (N = 441; 52%) accepted TH visits and 408 (48%) cancelled or rescheduled visits until after May 8, 2020 (Table I). The majority of patients whose parent/guardian accepted TH visits were younger in age (median age 9.9 years versus 10.6 yrs., p=0.038; 60% under 4 years of age, 46% between 4 and 12 years, 51% 13 years of age and above; of Hispanic vs non-Hispanic ethnicity (68% vs 52%; p=0.0034); non-Christian vs Christian (58% vs 47%; p=0.0020); of out-of-state vs in-state residence (60% vs 46%, p<0.0001); with children and adolescents scheduled for new visits vs follow-up visits (56% vs 46%, p=0.0046). Patients with a primary diagnosis of abnormal ECG/arrhythmia, chest pain, dysautonomia, dyslipidemia, and acquired heart disease were also more likely to accept TH visits, and patients with cardiomyopathy, patients with syndromes and patients with known or suspected connective tissue disorders, less likely (Figure I and Table I).

Table I:

Demographic Data of Telehealth Scheduled versus Cancelled Patients

	Telehealth	Cancelled	p-Value
N	441 (52%)	408 (48%)
Age (years)			0.003
<4	147 (60%)	96 (40%)
4–12	127 (46%)	151 (54%)
≥13	167 (51%)	161 (49%)
Gender			0.93
Female	210 (52%)	193 (48%)
Male	231 (52%)	215 (48%)
Language			0.0013
English	407 (51%)	389 (49%)
Spanish	22 (54%)	19 (46%)
Other	12 (100%)	0 (0%)
Language (simplified)			0.87
English	407 (51%)	389 (49%)
Spanish	22 (54%)	19 (46%)
Race			0.45
Other	61 (54%)	53 (46%)
Black	126 (55%)	103 (45%)
White	254 (50%)	252 (50%)
Ethnicity			0.0034
Non-Hispanic	373 (52%)	346 (48%)
Hispanic	62 (68%)	29 (32%)
Religion			0.0012
Christian	235 (47%)	262 (53%)
Other	206 (59%)	146 (41%)
Poverty status			0.22*
Percentage of families living below the federal poverty level:
≥ 40% = 1	5 (83%)	1 (17%)
< 40% = 0	436 (52%)	407 (48%)
Payer Type**			0.068
Private	206 (48%)	224 (52%)
Public	198 (54%)	166 (46%)
State			<0.0001
Delaware	223 (46%)	262 (54%)
Out of State	218 (60%)	146 (40%)
Visit-Type			0.0046
New	247 (56%)	194 (44%)
Follow-Up	181 (46%)	211 (54%)
Primary Diagnosis			0.0011
0 Abnormal ECG/Arrhythmia	96 (60%)	63 (40%)
1 Congenital Heart Disease	137 (54%)	116 (46%)
2 Innocent Murmur	48 (47%)	55 (53%)
3 Chest Pain	32 (53%)	28 (47%)
4 Cardiomyopathy/Heart Transplant	11 (42%)	15 (58%)
5 Dysautonomia	37 (58%)	27 (42%)
6 Elevated BP/ Hypertension	8 (50%)	8 (50%)
7 Dyslipidemia/Hypertriglyceridemia	19 (58%)	14 (42%)
8 Acquired heart disease/Rheumatic Fever	16 (70%)	7 (30%)
9 Screening, Cancer, Syndromes	30 (37%)	52 (63%)
10 Connective Tissue Disorder	2 (12%)	14 (88%)
11 Unknown	5 (36%)	9 (64%)

^*Fisher’s exact.

^**Excluding self-pay (N=32) and missing payer type (N=23).

Payer type was significant if self-pay and “missing/other” payer type categories were excluded (p < 0.001). Ethnicity “missing”, “self-pay” payer type, and “missing/other” payer type were excluded from the logistic regression model.

