Differential Impact of Virtual Family-Centered Rounds in the Neonatal Intensive Care Unit by Social Factors: A Post Hoc Subgroup Analysis

Abstract

Background:

Barriers to attending family-centered rounds (FCR) exist for socially disadvantaged families. Using telehealth to conduct virtual FCR could potentially promote equitable parent/guardian FCR access. The objective of this work was to assess whether the effects of a virtual FCR intervention on parent FCR attendance varied by subgroups defined by social factors.

Methods:

We conducted a post hoc analysis of a randomized controlled trial of virtual FCR in the neonatal intensive care unit. Parents of intervention arm infants were invited to participate in virtual FCR plus usual care; control arm infants received usual care. Participants were analyzed according to the assigned group and by race/ethnicity, insurance, mother’s education, and neighborhood health conditions. We used Poisson regression to estimate and compare FCR parent attendance rates. Heterogeneity of intervention effects was assessed using interaction terms to evaluate the relative benefit of the intervention in increasing parent FCR attendance.

Results:

We included all enrolled trial subjects (74 intervention, 36 control). Intervention arm infants had 3.36 (95% confidence interval [CI]: 2.66–4.23) times the FCR parent attendance rate of subjects in the control arm. Compared with the corresponding reference subgroup, intervention benefits were 2.15 times (95% CI: 1.30–3.56) better for racial/ethnic minorities, 3.08 times (95% CI: 1.59–5.95) better for those with private insurance, 2.68 times (95% CI: 1.12–6.40) better for those whose mother reported no college education, and 4.14 times (95% CI: 2.07–8.25) better for those from a neighborhood with worse health conditions.

Conclusions:

Virtual FCR improved parent FCR attendance overall, with even greater benefits for certain subgroups. Further research is needed to mitigate the differential benefit demonstrated for privately insured subjects.

Introduction

Family-centered, multidisciplinary bedside rounds (FCR) promote effective communication, active family engagement, and the core values of humanism in pediatrics.^1,2 Recognized benefits of FCR include fewer harmful errors, improved family understanding of their child’s condition and care treatment plan, reduced parental anxiety, increased parental trust in providers, and shortened hospital stays.^3–11 For these reasons, the American Academy of Pediatrics has endorsed FCR as best practice care for hospitalized children for over 20 years.^12,13

However, FCR are limited to patients whose families can be present at the bedside during the rounding process. Despite over 90% of parents wanting to attend FCR,^9,14 actual attendance proportions range from 19% to 66% (median: 52%).^5,9,15–20 Barriers to FCR attendance include competing caregiving and work duties, living too far from the hospital, lack of transportation, and high travel costs that prevent families from being at the bedside during FCR.^11,17 Families with fewer resources and families living further from the hospital face disproportionate challenges to attend FCR encounters. The requirement for physical presence in the hospital for standard FCR poses a dilemma for working parents and those with other caregiving responsibilities. Financial consequences from missed workdays or childcare costs unjustly disadvantage underresourced individuals in fully engaging in their child’s care. Due to these disproportionate challenges and additional factors stemming from systemic racism and classism, inequities in FCR access have been demonstrated along lines of race, education, and language.^17,21–25

Telehealth interventions have the potential to mitigate these challenges and thus support parents or guardians (“parents” hereafter) in attending FCR. To explore this promising solution, our team conducted a pilot randomized trial of virtual FCR in the neonatal intensive care unit (NICU).²⁶ Parents assigned to the intervention arm had the option to use Health Insurance Portability and Accountability Act compliant technology to be virtually present at the bedside in the NICU in order to virtually participate in FCR. This trial found that the option for parents to have a virtual bedside presence for FCR increased parent attendance 3.4-fold.²⁶

In addition, virtual FCR has potential to address some of the healthcare racial and socioeconomic inequities that result in the race-, education-, and language-based FCR access inequities.^17,21–25 Because virtual FCR allows parents to join FCR without being physically in the hospital, it helps parents overcome obstacles like competing caregiving and work responsibilities, time constraints, and travel challenges. Using data from the previously reported virtual FCR trial,²⁶ we aimed to examine the variation in intervention effects on parent FCR attendance among participant subgroups based on social factors.

