Abstract
Objective:
COVID-19 continues to have a profound impact on infant healthcare and health outcomes. In this study, we aimed to characterize the social impact of the first COVID-19 lockdown on families in a neonatal follow-up program (NFP). Given the ongoing increased use of telehealth across the medicine, we also evaluated for patient-level differences in virtual visit rates to identify patients at risk of follow-up challenges.
Methods:
In order to assess the impact of virtual healthcare utilization, we conducted a retrospective cohort study to describe challenges associated with telemedicine use in this vulnerable patient population during our telemedicine epoch (March 13, 2020-July 31, 2020). We also looked for patient-level factors associated with attending NFP visits as scheduled. Finally, we summarized caregiver responses to a COVID-19 Obstacles Assessment Survey and assessed for racial disparities in these responses.
Results:
When comparing patients who completed their virtual visit to those who did not, we found no differences by infants’ sex, birthweight, gestational age at birth or caregiver self-reported race and ethnicity. However, infants whose visits did not occur were more often discharged with equipment or covered by public insurance. Nine percent of families endorsed food insecurity.
Conclusion:
During the initial COVID-19 lockdown, families with infants discharged from a NICU faced significant obstacles caring for their infants and attending scheduled follow-up visits. Infants in families with lower socioeconomic status or with increased medical complexity faced increased challenges in attending virtual follow-up visits during this epoch. Given the ongoing reliance on telemedicine in healthcare and the need to better prepare for future epidemics/pandemics, this study offers critical information that can assist neonatal teams in bolstering transitions to home and creating stronger safety nets for their patients after discharge.
Keywords: SARS-CoV-2, neonatal follow-up, telehealth, food insecurity, racial disparities
INTRODUCTION
Infants discharged from neonatal intensive care units (NICUs) are vulnerable to a variety of adverse health outcomes.1–3 They are at increased risk for poor growth and development and are more likely to rely upon medical technology.3–8 Many infants discharged from the NICU have adjunct services involved in their care, further adding to their medical complexity.4,9–12 In addition to increased medical needs, infants discharged from the NICU may be coming home to a household experiencing increased stress.4,5 Not only do parents of children with disabilities or medical complexity report increased stress in general, but there is also evidence that parents of NICU graduates may experience difficulties trying to access comprehensive and appropriate services for their infants.4
Amidst these stressors and challenges, the global COVID-19 pandemic hit in March 2020, bringing not only unprecedented health and health care challenges, but also highlighting and exacerbating known health disparities.13–16 The COVID-19 pandemic quickly and radically changed the delivery of outpatient healthcare. During the height of social lockdowns aimed at slowing the spread of the virus, many medical practices transitioned services to a telemedicine model to ensure continuity of care while maintaining social distancing measures.17–19 While this led to the development of new technology and optimization of existing workflows, the switch to a predominantly virtual platform also quickly created deep digital disparities for patients made vulnerable by structural disenfranchisement.14 As such, despite its many potential benefits,20 telemedicine has the potential to compromise high-quality care, especially among communities with specific vulnerabilities21 to poor internet access or digital literacy.14,16,22,23
Given promising early data for the use of telemedicine among NICU graduate families even before the pandemic,24 in this study, we sought to explore the challenges experienced by NICU graduate families during the initial COVID-19 lockdown in 2020 when all follow-up visits were transitioned to a virtual platform. Evaluation of such data is paramount to the improvement of care provision during the ongoing pandemic, especially given the ongoing reliance on telemedicine in healthcare. However, learning from the challenges experienced by this population during that historical time can allow for the development of better infrastructure to support families during the transition from NICU to home, clarify the strengths and barriers to utilizing telemedicine for post-NICU medical and developmental care, and elucidate the challenges families of NICU graduates face in accessing necessary resources. Our specific aims were to: 1) describe telehealth challenges experienced by families during their virtual follow-up visits; 2) describe differences between families who completed a virtual NICU follow-up visit and families whose visits did not occur as scheduled with the aim of identifying risk factors for poor virtual follow-up; and 3) document pandemic-related obstacles self-reported by families of NICU graduates in order to help inform the creation of future safety nets that can create structural resilience25 for families during future public health crises.
Methods
Study Design
This was a retrospective cohort study of families with infants discharged from neonatal intensive care units. Historical control data were used for comparisons in some analyses. Data was mainly survey-based, but all data was acquired retrospectively from electronic medical records (EMR) or existing research databases.
Study Population and Time Frame
Our Neonatal Follow-Up Program (NFP) provides follow-up outpatient services to infants discharged from a quaternary level NICU and a tertiary level NICU in urban Philadelphia. Our program cares for children born less than 32 weeks’ gestation and children with other perinatal morbidities associated with neurodevelopmental deficits (e.g., hypoxic-ischemic encephalopathy, history of treatment with ECMO, presence of chronic lung disease, in-utero drug exposure, etc.) longitudinally through age 5. Prior to the pandemic, our clinic averaged approximately 300 visits per month.