Model fit was assessed via Hosmer-Lemeshow goodness of fit test. By multivariable logistic regression, significant predictors of TH acceptance include age <4 years versus age ≥13 years (OR 2.06, 95%CI: 1.34, 3.18), p <0.0001), Hispanic ethnicity (OR 1.77, 95% CI: 1.04, 3.03), p=0.036), new visit (OR 1.65, 95%CI: 1.2, 2.42, p=0.0029), primary diagnosis of abnormal ECG/arrhythmia/palpitations (OR 4.77, 95% CI: 1.30, 17.5), chest pain (OR 3.98, 95% CI: 1.005, 15.8), dysautonomia (OR 4.06, 95% CI: 1.03, 15.6), dyslipidemia/ hypertriglyceridemia (OR: 5.67, 95% CI: 1.3, 24.6), and acquired heart disease/rheumatic fever (OR: 6.92, 95% CI: 1.40, 34.2); public insurance vs private insurance coverage (OR 1.51, 95% CI: 1.09, 2.09, p=0.0124); as well as out of state vs DE residence (OR: 2.09, 95% CI: 1.51, 2.89, p <0.0001) (Figure 2). Religion was not a predictor of telemedicine acceptance by multivariable analysis and was thus removed from the model. Higher order associations (e.g., ethnicity*state; ethnicity*payer type; diagnosis*visit type, race*payer type, etc.) were not significant and were excluded from the model. Area under the curve for the receiver-operating curve for this model was 0.7 (good) for predicting TH acceptance (data not shown).

FIGURE II: Multivariable Logistic Regression Odds Ratios (Outcome Variable: Telehealth Acceptance).

Odds ratio and interquartile range. Predictors of TH acceptance. Legend: Logistic regression: excludes “other” visit-type, “missing” ethnicity, “missing and self-pay” insurance type, and Nemours’ Florida patients. The likelihood of a parent accepting a TH visit (vs rescheduling/cancelling) for a previously scheduled in-person pediatric cardiology visit during the early phase of the COVID-19 pandemic is modeled by the main effect variables: patient age (years), primary visit diagnosis, visit-type, ethnicity, insurance type and state.

Parents who completed surveys were predominantly White (77%), non-Hispanic (93%), English speakers (98%), Christian (64%), privately insured (62%) and their children were scheduled for follow-up pediatric cardiology visits (55%); 218 (26%) of parents chose to complete surveys (Figure 1 and Table 3 [available at www.jpeds.com]). Two surveys were distributed to parents who selected TH visits: TECHI and TUQ. One survey, the TECHI, was distributed to parents who chose to cancel/reschedule.

Table 3.

Demographic Data of Survey Completers versus Non-completers

	Completers	Non-Completers	p-Value
N, Total= 849	218	631
Age (years)	9.07	10.6	0.14
Gender			0.098
Female	114 (52%)	289 (46%)
Male	104 (48%)	342 (54%)
Language			0.044 (Fisher)
English	212 (98%)	584 (94%)
Spanish	5 (2%)	36 (6%)
Race (simplified)			0.0030
Black	44 (23%)	185 (34%)
White	150 (77%)	356 (66%)
Ethnicity (simplified)			0.035
Non-Hispanic	192 (93%)	527 (87%)
Hispanic	15 (7%)	76 (13%)
Religion			0.022
Christian	139 (64%)	358 (57%)
Other	79 (36%)	273 (43%)
Payer Type (simplified)			0.0062
Private	129 (62%)	301 (51%)
Public	78 (38%)	286 (49%)
State			0.94
Delaware	125 (57%)	360 (57%)
Out of State	93 (43%)	271 (43%)
Visit-Type			0.0047
New	97 (45%)	344 (56%)
Follow-Up	120 (55%)	272 (44%)
Primary Diagnosis			0.49
0 Abnormal ECG/Arrhythmia	41 (18%)	118 (19%)
1 Congenital Heart Disease	70 (32%)	183 (29%)
2 Innocent Murmur	28 (13%)	75 (12%)
3 Chest Pain	10 (4%)	50 (8%)
4 Cardiomyopathy/Heart Transplant	9 (4%)	17 (3%)
5 Dysautonomia	15 (7%)	49 (8%)
6 Elevated BP/ Hypertension	4 (2%)	12 (2%)
7 Dyslipidemia/Hypertriglyceridemia	10 (4%)	23 (4%)
8 Acquired heart disease/Rheumatic Fever	5 (3%)	18 (3%)
9 Screening, Cancer, Syndromes	18 (10%)	64 (10%)
10 Connective Tissue Disorder	6 (2%)	10 (2%)
11 Unknown	2 (2%)	12 (2%)

Parents of children scheduled for visits between March 30 and May 8, 2020 were offered the opportunity to complete surveys. N=218 parent/guardians completed surveys while N=631 did not choose to complete surveys.

