Introduction
Despite advances in surgical interventions and diagnostic capabilities over the past several decades, CHD remains a leading cause of death in children less than 10 years old [
1,
2]. Even with surgery, many children with CHD may have residual disease and adverse sequela [
3]. Thus, lifelong care by a congenital cardiologist is often required for patients with CHD. For adults living with CHD, the American College of Cardiology (ACC) and the American Heart Association (AHA) recommend follow-up surveillance with an adult CHD specialist at least every 2 years for most congenital lesions [
4]. However, there are limited guidelines regarding follow-up in children. Some consensus groups have attempted to provide guidance on follow-up frequency in pediatric populations by specific cardiac lesion [
5]. Despite follow-up recommendations, many CHD patients are not seen by pediatric cardiology specialists following hospital discharge, which can lead to increased morbidity and mortality [
6‐
8].
CHD survival has been previously associated with several factors, including race, insurance status, and adherence to medical care [
8‐
11]. Furthermore, there is evidence that an individual’s surroundings can have an impact on their health outcomes and likelihood of follow-up, as has been shown in asthma and preterm births [
12‐
14]. Recently, several geospatially linked variables have been validated and have shown to strongly associate with healthcare access and outcomes [
15]. Examples of validated, geospatially linked indices include Neighborhood Deprivation Index (NDI), Racial Isolation (RI), Education Isolation (EI), and drive time to the nearest cardiology clinic.
Despite the importance of aftercare for complex CHD, there are few studies with large samples showing factors associated with gaps in cardiology follow-up in children with CHD with subsequent differences in health outcomes. Using a large retrospective database spanning multiple tertiary care centers across the state of North Carolina (NC) and multivariable analysis techniques, we aim to determine socioeconomic, demographic, and medical factors associated with gaps in follow-up and the potential consequences of a gap in care, including increased morbidity, mortality, and health care utilization.
Discussion
In this retrospective cohort study of children with CHD in NC, we examined risk factors for gaps in cardiology aftercare, and differences in healthcare utilization and outcomes when gap in follow-up occurs. Factors associated with reduced rates of follow-up in cardiology care included increased drive time to a cardiology clinic, worse EI, older age at initial cardiology encounter, and non-severe CHD lesion. Healthcare utilization differs in individuals who had gaps in cardiology follow-up, with more subsequent ED visits and fewer inpatient encounters and procedures. Finally, in this cohort, there was a low event rate for death, heart transplant, and LVAD placement.
Our findings for risk factors associated with a reduced rate of cardiology follow-up add to the scope of known risks in children with CHD. There is previous evidence that as children with complex congenital diseases, including CHD age, they are less likely to attend regularly scheduled follow-up encounters [
19‐
21]. Our results show that with increased age in infants, even on the scale of months, there is a substantial drop-off in likelihood of cardiology follow-up. The likelihood of follow-up continues to decline at a slower rate as children age from 1 to less than 10 years old. Explanations for why older children may be less likely to experience regular follow-up for congenital diseases include greater travel and scheduling constraints, including time in school, compared with younger children [
22]. An additional risk for decreased follow-up, especially when moving from infancy to childhood, is the potential loss of peripartum insurance benefits. For example, in NC, neonates with mothers on Medicaid are automatically eligible for Medicaid but may lose this coverage after their first birthday [
23]. Also, our finding that individuals with severe disease have a higher probability of follow-up with cardiology is concordant with previous findings that individuals whose daily lives are more impacted by their disease are more likely to follow-up [
7]. Additionally, as a child ages, longer time to follow-up with spaced out clinic visits may be expected based on the patient’s clinical course. We acknowledge that age may not necessarily be a risk factor for loss to follow-up. Because of this, analyses adjusted for age and CHD severity.
A novel finding from this study is the substantial impact of drive time to a cardiology clinic on likelihood of follow-up in children with CHD. It has been previously shown that outcomes for hypoplastic left heart syndrome (HLHS) are worse for individuals who live further from tertiary care centers [
24]. Our results show that drive time to a cardiology clinic has a major impact on cardiology follow-up, which may explain those previously described differences in outcomes. As about 40% of the NC population live in rural areas, there is an especially significant challenge to provide accessible cardiology care to all. Another novel finding of this study is the relationship between EI and likelihood of follow-up. Parental health literacy and education are prognostic factors of health outcomes in children with complex congenital disease [
25]. We provide evidence that one way in which EI may impact outcomes involves greater likelihood of reduced rates of cardiology follow-up. Surprisingly, we did not find a significant association between length of time to follow-up and either race or NDI.