In response to rising COVID-19 case numbers and public health guidance, our healthcare system closed all outpatient clinics on March 13, 2020. In line with institutional policies, in-person visits fully resumed by September 1, 2020. Given our focus on describing the social and health care experiences of our NFP families during the initial lockdown, we defined the study period as March 13, 2020-July 31, 2020 and included all children scheduled for NFP telemedicine visits during this epoch.
COVID-Related Obstacles Assessment Survey
Given the immediately apparent widespread and profound impact the pandemic was having on crucial social determinants of health such as housing, employment and access to food and other essential goods,14,16,26 the multidisciplinary NFP team rapidly designed and implemented an Obstacles Assessment Survey into all telemedicine visits. The survey, which was administered by either physicians or psychologists, aimed to capture both qualitative and quantitative data regarding new or worsened obstacles families might be facing as a direct result of the COVID-19 pandemic (Online Supplementary Table). Questions focused on difficulties accessing health-related services, medical supplies, food or other essential items for infants, as well as new familial stressors related to housing and employment. It included a validated 2-question food insecurity screening tool known as “The Hunger Vital Sign.”27 The goal of our survey was to identify needs that our team might be able to address directly or through resource provision.
Data Collection
All data from the virtual visits related to the Obstacles Assessment Survey and telemedicine visit characteristics were abstracted from the EMR by two team members. Data collectors each abstracted 10 randomly selected charts in order to measure interrater agreement in EMR data abstraction. This allowed queries between the two data collectors to be resolved, in consultation with a third research team member. As interrater agreement was noted to be high on this random sample (Cohen’s kappa=0.8), data collectors then each subsequently completed half of the remaining sample of charts. All data were abstracted into a HIPAA-compliant REDCap database. Demographic information not readily available in the EMR, including maternal education, marital status, insurance status and preferred language, were pulled from existing research databases of ongoing research studies,28 which use NFP-based data and are merged with our study-specific REDCap database.
Statistical Analyses
For the first aim of describing challenges related to the virtual visit experience, data collectors manually extracted data to address three questions: 1) whether a visit was noted to be complicated by technology (i.e. internet connectivity issues), 2) whether a visit was transitioned from the virtual platform to a telephone call, and 3) whether a visit was terminated early due to technological challenges. We could not ascertain from the EMR whether a visit did not occur at all due to technological challenges if there was no documentation to that effect (i.e. no provider note about a telephone call from family stating such challenges).
For the second aim of understanding characteristics of children and families for whom virtual NFP visits did not occur as scheduled, we created two groups of predictor variables: sociodemographic variables and infants’ medical characteristics. Potential sociodemographic predictors of a virtual NFP visit not occurring were maternal education, marital status, insurance type, combined race and ethnicity, and preferred language. Medical characteristics included gestational age at birth, the need for equipment at discharge, and corrected age at the time of the scheduled visit. Corrected age was included as a proxy for variation that might exist between patients who were new to the follow-up clinic and those who were more likely to be established patients. At our institution, patients are scheduled for their first NFP visit when they are 3-4 months corrected age. As a result, we queried the impact of an infant being > 4.5 months corrected age at the time of the scheduled visit vs. < 4.5 months corrected age, under the hypothesis that how established a family was with the practice might be associated with the odds of an NFP visit occurring during this telehealth period.
We conducted univariate analyses to describe differences in familial sociodemographic and infant medical characteristics between those whose virtual visits were completed and those whose visit did not occur. Multivariable logistic regression models were then sequentially constructed, via the enter method, to determine the associations between virtual visit non-occurrence and sociodemographic and medical characteristic variables. Variables were assessed separately and then as groups. Final models included any variable which remained significantly associated with non-occurrence of a virtual visit after adjusting for all other sociodemographic and medical risk variables.
For the final aim of this study, the primary outcomes of interest were caregivers’ responses to the COVID-related Obstacles Assessment Survey. Families were offered the survey at each telemedicine visit. Survey responses were first summarized for the overall cohort. Given well-established racial disparities in food, housing and economic security29,30 and the inequitable over-representation of Black infants within NICUs in the United States,31 we were also interested in documenting how the pandemic might be influencing NICU families of different races and ethnicities. Race and ethnicity data were self-reported and used to construct a variable that captured both dimensions in one variable (non-Hispanic Black, non-Hispanic White, Hispanic, non-Hispanic Asian and Other). These five categories were used in the univariate analyses in order to provide descriptive information about the entire cohort with as much granularity as possible. However, given the demographic makeup of our local community, the sample sizes for the Asian, Hispanic and “Other” racial/ethnic subgroups were too small to allow for multivariable comparisons and to compare participants who endorsed resource needs on the COVID-related Obstacles Assessment Survey. Thus, these three groups were consolidated into one larger “Other” group for the multivariable analyses to create cell sizes that allowed for comparisons.