Of the parents who completed TECHI survey questions, TH acceptance was associated with “regular use of Skype or another video-conferencing service” and general agreement that “telehealth is helpful” (Table II). Other questions related to DL were not associated with TH acceptance (e.g., smartphone, computer, tablet, email, text message, and Bluetooth use) or perceived expertise with use of technology (e.g., ability to solve technology related problems) (Table II). A separate logistic regression carried out to include TECHI survey response data, followed by stepwise selection of independent variables (data not shown), revealed that the strongest predictors of TH acceptance among survey completers included strong agreement vs strong disagreement with TECHI6 “I regularly use Skype or another videoconferencing service” (OR: 4.23, 95% CI: 1.51, 11.8, p=0.012) and Hispanic vs non-Hispanic ethnicity (OR 7.8, 95%CI: 1.6, 38.2, p=0.011). Among survey completers, patient age in years, payer type, visit type, primary diagnosis, TECHI10 “Technology is helpful”, and state of residence were no longer predictive of TH acceptance.

Table II.

Digital Literacy (DL) Assessment: Technological Ease and Computer-based Habits Inventory (TECHI) Survey Results

	Telehealth	Cancelled	p-Value
N	105 (48%)	113 (52%)
I use a Smartphone on a daily basis (TECHI 1)			0.53*
Mildly Agree	1 (33%)	2 (67%)
Agree	1 (20%)	4 (80%)
Strongly Agree	103 (49%)	107 (51%)
I use a computer or tablet on a daily basis (TECHI 2)			0.071*
Strongly Disagree	7 (58%)	5 (42%)
Disagree	6 (55%)	5 (45%)
Mildly Disagree	10 (59%)	7 (41%)
Mildly Agree	12 (48%)	13 (52%)
Agree	5 (20%)	20 (80%)
Strongly Agree	65 (51%)	63 (49%)
I check my email on a daily basis (TECHI 3)			0.91*
Strongly Disagree	1 (100%)	0 (0%)
Disagree	2 (67%)	1 (33%)
Mildly Disagree	3 (50%)	3 (50%)
Mildly Agree	8 (44%)	10 (56%)
Agree	14 (54%)	12 (46%)
Strongly Agree	77 (47%)	87 (53%)
I send emails on a daily basis (TECHI 4)			0.13*
Strongly Disagree	9 (90%)	1 (10%)
Disagree	7 (44%)	9 (56%)
Mildly Disagree	14 (44%)	18 (56%)
Mildly Agree	15 (47%)	17 (53%)
Agree	10 (38%)	16 (62%)
Strongly Agree	50 (49%)	52 (51%)
I send text messages on a daily basis (TECHI 5)			0.49*
Disagree	1 (100%)	0 (0%)
Mildly Disagree	2 (67%)	1 (33%)
Mildly Agree	4 (80%)	1 (20%)
Agree	6 (43%)	8 (57%)
Strongly Agree	92 (47%)	103 (53%)
I regularly use Skype or another video-conferencing service (TECHI 6)			0.012*
Strongly Disagree	10 (33%)	20 (67%)
Disagree	16 (59%)	11 (41%)
Mildly Disagree	13 (38%)	21 (62%)
Mildly Agree	23 (49%)	24 (51%)
Agree	7 (29%)	17 (71%)
Strongly Agree	36 (64%)	20 (36%)
I regularly use Bluetooth Technology or pair my devices to one another (TECHI 7)			0.59*
Strongly Disagree	8 (44%)	10 (56%)
Disagree	6 (46%)	7 (54%)
Mildly Disagree	11 (65%)	6 (35%)
Mildly Agree	11 (38%)	18 (62%)
Agree	11 (42%)	15 (58%)
Strongly Agree	58 (50%)	57 (50%)
People close to me would refer to me as plugged in			0.070*
Strongly Disagree (TECHI 8)	5 (29%)	12 (71%)
Disagree	8 (62%)	5 (38%)
Mildly Disagree	17 (71%)	7 (29%)
Mildly Agree	22 (39%)	34 (61%)
Agree	23 (47%)	26 (53%)