Methods

STUDY DESIGN AND PARTICIPANTS

We conducted a post hoc subgroup study of the virtual FCR trial, a pilot randomized trial conducted in the NICU. The methods of the virtual FCR trial have previously described in detail.²⁶ In brief, we randomized subjects with a 2:1 intervention-to-control arm ratio. Parents of subjects assigned to the intervention arm could participate in virtual FCR as frequently as they chose. They could also participate in FCR in person, if desired. Parents of subjects assigned to the control arm received standard of care; they had the option to participate in FCR in person or to not participate in FCR. The original trial was approved by the University of California Davis Institutional Review Board and was registered with ClinicalTrials.gov (NCT04265677). The Institutional Review Board approved this study with a waiver of consent to collect data from observations of FCR and the electronic health record.

We examined intervention effects by subgroups for the outcome of parent FCR attendance proportion, which was the number of weekday round encounters with at least one parent present—either virtually or in person—divided by the infant’s total number of weekday round encounters. Subgroups of interest were defined a priori and consisted of the following:

• •Subgroups according to race and ethnicity, categorized as “non-Hispanic White subjects” or “racial/ethnic minority subjects.” We defined racial/ethnic minorities as infants with a documented identity in the electronic health record that was other than non-Hispanic White.
• •Subgroups according to insurance, categorized as “private insurance” or “public insurance” as documented in the electronic health record.
• •Subgroups according to mother’s highest education attainment, categorized as “some college education or more” or “no college education.” These data were obtained via parent-reported survey data collected during the original trial.
• •Subgroups according to neighborhood health conditions, categorized as “lowest quartile” (≤25th percentile) or “higher quartiles” (>25th percentile) using the California Healthy Places Index (HPI) Score²⁷ associated with the residential address documented in the electronic health record.

STUDY INTERVENTION

For the virtual FCR intervention, we used Zoom (San Jose, CA) as the secure software interface. Parents downloaded Zoom onto their personal computer or smart device. The NICU care team used Zoom on a computer on wheels with a speaker and pan-tilt-zoom camera. A research coordinator messaged parents with ∼15 minute warnings prior to the start of rounds for their infant and admitted parent(s) from the virtual waiting room to establish a bidirectional audio and visual connection. Once the medical team was at the bedside, the coordinator admitted the parents(s) from the Zoom waiting room to establish the real-time, bidirectional audio and visual connection. FCR then proceeded in usual fashion with a neonatologist, neonatal fellow, two pediatric residents, charge nurse, bedside nurse, pharmacist, dietician, and social worker.

ANALYSIS

We examined participant data according to their assigned group (intervention versus control). We reported subgroup-specific descriptive statistics and summarized effect sizes (with 95% confidence interval [CI]) with differences in proportions. To account for increased exposure (opportunity) among subjects with longer NICU hospitalizations, we used Poisson regression to estimate and compare FCR parent attendance rates, using the logarithm of the number of weekdays as offset terms. All available data were analyzed; we did not use imputation for nonresponse or missing data. Heterogeneity of intervention effects was assessed using interaction terms to evaluate the relative benefit of the intervention in increasing parent FCR attendance. A separate model was fit for each subgroup variable that included an interaction term for the main effect of the intervention and neighborhood health conditions.

Results

We included all enrolled trial subjects (74 intervention, 36 control). The intervention and control groups were similar in regard to each measured infant and maternal characteristics (Table 1). Intervention arm infants had 3.36 (95% CI: 2.66–4.23) times the FCR parent attendance rate of subjects in the control arm. Fig. 1 shows subgroup analyses of the intervention effect on the FCR parent attendance rate.

Table 1.