This study was approved as exempt by our local institutional review board. STATA/IC Version 16.1 and R Studio were used for analyses.
Results
Virtual visit occurrence and telemedicine complications
During our study period, our NFP clinic scheduled 780 visits for 716 individual patients. After excluding repeat visits and visits canceled by clinic staff (for reasons such as the infant still being inpatient and rescheduling requested by family or provider), we were left with 678 unique family visits in our final dataset. Within this cohort, 490 scheduled visits occurred (72%) and 188 did not occur as scheduled (28%). Reasons for non-occurrence were mainly related to a family canceling a visit before the visit date without rescheduling it or a family not connecting to the virtual visit (i.e., “no show”).
Clinic notes for visits which did occur mentioned technology or connection complications in 2.2% of the time (n=11 visits), and 1.8% of telemedicine visits (n= 9) were unable to be completed in their entirety due to technological issues. Two visits began on the virtual platform and were converted to a phone call, and only one visit occurred entirely via telephone from the start.
Predictors of Non-Occurrence
Infants scheduled for telemedicine visits were, on average, 31 weeks’ gestation at birth with a mean birthweight of 1690 grams (Table 1). Eighteen percent of this study cohort required equipment such as feeding tubes, apnea monitors and pulse oximeters at discharge. In univariate analyses, there were no differences in virtual visit occurrence by infant sex, birthweight, gestational age at birth, need for equipment at discharge or corrected age at the time of the visit (Table 1). Conversely, the caregivers of infants whose visits did not occur had lower levels of education or did not report their education (p <0.001), were less often married (p<0.001), and were more often covered by public rather than private insurance (<0.001; Table 2). Race/ethnicity and preferred language were similar between the groups of infants for whom visits were and were not completed.
Table 1.
Infant Characteristics by Occurrence of Scheduled Neonatal Follow-Up Program Telemedicine Visit (March 13, 2021-July 31, 2021)
| Total (N=678) | Scheduled visit occurred (N=490) | Scheduled visit did not occur (N=188) | P-value | |
|---|---|---|---|---|
| n (%) or mean (standard deviation) | ||||
| Infant Sex (% female) | 306 (45.1%) | 215 (43.9%) | 91 (48.4%) | 0.330 |
| Birthweight, grams | 1690.4 (954.9) | 1696.9 (953.4) | 1672.0 (961.9) | 0.771 |
| Gestational age at birth, weeks | 31.2 (4.9) | 31.2 (4.8) | 31.0 (5.1) | 0.525 |
| Corrected Age at Visit | ||||
| Corrected age < 4.5 months | 141 (20.8%) | 108 (22.0%) | 33 (17.6%) | .2367 |
| Corrected age > 4.5 months | 537 (79.2%) | 382 (78.0%) | 155 (82.4%) | |
| Equipment required at discharge | 0.162 | |||
| None | 555 (81.9% | 408 (83.3%) | 147 (78.2%) | |
| Apnea Monitor | 16 (2.4%) | 11 (2.2%) | 5 (2.7%) | |
| Pulse Oximetry | 43 (6.3%) | 32 (6.5%) | 11 (5.9%) | |
| Other* | 64 (9.4%) | 39 (8.0%) | 25 (13.3%) | |
Other includes feed-related equipment such as gastrostomy tubes and nasogastric tubes, as well as adaptive equipment such as splints.
Table 2.
Family demographics for infants scheduled for an NFP Telemedicine Visit
| Total (N=678) | Scheduled visit occurred (N=490) | Scheduled visit did not occur (N=188) | P-value* | |
|---|---|---|---|---|
| n (%) | ||||
| Caregiver educational level | <0.001 | |||
| Did not complete high school | 17 (2.5%) | 12 (2.5%) | 5 (2.7%) | |
| High school diploma/GED | 102 (15.0%) | 72 (14.7%) | 30 (16.0%) | |
| Some college/college degree | 257 (37.9%) | 201 (41.0%) | 56 (29.8%) | |
| More than college degree | 81 (12.0%) | 67 (13.7%) | 14 (7.5%) | |
| Did not report | 221 (32.6%) | 138 (28.2%) | 83 (44.2%) | |
| Caregiver marital status | <0.001 | |||
| Single, divorced or separated | 204 (30.1%) | 148 (30.2%) | 56 (29.8%) | |
| Married | 296 (43.7%) | 239 (48.8%) | 57 (30.3%) | |
| Did not report | 178 (26.3%) | 103 (21.0%) | 75 (39.9%) | |
| Caregiver insurance | <0.001 | |||
| Public | 252 (37.2%) | 163 (33.3%) | 89 (47.3%) | |
| Private | 353 (52.0%) | 295 (60.2%) | 58 (30.9%) | |
| Did not report | 73 (10.8%) | 32 (6.5%) | 41 (21.8%) | |
| Caregiver race/ethnicity | 0.086 | |||
| Non-Hispanic White | 238 (35.1%) | 185 (37.8%) | 53 (28.2%) | |
| Non-Hispanic Black | 210 (31.0%) | 142 (29.0%) | 68 (36.2%) | |
| Hispanic | 72 (10.6%) | 55 (11.2%) | 17 (9.0%) | |
| Asian | 18 (2.7%) | 11 (2.2%) | 7 (3.7%) | |
| Other | 140 (20.7%) | 97 (19.8%) | 43 (22.9%) | |
| Preferred Language | 0.261 | |||
| English | 637 (94.0%) | 463 (94.5%) | 174 (92.6%) | |
| Spanish | 9 (1.3%) | 4 (0.8%) | 5 (2.7%) | |
| Other | 8 (1.2%) | 5 (1.0%) | 3 (1.6%) | |
| Unknown/Missing | 24 (3.5%) | 18 (3.7%) | 6 (3.2%) | |
Denotes the results of Chi square tests or Fisher’s exact tests for cells with less than 5 counts.