Strongly Agree	30 (51%)	29 (49%)
I always have my cell phone at my side (TECHI 9)			0.98 *
Strongly Disagree	0 (0%)	1 (100%)
Disagree	2 (40%)	3 (60%)
Mildly Disagree	6 (55%)	5 (45%)
Mildly Agree	10 (48%)	11 (52%)
Agree	26 (46%)	31 (54%)
Strongly Agree	61 (50%)	62 (50%)
Technology is helpful (TECHI 10)			0.0014*
Mildly Disagree	0 (0%)	2 (100%)
Mildly Agree	4 (25%)	12 (75%)
Agree	14 (35%)	26 (65%)
Strongly Agree	87 (54%)	73 (46%)
People close to me would refer to me as tech savvy (TECHI 11)			0.51*
Strongly Disagree	2 (33%)	4 (67%)
Disagree	8 (47%)	9 (53%)
Mildly Disagree	14 (50%)	14 (50%)
Mildly Agree	25 (44%)	32 (56%)
Agree	25 (43%)	33 (57%)
Strongly Agree	31 (60%)	21 (40%)
If we have technology related issues in my household, I am the one to address them (TECHI 12)			0.52
Strongly Disagree	11 (44%)	14 (56%)
Disagree	10 (38%)	16 (62%)
Mildly Disagree	16 (57%)	12 43%)
Mildly Agree	17 (45%)	21 (55%)
Agree	25 (53%)	22 (47%)
Strongly Agree	26 (48%)	28 (52%)
If we have technology related issues in my household, I ask someone to address them (TECHI 13)			0.33
Strongly Disagree	17 (50%)	17 (50%)
Disagree	19 (59%)	13 (41%)
Mildly Disagree	13 (54%)	11 (46%)
Mildly Agree	20 (48%)	22 (52%)
Agree	12 (28%)	31 (72%)
Strongly Agree	24 (56%)	19 (44%)
I am easily frustrated by technology (TECHI 14)			0.096*
Strongly Disagree	31(65%)	17(35%)
Disagree	25 (50%)	25 (50%)
Mildly Disagree	20 (47%)	23 (53%)
Mildly Agree	21 (37%)	36 (63%)
Agree	6 (46%)	7 (54%)
Strongly Agree	2 (29%)	5 (71%)
When something goes wrong with my computer, I can fix it (TECHI 15)			0.49*
Strongly Disagree	7 (44%)	9 (56%)
Disagree	7 (35%)	13 (65%)
Mildly Disagree	21 (57%)	16 (43%)
Mildly Agree	34 (44%)	43 (56%)
Agree	26 (57%)	20 (43%)
Strongly Agree	10 (45%)	12 (55%)
Technology creates more problems than it solves (TECHI 16)			0.25*
Strongly Disagree	35 (54%)	30 (46%)
Disagree	35 (52%)	32 (48%)
Mildly Disagree	19 (49%)	20 (51%)
Mildly Agree	13 (39%)	20 (61%)
Agree	3 (27%)	8 (73%)
Strongly Agree	0 (0%)	3 (100%)
When I don’t know how to do something on my Computer, I can figure out how to do it on my own with the help of online resources (TECHI 17)			0.16*
Strongly Disagree	4 (100%)	0 (0%)
Disagree	5 (45%)	6 (55%)
Mildly Disagree	9 (47%)	10 (53%)
Mildly Agree	24 (45%)	29 (55%)
Agree	31 (44%)	40 (56%)
Strongly Agree	32 (53%)	28 (47%)
I have patience when it comes to technology - related issues (TECHI 18)			0.16*
Strongly Disagree	3 (50%)	3 (50%)
Disagree	4 (27%)	11 (73%)
Mildly Disagree	14 (48%)	15 (52%)
Mildly Agree	28 (45%)	34 (55%)
Agree	31 (46%)	37 (54%)
Strongly Agree	25 (66%)	13 (34%)
When I want to learn about something on the Internet, I can find it (TECHI 19)			0.25*
Strongly Disagree	0 (0%)	1 (100%)
Disagree	1 (100%)	0 (0%)
Mildly Disagree	1 (25%)	3 (75%)
Mildly Agree	5 (29%)	12 (71%)
Agree	28 (46%)	33 (54%)
Strongly Agree	70 (52%)	64 (48%)
When I encounter a technology-related problem, I give up and walk away (TECHI 20)			0.18*
Strongly Disagree	49 (58%)	36 (42%)
Disagree	29 (45%)	36 (55%)
Mildly Disagree	16 (39%)	25 (61%)
Mildly Agree	10 (45%)	12 (55%)
Agree	1 (20%)	4 (80%)