Characteristics of Subjects by Intervention Versus Control Group

	INTERVENTION n = 74		CONTROL n = 36
Infant Characteristics
Gestational age, weeks, mean (95% CI)	35.6	(34.6–36.7)	34.7	(33.1–36.3)
Admission age, days, median (IQR)	0	(0–1)	0	(0–0)
Sex, n (%)
Male	37	(50.0)	21	(58.3)
Female	37	(50.0)	15	(41.7)
Race, n (%)a
White	37	(50.0)	15	(41.7)
Asian	8	(10.8)	5	(13.9)
Black	4	(5.4)	3	(8.3)
Other or multiple	6	(8.1)	7	(19.4)
Unknown	19	(25.7)	6	(16.7)
Ethnicity, n (%)a
Not Hispanic	51	(68.9)	26	(72.2)
Hispanic	20	(27.0)	8	(22.2)
Unknown	3	(4.0)	2	(5.6)
Insurance, n (%)
Public	45	(60.8)	24	(66.7)
Private	19	(25.7)	7	(19.4)
Other	10	(13.5)	5	(13.9)
Birth weight, kilograms, mean (95% CI)	2.5	(2.3–2.7)	2.4	(2.1–2.7)
Ventilator use, n (%)	12	(16.4)	8	(22.9)
Disposition, n (%)
Discharged home	56	(75.7)	28	(77.8)
Transfer to other unit or hospital	11	(14.9)	7	(19.4)
Still in NICU at end of study	3	(4.0)	1	(2.8)
Died	4	(5.4)	0	(0.0)
Maternal characteristics
Maternal age, mean (95% CI)	30.3	(28.9–31.7)	29.9	(28.1–31.7)
Maternal parity, median (IQR)	2	(1–3)	1	(1–3)
Maternal education, n (%)
No college	20	(27.0)	7	(20.0)
Some college or more	26	(35.1)	14	(40.0)
Unknown	28	(37.8)	14	(40.0)
Neighborhood condition, n (%)b
1st quartile (worst health)	20	(28.6)	10	(29.4)
2nd quartile	21	(30.0)	12	(35.3)
3rd quartile	17	(24.3)	8	(23.5)
4th quartile (best health)	12	(17.1)	4	(11.8)
Distance, miles, median (IQR)c	39.9	(18.7–78.2)	30.6	(16.4–63.4)

^a Subgroups according to race and ethnicity, categorized as “racial/ethnic minority subjects” (n = 34 intervention, n = 17 control) or “non-Hispanic White subjects” (n = 30 intervention, n = 14 control); racial/ethnic minorities defined as other than non-Hispanic White.

^b Neighborhood health condition quartiles based on the California Healthy Places Index. Among the 110 subjects, 104 (94.5%) resided in California.

^c Distance was the shortest driving distance from the home to the study hospital.

CI, confidence interval; IQR, interquartile range; NICU, neonatal intensive care unit.

Fig. 1. Subgroup analyses of the intervention effect on FCR parent attendance rate. Neighborhood health condition quartiles based on the California Healthy Places Index. Logarithmic scale is employed on the horizontal axis. Adjusted rate ratios estimated in separate Poisson regression models for each of the top three subgroups, each model adjusting for neighborhood health condition. Estimates favor the intervention for all subgroups. FCR, family-centered rounds.

Subgroup analysis according to race and ethnicity demonstrated relative benefit for increasing parent FCR attendance in racial/ethnic minorities. Among non-Hispanic White subjects, intervention arm infants had 2.61 (95% CI: 1.94–3.51) times the FCR parent attendance rate of subjects in the control arm. Among racial/ethnic minority subjects, intervention arm infants had 5.62 (95% CI: 3.74–8.44) times the FCR parent attendance rate of subjects in the control arm. The intervention effect was 2.15 times (95% CI: 1.30–3.56) more beneficial for racial/ethnic minorities than for non-Hispanic White subjects. Relative to non-Hispanic White subjects, the minority FCR attendance relative rate was 0.32 (95% CI: 0.20–0.50) in the control arm. This relative rate increased to 0.68 (95% CI: 0.56–0.83) in the intervention arm.

Subgroup analysis according to insurance demonstrated relative benefit of the intervention in increasing parent FCR attendance for subjects with private insurance. Among subjects with private insurance, intervention arm infants had 9.87 (95% CI: 5.67–17.18) times the FCR parent attendance rate of subjects in the control arm. Among subjects with public insurance, intervention arm infants had 3.21 (95% CI: 2.24–4.59) times the FCR parent attendance rate of subjects in the control arm. The intervention effect was 3.08 times (95% CI: 1.59–5.95) more beneficial for privately insured subjects than for publicly insured subjects. Relative to privately insured subjects, the publicly insured FCR attendance relative rate was 1.74 (95% CI: 1.39–2.19) in the control arm. This relative rate was decreased to 0.57 (95% CI: 0.30–1.05) in the intervention arm.