In unadjusted logistic regression models that examined each sociodemographic and medical risk variable independently (Model 1 in Table 3), we found that, compared to caregivers with more than a college degree, infants whose caregivers had not completed high school or had a high school diploma but no college education were more than twice as likely to have their virtual NFP visit not occur (Table 3). Other variables which significantly increased the odds of non-occurrence were an infant being discharged with equipment, caregivers reporting Medicaid insurance or no insurance, non-Hispanic Black or Other racial/ethnic identity, and unmarried primary caregiver (Table 3). When adjusting for all covariates, the odds of a virtual visit not occurring were nearly three times higher for infants whose caregivers reported no insurance or for whom this information was missing compared to infants whose caregivers reported private insurance (adjusted odds ratio or aOR 2.89, 95% CI 1.54-5.43; Table 3). The odds of a visit non-occurrence were also higher for infants whose caregivers reported Medicaid insurance compared to those with caregivers on private insurance (aOR 2.33, 95% CI 1.55-3.5). Finally, infants who were greater than 4.5 months corrected at the time of the visit as well as those who were discharged with equipment had higher odds of their virtual visit not occurring compared to infants who were younger than 4.5 months corrected or discharged without equipment, after adjusting for caregiver sociodemographic variables and infant gestational age at birth (Table 3).
Table 3.
Predictors of a Virtual NFP Visit Not Occurring as Scheduled
| Model 1: Unadjusted | Model 2: Adjusted for sociodemographic medical risk and visit variables | |
|---|---|---|
| Odds Ratios (95% Confidence Interval) | ||
| Education | ||
| More than college degree | Ref | - |
| Did not complete high school | 2.70 (1.01-7.17) | 1.15 (0.40-3.27) |
| High school diploma/GED | 2.29 (1.26-4.27) | 1.51 (0.77-3.01) |
| Some college/college degree | 1.30 (0.76-2.30) | 1.10 (0.62-2.01) |
| Marital Status | ||
| Married | Ref | - |
| Single/Divorced/Separated | 1.70 (1.19-2.44) | 0.90 (0.58-1.41) |
| Insurance | ||
| Private insurance | Ref | - |
| Medicaid insurance | 2.67 (1.93-3.72) | 2.33 (1.55-3.53) |
| None/Unknown | 5.94 (3.65-9.81) | 2.89 (1.54-5.43) |
| Race * | ||
| Non-Hispanic White | Ref | - |
| Non-Hispanic Black | 2.09 (1.45-3.02) | 1.37 (0.86-2.17) |
| Other | 1.50 (1.03-2.18) | 1.26 (0.83-1.90) |
| Medical Risk Variables | ||
| Discharged with equipment | 1.5 (1.02-2.21) | 1.65 (1.10-2.48) |
| Gestational age (continuous) | 0.99 (0.96-1.02) | 1.01 (0.98-1.04) |
| Corrected Age at Visit | ||
| Corrected age < 4.5 months | Ref | - |
| Corrected age > 4.5 months | 1.44 (0.99-2.13) | 1.80 (1.19-2.77) |
Due to small sample sizes, respondents self-identifying as Hispanic, Asian, or Other were consolidated into one group to allow for statistical modeling.
Prevalence of pandemic-related obstacles
Response rates to our Obstacles Assessment Survey varied from 53% to 61% for each of the obstacles assessed (Table 4). Eight percent (n=24) of families who completed the survey endorsed having trouble accessing an indispensable infant resource such as formula, diapers, medicine or medical supplies due to cost; the item most often reported as difficult to access was formula. A slightly higher proportion of respondents reported that basic infant supplies were difficult to access due to availability (10% of respondents, n=30 families). A total of 33 families (%) endorsed food insecurity. In addition, 44% of respondents endorsed that at least one of the infant’s parents had their work hours cut or had become unemployed.