^*Fisher’s exact categorical analysis. Test for trend was performed p-values represent composite assessment trend analysis toward strongly agree (significance, p<0.05).

Among parents who chose TH visits, we assessed perception of TH usability and reliability (Figure 4). Perceptions of usefulness, ease of use, learnability, interaction quality, and satisfaction with the TH experience were high. Survey responses were the lowest in answer to the question of “I think the visits provided over the TH system are the same as in-person visits” (TUQ15) (Figure III). There are very few reported cases of “whenever I made a mistake using the system, I could recover easily and quickly” (TUQ16) and “the system gave error messages that clearly told me how to fix problems” (TUQ17) with nearly 50% of TUQ16 and nearly 80% of TUQ17 survey participants responding that these questions were not applicable to their experience. Of those who responded, the majority agreed (data not shown).

FIGURE III: Telehealth Usability Questionnaire Results.

Usefulness (1–3). Ease of Use and Learnability (4–6). Interface Quality (7–10). Interaction Quality (11–14). Reliability (15–17). Satisfaction and Future Use (18–21). Usefulness questions: TUQ1: “Telehealth improves my access to healthcare services”‘ TUQ2: “Telehealth saves me time traveling to a hospital or specialist clinic”, TUQ3: “Telehealth provides for my healthcare needs”. Ease of Use and Learnability: TUQ4: “It was simple to use this system”; TUQ5 “It was easy to learn to use the system”; TUQ6 “I believe I could become productive quickly using this system”. Interface Quality: TUQ7 “The way I interact with this system is pleasant”, TUQ8 “I like using the system”, TUQ9 “The system is simple and easy to understand”, TUQ10” This system is able to do everything I would want it to be able to do”. Interaction Quality: TUQ11” I could easily talk to the clinician using the telehealth system”, TUQ12” I could hear the clinician clearly using the telehealth system”, TUQ13” I felt I was able to express myself effectively”, TUQ14” Using the telehealth system, I could see the clinician as well as if we met in person”. Reliability: TUQ15 “I think the visits provided over the telehealth system are the same as in-person visits”, TUQ16 “Whenever I made a mistake using the system, I could recover easily and quickly”, TUQ17 “The system gave error messages that clearly told me how to fix problems”. Satisfaction and Future Use: TUQ18 “I feel comfortable communicating with the clinician using the telehealth system”, TUQ19” Telehealth is an acceptable way to receive healthcare services”, TUQ20” I would use telehealth services again”, TUQ21 “Overall, I am satisfied with this telehealth system”.

DISCUSSION:

Rapid adaptation to a changing health care environment is successful if it does not further disadvantage under-resourced communities or exacerbate the divide in health access by race, ethnicity, sex, primary language, poverty status or geographic location. Social determinants of health (SDoH) are defined, by the World Health Organization, as “the circumstances in which people are born, grow, live, work and age and the systems in place to deal with illness.” (1) (8) Traditionally, SDoH have been defined by factors such as poverty status/wealth, race, ethnicity, language, access to care, and residential environment (8). Among children with CHD that SDoH may negatively affect outcomes (9). Some studies of adults have demonstrated that factors such as employment status, primary language, and race/ethnicity may negatively impact selection of TH (10) (11); speaking a language other than English increasing the chance that a person would find it difficult to navigate the application used to access the virtual visit (10). However, TH may be a suitable alternative for making health care more available to under-resourced communities, particularly, rural communities (12). Delays in care and concerns regarding the potential for biased care have negatively impacted ethnic and racial minority communities and concerns that transition to TH could further delay care have been raised (13). However, we found that SDoH did not result in cancelled/rescheduled visits or refusal to accept TH among such groups. We found either no difference in acceptance or that potentially under-resourced portions of the population (e.g., publically insured), including Ethnic minorities and persons whose primary language was not English, were more likely to accept TH visits.