Regarding maternal education, subgroup analysis demonstrated relative benefit of the intervention in increasing parent FCR attendance for those whose mother reported no college education. Among subjects whose mother reported some college education or more, intervention arm infants had 2.91 (95% CI: 2.18–3.89) times the FCR parent attendance rate of subjects in the control arm. Among subjects whose mother reported no college education, intervention arm infants had 7.81 (95% CI: 3.44–17.73) times the FCR parent attendance rate of subjects in the control arm. The intervention effect was 2.68 times (95% CI: 1.12–6.40) more beneficial for those whose mother reported no college education than for those with some college education. Relative to the college education group, the no college education group’s FCR attendance relative rate was 0.22 (95% CI: 0.09–0.50) in the control arm. This relative rate increased to 0.58 (95% CI: 0.46–0.73) in the intervention arm.

Subgroup analysis according to neighborhood health conditions demonstrated relative benefit of the intervention in increasing parent FCR attendance for subjects in the lowest quartile of HPI score, indicating worse health conditions. Among subjects with a residence in a neighborhood with a higher HPI (better health conditions), intervention arm infants had 2.45 (95% CI: 1.90–3.16) times the FCR parent attendance rate of subjects in the control arm. Among subjects from a lower HPI neighborhood (worse health conditions), intervention arm infants had 10.15 (95% CI: 5.34–19.26) times the FCR parent attendance rate of subjects in the control arm. The intervention effect was 4.14 times (95% CI: 2.07–8.25) more beneficial for those from a neighborhood with worse health conditions than for those from a neighborhood with better health conditions. Relative to the better health conditions group, the worse health conditions group’s FCR attendance relative rate was 0.23 (95% CI: 0.12–0.44) in the control arm. This difference resolved in the intervention arm; the relative rate became 0.94 (95% CI: 0.76–1.15).

Discussion

This post hoc subgroup study of the virtual FCR trial suggests that the impact of virtual FCR on parent FCR attendance varies across participant subgroups based on social factors. The intervention of virtual FCR—having the option to attend FCR virtually in addition to in person—demonstrated improved parent FCR attendance overall, with greater benefits for racial and ethnic minorities, those with no college education, and those from neighborhoods with worse health conditions. This study suggests that the virtual FCR intervention mitigates race-, education-, and neighborhood health condition-based FCR attendance difference. In contrast, however, the intervention effect demonstrated a differential benefit for privately insured subjects. Overall, virtual FCR is a promising tool to promote more equitable parent FCR access to engage in their infant’s care during NICU hospitalization.

Considering the expanding body of evidence indicating that telehealth tools may exacerbate inequities in health care access, the insurance-based finding that virtual FCR had greater benefit in increasing FCR attendance for patients with private insurance relative to public insurance warrants further research to examine and address this difference. Having private versus public insurance can be a proxy for socioeconomic status, and exploration to identify the potential economic and social factors contributing to this insurance-based inequity is needed. Prior telehealth research has demonstrated that having public insurance is associated with lower telehealth use.^28–33 Previously identified barriers include limited access to devices and broadband connection required for telemedicine visits, low digital literacy, selective offering of telemedicine, lack of private space for conducting telemedicine visits, and concern about availability and quality of interpreter services for telemedicine visits.^33–36 These drivers could have potentially contributed to the insurance-based findings of this present study. In addition, individuals with private insurance might have more flexible schedules that permit them to join FCR virtually. However, those with public insurance might have, for example, a work schedule that does not permit the flexibility to step away to join FCR virtually. Despite these important concerns, virtual FCR did demonstrate greater benefit in increasing parent FCR attendance for groups previously recognized to be at risk for low telehealth access.