Table 4.
Prevalence of pandemic-related challenges overall and by self-reported race1
| Survey Item | Overall Cohort | Non-Hispanic White | Non-Hispanic Black | Other1 | P-value |
|---|---|---|---|---|---|
| n(%) for positive responses | |||||
| Have you had any trouble accessing any of the following for your child due to cost? (n=301)2 | 24 (8.0%) | 7 (29%) | 10 (42%) | 7 (29%) | 0.458 |
| Formula | 10 (3.3%) | 4 (40%) | 3 (30%) | 3 (30%) | |
| Diapers | 5 (1.7%) | 0 | 2 (40%) | 3 (60%) | |
| Medicine | 3 (1.0%) | 1 (33.3%) | 2 (67%) | 0 | |
| Medical Supplies | 2 (0.7%) | 1 (50%) | 1 (50%) | 0 | |
| More than one of these things. | 4 (1.3%) | 1 (25%) | 2 (50%) | 1 (25%) | |
| Have you had any trouble accessing any of the following for your child due to availability? (n=298) | 30 (10.1%) | 6 (20%) | 17 (57%) | 7 (23%) | 0.004 |
| Formula | 14 (4.7%) | 2 (14%) | 9 (64%) | 3 (21%) | |
| Diapers | 7 (2.3%) | 2 (29%) | 3 (43%) | 2 (29%) | |
| Medicine | 1 (0.7%) | 0 | 1 (100%) | 0 | |
| Medical Supplies | 3 (1.0%) | 0 | 1 (33%) | 2 (67%) | |
| More than one of these things. | 5 (1.7%) | 2 (40%) | 3 (60%) | 0 | |
| “We worried whether our food would run out before we got money to buy more.”3 (n=267) | 24 (9.0%) | 6 (24%) | 13 (54%) | 5 (21%) | 0.032 |
| “The food we bought just didn’t last and we didn’t have money to get more.” 3 (n=262) | 9 (3.4%) | 3 (33%) | 4 (44%) | 2 (22%) | 0.646 |
| “We were unable to get food for another reason.”3(n=248) | 12 (4.8%) | 2 (17%) | 6 (50%) | 4 (33%) | 0.207 |
| Either parent’s employment changed (i.e. hours were reduced or become unemployed) (n=294) | 129 (44%) | 57 (44%) | 36 (28%) | 36 (28%) | 0.150 |
| Housing status changed (n=294) | 18 (6.1%) | 8 (44%) | 3 (17%) | 7 (39%) | 0.952 |
Due to small sample sizes, respondents self-identifying as Hispanic, Asian, or Other were consolidated into one group to allow for cross-group comparisons.
These sample sizes indicate the total responses received for this question.
A positive response meant the caregiver endorsed feeling this way “often” or “sometimes.
When exploring racial differences in who endorsed each of these needs or challenges, we found that non-Hispanic Black families were more likely to report having trouble accessing a needed infant supply due to availability, and more likely to be concerned that their family’s food would run out before being able to buy more than non-Hispanic White families (p=0.004 and 0.032, respectively) (Table 4).
Discussion
In this study of neonatal follow-up patients followed virtually during the first wave of the COVID-19 pandemic in 2020, we documented several predictors of a virtual NFP visit not occurring, including coverage through public insurance, older corrected age at the time of a visit and an infant being discharged on equipment. In addition, we found that roughly 90% of visits that began virtually were successfully conducted without technologic issues.
When the pandemic began, there were significant concerns about the potential for telemedicine to exacerbate health disparities given known challenges to virtual health visits.22 Despite the validity of these concerns, the evidence for telemedicine widening disparities has been mixed, with some studies reporting racial, ethnic, age and income disparities in telemedicine use32 and others finding increased access to healthcare for certain marginalized groups as a result of the telehealth expansion prompted by COVID-19.33 Telemedicine has been demonstrated to be a feasible tool for follow-up of NICU discharges.24 Furthermore, compared to in-person nursing visits, telemedicine visits for newborns have been associated with fewer ER visits after discharge from the hospital.34 Our team previously described our NFP’s experiences with transitioning to telemedicine visits, including the steps taken to conduct neurodevelopmental assessments virtually.35 This study provides evidence that, with the right technology, NICU families who can connect to a virtual visit can subsequently complete these visits without significant technological challenges, even when faced with significant new challenges related to a worldwide pandemic.