The choice to accept TH may be more suitable for particular visit diagnoses. TH does not offer the opportunity for a detailed physical examination or even real-time testing (e.g., ECG, echocardiogram). However, some diagnoses, including previously diagnosed arrhythmia, concern regarding abnormal ECG, chest pain, dysautonomia and dizziness, dyslipidemia and acquired heart disease may be particularly suitable for care delivery via a virtual platform. Patients with a history or concern for arrhythmia or a chief complaint of chest pain can undergo a review of their symptoms. Remote wearable devices can be mailed to the home-including Holters and event recorders for patients complaining of palpitations. The American College of Cardiology has advocated for the use of remote cardiac-rhythm monitoring technologies during the pandemic (14). Patients in need of follow-up regarding lifestyle interventions, review of previously obtained laboratory testing and assessment of medication compliance, when appropriate, such as patients with dyslipidemia. Patients in need of adjustments to medication and follow-up of symptoms are a suitable population to adopt TH.

Advantages of TH include less time away from home, decreased travel, less time missed from work due to decreased travel, and lower risk of spread of COVID-19.We found by univariate and multivariable analysis that Hispanic ethnicity was associated with greater TH acceptance (N=62 vs N=29; Table 1). Race, however, was not a significant predictor. There have been limited studies assessing how to deliver care most effectively to racially and ethnically diverse populations (15) (16).

Insurance payer type also did not influence TH acceptance. Rules regarding reimbursement for TH visits were relaxed and this may have contributed to greater willingness on the part of families and referring providers (referring for new pediatric cardiology visits) to accept and recommend TH visits, respectively (2) (5).

Surveys completed by parents of children with acquired and congenital heart disease revealed that experience with a videoconferencing platform increased the chances of acceptance of TH. Our study demonstrates that DL may be key driver in TH acceptance. Similarly, we found that parents who perceived of technology positively were also more likely to choose TH. Results related to technological ease and telehealth usability must be considered with caution as most survey completers were white, Christian, English speaking, and privately insured. Although concerns may arise regarding access to internet services, more recent data demonstrates that as of 2017, 80% or more of households have access to some form of internet; broadband or via use of a cellphone. (17).

Our study was limited in that we evaluated data from a single center. Unfortunately, the impact of DL on the larger population of non-English speaking, non-white, non-Christian, publicly insured parents could not be assessed (data not shown) as these families were less likely to complete surveys. Finally, although requests for study participation and surveys were largely distributed electronically, potentially not reaching those without access to the internet or those who may have the lowest DL, all families who did not respond to email requests were contacted by phone. Printed consents and surveys were made available.

An additional limitation is that it is possible that parents may have chosen telehealth because they felt that it was “the least worst” of the two options rather than truly accepting it. Outside of this initial phase of COVID-19 and under usual circumstances it is unclear whether the same factors identified in our study would predict TH acceptance.

We found that TH is a useable and reliable method for the delivery of health care. The fact that families were provided assistance with establishing TH accounts, translation services were made available during the scheduling call for Spanish Speakers (as well as for other languages), and immediate assistance was available during the TH visit if technical issues arose, made use of this technology more acceptable. We did not find that SDoH negatively impact TH acceptance and key drivers for acceptance of TH have yet to be fully elucidated.

List of Abbreviations:

AHA

American Heart Association

CHD

Congenital Heart Disease

COVID-19

Corona virus disease 2019

ECG

Electrocardiogram

EHR

Electronic Health Record

REDCap

Research Electronic Data Capture

SARS-CoV-2

Severe Acute Respiratory Syndrome Corona Virus 2

SAS

Statistical Analysis System

SDoH

Social Determinants of Health

TECHI

Technological Ease and Computer-based Habits Inventory

TUQ

Telehealth Usability Questionnaire

US

United States

WHO

World Health Organization