Previous research has demonstrated that identifying as a minority racial or ethnic group is associated with poor telehealth access.^{28,30,37–39} Our present study, however, identified that subjects of a minority racial or ethnic group experienced greater benefit from virtual FCR, indicating that these subjects successfully used this telehealth tool to improve FCR access. These findings diverge from the large body of literature on telehealth inequities by race and ethnicity. This divergence is likely attributed to the unique use of telehealth for virtual FCR. Although most telehealth research has examined ambulatory provider-to-consumer telehealth visits, the use of telehealth for virtual FCR is an inpatient-based telehealth tool. The unique contexts, purposes, and workflows for virtual FCR likely enabled the relative benefit for racial and ethnic minorities. Importantly, race and ethnicity are proxy measures for constructs like experienced racism or discrimination. Findings based on these demographics do not provide conclusive understanding of the relevance and impact of the intersection of these social constructs on individual health care experiences.

This study’s findings that virtual FCR can mitigate race-, education-, and neighborhood health condition-based FCR attendance differences are relevant for improving more equitable healthcare delivery practices. Families with low income, public insurance, and from racial/ethnic minority backgrounds are less likely to receive parent-reported family-centered care.⁴⁰ Compared with their most affluent counterparts, parents living below 100% of the federal poverty line are twice as likely to report that their child’s providers were not sensitive to their values.⁴⁰ Parents of minority races are more likely to report that their child’s hospital doctors did not listen to them when compared with White parents. They are also more likely to report instances of experienced or observed discrimination.⁴¹ Black parents of hospitalized infants report more harmful provider interactions compared to White parents, including being disrespectful, impersonal, and dismissive.⁴² These interactions harm parents and impact the child by impeding parents’ ability to effectively engage in their child’s care. Virtual FCR cannot overcome all factors that result in harmful communication experiences; however, virtual FCR offers parents an additional resource to access FCR and thus increase their opportunities to strengthen communication and trust.

Although our study examined the heterogeneity of intervention effects of virtual FCR by subgroups, we lack data on the impact of this intervention by language preference. The original virtual FCR trial was limited to parents with English language preference. Expanding virtual FCR to include those with language preferences other than English (LOE) is imperative. Hospitalized children of parents with LOE are twice as likely to experience adverse events.^43,44 In addition, parents with LOE are more likely to report poor satisfaction, low comprehension, lack of knowledge of discharge plans, low parent activation, and inadequate family-centered care.^45–48 Although FCR addresses these very problems, parents with LOE attend FCR at rates less than half of parents with English proficiency: 26% versus 58%, respectively.¹⁷ Therefore, children and families at increased risk of the poor outcomes FCR is designed to mitigate are also the ones who struggle the most with accessing FCR. Increasing access to FCR is thus needed for parents with LOE, and virtual FCR is a potential strategy that can address this FCR access issue.

LIMITATIONS

Additional limitations of our study include having a local focus; the findings from this study may not be generalizable to other hospitals. It was also conducted at a level IV NICU within a teaching children’s hospital, which has unique contextual features. The telehealth program at our local institution provides equipment, software, and other supported services that enhanced the successful implementation of virtual FCR. We recognize that this unique telehealth infrastructure might not be widely available to other NICU settings. This study used post hoc subgroup analysis. Rigor is thus limited due to lack of specifying the subgroups a priori and designing the trial to analyze heterogeneity of intervention effects with sufficient power to detect modest effects for the outcomes across different subgroups. Despite these limitations, this study provided valuable information that supports the promise of virtual FCR as a health care delivery tool that can be used to improve FCR access inequities for socially disadvantaged families.

Conclusions

In summary, virtual FCR demonstrated improvements in parent FCR attendance overall, with greater benefits in increasing attendance for racial and ethnic minorities, those with no college education, and those from neighborhoods with worse health conditions. In light of the limitations of this post hoc subgroup study, future research is needed that specifies subgroups a priori and is adequately powered for hypothesis testing of heterogeneity of intervention effects. Nevertheless, virtual FCR is a telehealth tool with high potential to address some of the health care racial and socioeconomic inequities that result in FCR access inequities.

Disclosure Statement

The authors have no conflicts of interest to declare.

Funding Information

This work was supported by a pilot research grant from the University of California Davis Center for Healthcare Policy and Research and Center for Health Technology; the National Center for Advancing Translational Sciences, National Institutes of Health [grant number UL1 TR001860 and linked award KL2 TR001859]; and the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health [grant number K23HD101550 to Dr. Rosenthal]. The content is solely the authors’ responsibility and does not necessarily represent the official views of the National Institutes of Health.