Neonatal follow-up programs (NFPs) are specialized models of care that can provide outpatient access to neonatologists and other specialists who focus on identifying and addressing neurodevelopmental and medical morbidities of prematurity after discharge from a NICU. Such programs typically focus on close monitoring of growth and development, early referrals, care coordination, and overall transition to primary care-based medical homes.7 Neonatal follow-up programs are thus uniquely situated to provide comprehensive multidisciplinary care as well as psychosocial support to families of these high-risk children. Despite this, some studies have documented attendance rates as low as 50% in some areas.36
We found that nearly a quarter of telemedicine visits did not occur as scheduled, and that there were socioeconomic differences between the cohort of families whose visits did occur and those for whom they did not occur as scheduled. Notably, families who did not connect for their visit or who canceled their visit without rescheduling it were more likely to have a caregiver who had previously self-identified as being unmarried, possessed lower education levels or were covered by public insurance. A recent study examining a clinical cohort in Boston, Massachusetts reported racial and preferred language disparities in neonatal follow-up visit attendance rates.37 The different findings we have documented may be related to demographic differences between study cohorts and the small number of families who preferred a language other than English in our cohort. Notably, we also found that infants who were discharged with medical equipment (such as oxygen or feeding tubes) and those who had a higher corrected age at the time of the scheduled visit were also more likely to have their visits not occur as scheduled. Although our retrospective EMR dataset could not explore the reasons for these findings, it is possible that families with more medically complex children or those for whom more time had passed since discharge from the NICU were prioritizing other types of visits or perhaps facing more difficulties with their time. These findings represent areas critical for future study given the increased risk for poor health outcomes among medically complex infants and their need for continued long-term follow-up.
In addition, our Obstacles Assessment Survey revealed that 9% of our neonatal follow-up cohort experienced food insecurity during this difficult time. Food insecurity affected about 11% of American households overall before COVID-19 emerged.38 Since the onset of the pandemic, studies have documented the prevalence of moderate to severe food insecurity in the US rising to close to 40%.39 Concerningly, food insecurity during the pandemic appeared to rise earlier among households containing children.40 Food insecurity has been associated with infant mortality,41 and as such represents a critical social determinant of health to address among families with infants. However, there remains little published literature estimating the prevalence of food insecurity among families whose infants had been admitted to the NICU.
Although we screened families during a time of heightened national economic insecurity and supply chain challenges, we were still concerned to find that a proportion of the families who had recently been discharged from our highly-resourced clinical environment would have difficulty finding formula for their medically vulnerable infants. We believe NICU families not being able to access formula for their NICU graduate represents a “never” event, to use patient quality and safety terminology for an event that is identifiable, has the potential for serious consequences, and importantly, preventable.42 Our findings indicate the need to better understand how to support NICU families who might be experiencing (or are at risk for) food insecurity before discharge. Locally, our findings have informed ongoing quality improvement projects to address this issue at our own institution.
Limitations to this study include the low number of participants who did not identify as non-Hispanic Black or non-Hispanic White in our cohort, which limited our ability to generate meaningful conclusions about such families. In addition, our EMR-based dataset could not fully ascertain why certain visits did not occur as scheduled, and in particular, why families did not show up for their scheduled virtual visits. It is possible that technologic barriers proved insurmountable for some families, preventing them from initiating a visit at all and causing them to fall into the group of families for whom visits did not occur as scheduled. Thus, the rate of technological barriers to neonatal follow-up visits may be higher than we calculated in this analysis. Mixed-methods approaches which combine administrative or EMR datasets with qualitative work interviewing families of NICU graduates represent one way to further explore these issues.
In conclusion, in this study of the impact of COVID-19 on families followed in a large urban neonatal follow-up program during the early stage of the pandemic when follow-up visits were entirely virtual, we found socioeconomic disparities in which families ultimately connected to and completed their virtual visit. In addition, we found a concerning amount of food insecurity in our patient population. Neonatal providers are optimally suited to address the social needs of NICU families as they transition to life at home with their infant.43 These findings can be used by NICUs locally, but also across the country to explore regional needs and develop programs and policies to better prepare NICU families for the challenges they may face after discharge from the NICU.
Supplementary Material
KEY POINTS.
Telemedicine works well for high-risk neonatal populations.
Infant medical complexity may be a risk factor for challenges attending neonatal follow-up visits.
NICUs should work to prevent food insecurity post-discharge.
Acknowledgements:
The authors would like to acknowledge Mr. Matthew Devine & Ms. Liseth Cabrera-Hansali for her assistance with this project.
Funding Sources:
Dr. Peña is supported by the National Institutes of Health (grant T32HL098054-11). Dr. Montoya-Williams is supported by the National Institutes of Health (grant K23 HD102526).
Abbreviations
- NFP
Neonatal Follow-Up Program
- NICU
Neonatal intensive care unit
- EMR
Electronic medical record
Footnotes
Conflicts of Interest: The authors disclose no conflicts of interest.
Data Sharing: De-identified data available upon request.
References
- 1.Hintz SR, Gould JB, Bennett MV, et al. Referral of very low birth weight infants to high-risk follow-up at neonatal intensive care unit discharge varies widely across California. J Pediatr 2015;166(2):289–295 [DOI] [PubMed] [Google Scholar]
- 2.Wolke D, Söhne B, Ohrt B, Riegel K. Follow-up of preterm children: important to document dropouts. Lancet 1995;345(8947):447. [DOI] [PubMed] [Google Scholar]
- 3.Tin W, Fritz S, Wariyar U, Hey E. Outcome of very preterm birth: children reviewed with ease at 2 years differ from those followed up with difficulty. Arch Dis Child Fetal Neonatal Ed 1998;79(2):F83–F87 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Warfield ME, Krauss MW, Hauser-Cram P, Upshur CC, Shonkoff JP. Adaptation during early childhood among mothers of children with disabilities. J Dev Behav Pediatr 1999;20(1):9–16 [DOI] [PubMed] [Google Scholar]
- 5.Bailey DB Jr, Hebbeler K, Spiker D, Scarborough A, Mallik S, Nelson L. Thirty-six-month outcomes for families of children who have disabilities and participated in early intervention. Pediatrics 2005;116(6):1346–1352 [DOI] [PubMed] [Google Scholar]
- 6.Fazzi E, Galli J. New clinical needs and strategies for care in children with neurodisability during COVID-19. Dev Med Child Neurol 2020;62(7):879–880 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Harmon SL, Conaway M, Sinkin RA, Blackman JA. Factors associated with neonatal intensive care follow-up appointment compliance. Clin Pediatr 2013;52(5):389–396 [DOI] [PubMed] [Google Scholar]
- 8.Campbell MK, Halinda E, Carlyle MJ, Fox AM, Turner LA, Chance GW. Factors predictive of follow-up clinic attendance and developmental outcome in a regional cohort of very low birth weight infants. Am J Epidemiol 1993;138(9):704–713 [DOI] [PubMed] [Google Scholar]
- 9.Barfield WD, Clements KM, Lee KG, Kotelchuck M, Wilber N, Wise PH. Using linked data to assess patterns of early intervention (EI) referral among very low birth weight infants. Matern Child Health J 2008;12(1):24–33 [DOI] [PubMed] [Google Scholar]
- 10.Leventhal T, Brooks-Gunn J, McCormick MC, McCarton CM. Patterns of service use in preschool children: correlates, consequences, and the role of early intervention. Child Dev 2000;71(3):802–819 [DOI] [PubMed] [Google Scholar]
- 11.Hill JL, Brooks-Gunn J, Waldfogel J. Sustained effects of high participation in an early intervention for low-birth-weight premature infants. Dev Psychol 2003;39(4):730–744 [DOI] [PubMed] [Google Scholar]
- 12.Patel S, Corter C, Pelletier J, Bertrand J. “Dose-response” relations between participation in integrated early childhood services and children’s early development. Early Child Res Q 2016;35:49–62 [Google Scholar]
- 13.Ye S, Kronish I, Fleck E, et al. Telemedicine Expansion During the COVID-19 Pandemic and the Potential for Technology-Driven Disparities. J Gen Intern Med 2021;36(1):256–258 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lemmon ME, Chapman I, Malcolm W, et al. Beyond the First Wave: Consequences of COVID-19 on High-Risk Infants and Families. Am J Perinatol 2020;37(12):1283–1288 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Smith CB, Bhardwaj AS. Disparities in the use of telehealth during the COVID-19 pandemic. Journal of Clinical Oncology 2020;38(29_suppl):87–87. doi: 10.1200/jco.2020.38.29_suppl.87 [DOI] [Google Scholar]
- 16.Menon DU, Belcher HME. COVID-19 Pandemic Health Disparities and Pediatric Health Care—The Promise of Telehealth. JAMA Pediatrics Published online 2020. doi: 10.1001/jamapediatrics.2020.5097 [DOI] [PubMed] [Google Scholar]
- 17.Cutler DM, Nikpay S, Huckman RS. The business of medicine in the era of COVID-19. JAMA 2020;323(20):2003–2004. [DOI] [PubMed] [Google Scholar]
- 18.Mann DM, Chen J, Chunara R, Testa PA, Nov O. COVID-19 transforms health care through telemedicine: Evidence from the field. J Am Med Inform Assoc 2020;27(7):1132–1135 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Caporali C, Pisoni C, Naboni C, Provenzi L, Orcesi S. Challenges and opportunities for early intervention and neurodevelopmental follow-up in preterm infants during the COVID-19 pandemic. Child Care Health Dev 2021;47(1):140–141 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Marcin JP, Shaikh U, Steinhorn RH. Addressing health disparities in rural communities using telehealth. Pediatr Res 2016;79(1–2):169–176 [DOI] [PubMed] [Google Scholar]
- 21.Demographics of Mobile Device Ownership and Adoption in the United States. Accessed March 4, 2021. https://www.pewresearch.org/internet/fact-sheet/mobile/
- 22.Velasquez D, Mehrotra A. Ensuring the growth of telehealth during COVID-19 does not exacerbate disparities in care. Health Affairs Blog Published online 2020 [Google Scholar]
- 23.Walker DM, Hefner JL, Fareed N, Huerta TR, McAlearney AS. Exploring the Digital Divide: Age and Race Disparities in Use of an Inpatient Portal. Telemed J E Health 2020;26(5):603–613 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Das A, Cina L, Mathew A, Aziz H, Aly H. Telemedicine, a tool for follow-up of infants discharged from the NICU? Experience from a pilot project. J Perinatol 2020;40(6):875–880 [DOI] [PubMed] [Google Scholar]
- 25.Szanton SL, Gill JM. Facilitating resilience using a society-to-cells framework: a theory of nursing essentials applied to research and practice. ANS Adv Nurs Sci 2010;33(4):329–343 [DOI] [PubMed] [Google Scholar]
- 26.Schiariti V The human rights of children with disabilities during health emergencies: the challenge of COVID-19. Dev Med Child Neurol 2020;62(6):661. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.The Hunger Vital SignTM. Children’s HealthWatch. Published November 8, 2018. Accessed February 10, 2022. https://childrenshealthwatch.org/public-policy/hunger-vital-sign/
- 28.Bell EF, Hintz SR, Hansen NI, et al. Mortality, In-Hospital Morbidity, Care Practices, and 2-Year Outcomes for Extremely Preterm Infants in the US, 2013–2018. JAMA 2022;327(3):248–263 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Odoms-Young A Examining the Impact of Structural Racism on Food Insecurity: Implications for Addressing Racial/Ethnic Disparities. Fam Community Health 2018;41 Suppl 2 Suppl, Food Insecurity and Obesity:S3–S6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Racism and Discrimination Contribute to Housing Instability for Black Families During the Pandemic. Child Trends. Published March 17, 2021. Accessed February 3, 2022. https://www.childtrends.org/publications/racism-and-discrimination-contribute-to-housing-instability-for-black-families-during-the-pandemic
- 31.Kim Y, Ganduglia-Cazaban C, Chan W, Lee M, Goodman DC. Trends in neonatal intensive care unit admissions by race/ethnicity in the United States, 2008–2018. Sci Rep 2021;11(1):23795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Haynes N, Ezekwesili A, Nunes K, Gumbs E, Haynes M, Swain J. “Can you see my screen?” Addressing Racial and Ethnic Disparities in Telehealth. Curr Cardiovasc Risk Rep 2021;15(12):23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Pagán VM, McClung KS, Peden CJ. An Observational Study of Disparities in Telemedicine Utilization in Primary Care Patients Before and During the COVID-19 Pandemic. Telemed J E Health Published online December 20, 2021. doi: 10.1089/tmj.2021.0412 [DOI] [PubMed] [Google Scholar]
- 34.Robinson C, Gund A, Sjöqvist BA, Bry K. Using telemedicine in the care of newborn infants after discharge from a neonatal intensive care unit reduced the need of hospital visits. Acta Paediatr 2016;105(8):902–909 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.DeMauro SB, Duncan AF, Hurt H. Telemedicine use in neonatal follow-up programs - What can we do and what we can’t - Lessons learned from COVID-19. Semin Perinatol Published online April 6, 2021:151430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Litt JS, Edwards EM, Lainwala S, et al. Optimizing High-risk Infant Follow-up in Nonresearch-based Paradigms: The New England Follow-up Network. Pediatr Qual Saf 2020;5(3):e287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Fraiman YS, Stewart JE, Litt JS. Race, language, and neighborhood predict high-risk preterm Infant Follow Up Program participation. J Perinatol Published online August 17, 2021. doi: 10.1038/s41372-021-01188-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Coleman-Jensen A, Gregory C, Singh A. Household Food Security in the United States in 2013. Published online September 1, 2014. doi: 10.2139/ssrn.2504067 [DOI] [Google Scholar]
- 39.Fitzpatrick KM, Harris C, Drawve G, Willis DE. Assessing Food Insecurity among US Adults during the COVID-19 Pandemic. J Hunger Environ Nutr 2021;16(1):1–18 [Google Scholar]
- 40.Ahn S, Norwood FB. Measuring food insecurity during the COVID‐19 pandemic of spring 2020. Appl Econ Perspect Policy 2021;43(1):162–168 [Google Scholar]
- 41.Cassidy-Vu L, Way V, Spangler J. The correlation between food insecurity and infant mortality in North Carolina. Public Health Nutr Published online January 31, 2022:1–7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Patient Safety Network. Never Events. Agency for Healthcare Research and Quality. Published September 7, 2019. Accessed July 14, 2022. https://psnet.ahrq.gov/primer/never-events [Google Scholar]
- 43.Montoya-Williams D, Fraiman YS, Peña MM, Burris HH, Pursley DM. Antiracism in the Field of Neonatology: A Foundation and Concrete Approaches. Neoreviews 2022;23(1):e1–e12. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.