We found differences in healthcare utilization among children who had gaps in cardiology follow-up. Specifically, we found that children with CHD who have a gap in cardiology follow-up are more likely to be seen in the ED and less likely to have inpatient encounters or receive interventions. There are several possible explanations for why ED utilization differs in patients who have gaps in care. The first is that children with CHD who do not follow-up with cardiology may receive less focused subspecialty treatment and health optimization. As a child with CHD ages, especially when their congenital lesion is severe, their cardiovascular physiology has been shown to change, requiring medication adjustments or further intervention even when prior medical and/or surgical interventions were initially effective [
26‐
29]. Patients who do not follow-up with cardiology within a 2-year period may miss out on these critical adjustments. Furthermore, signs of worsening clinical course, which may be amenable to intervention, may be missed. Any decrement in treatment efficacy, medical, or surgical could increase the likelihood of a CHD patient visiting the ED. A second potential explanation for the observed difference in ED utilization with reduced rates in cardiology follow-up is that subpopulations of individuals have been shown to regularly seek care in the ED and use it in place of an outpatient or primary care provider [
30,
31]. This utilization pattern is both more burdensome on the healthcare system and leads to worse health outcomes [
32]. Regardless of the relative contribution of these two hypothetical explanations, the increase in ED utilization in children with CHD who have a reduced rates of cardiology follow-up provides strong evidence in support of standardized of follow-up guidelines and outreach for vulnerable populations. It additionally provides an opportunity to reengage patients with pediatric cardiology care. The observed decrease in inpatient encounters and interventions may be representative of multiple phenomena, including decreased routine procedures, such as screening echocardiograms, in children who have a gap in care. Additionally, people with greater burden of disease and require more procedures may be more likely to follow-up outpatient. We found a low rate of death, cardiac transplantation, and LVAD placement in our cohort, with no recorded LVADs and < 1% of individuals experiencing death or transplant. This rate is considerably lower than in past studies of individuals with CHD, likely owing to improvements in surgical and medical management and a relatively young patient cohort [
9]. We likely did not have power to detect difference in death and transplantation due to the very low event rate.
Overall, we find that there are strong predictors of reduced rates of cardiology follow-up in children with CHD, including drive time to the nearest cardiology clinic and EI. Both factors can be further explored via qualitative follow-up studies to identify specific barriers to care and potential interventions, including broader distribution of transportation vouchers or telemedicine when appropriate. Furthermore, we provide evidence for the creation of standardized follow-up guidelines for children with CHD given the observed increase in ED utilization in children without follow-up within 2 years, compared to children with follow-up.
Limitations
In the absence of clear national guidelines, it is difficult to know when patients with CHD should follow up. Individual practitioners may recommend various intervals for their patients and the recommendations are likely based on many individual patient factors. To account for variations in recommendations, we chose a long follow-up period of 2 years with a 60-day grace period. Our analysis focused on identifying risk factors with a patient having their first follow-up visit in a reasonable time frame. We acknowledge that some patient might later be lost to follow-up and we are not capturing those patients. We also included lesion severity in our model as this is likely one of the most important factors clinicians use to determine follow-up guidelines. Regardless of when patients were recommended to follow-up our data showed differences in health care utilization in patients who had longer intervals to follow up.
There is an intrinsic potential bias in this study since people who do not follow-up with cardiology may have systematically lower data quality (more missingness) when compared with those who do follow-up. This limitation is mitigated by the fact that we only included individuals who have some form of recorded interaction with the healthcare system over a defined period. Another limitation is that we did not have information if a patient moved during the study and followed up with cardiology care outside of NC. Furthermore, CHD patients followed outside of the included major CHD centers would not be captured; some patients who were first seen at an included center and then followed up elsewhere (e.g., after an out-of-state move) would be erroneously labeled as lost to follow-up. However, based on 2013 US Census Bureau data showing the rate of total migration out of NC to be about 2.5% per year, it is unlikely that this significantly affected our results. Furthermore, 95% of pediatric cardiologists, all but one in NC, are captured in our database and thus it is unlikely that many in-state patients are getting appropriate CHD care elsewhere. If patients moved out of state and were considered to have a gap in follow-up, they likely also had lower or zero health care utilization. This limits our ability to detect greater utilization in those with gaps in care.
Our inability to include insurance status as part of our analysis limited our understanding of how this variable affects follow-up time and overall health care utilization in our specific population. As previously noted, there was a high percentage of missing insurance data in our population (40%). In addition to lack of insurance information, there may have been other important variables that were not included. Many care decisions both from the patient and provider perspective are complex and may not be entirely understood by the variables included in this study.
Additionally, while we chose a landmark study design to more effectively assess utilization and outcomes after gap in follow-up, we are still not able to assess causation (i.e., we cannot say that utilization changes occurred because of gap in follow-up). This study was performed in major, tertiary care hospital systems. The patient population seen in these systems may not be entirely representative of the total population of individuals with CHD.
Conclusion
In this study, we have identified risk factors for gaps in cardiology follow-up in children with CHD and consequences of having a gap in care. We find that there are substantial predictors of gaps in cardiology follow-up based on patient factors, socioeconomic indicators, and barriers to healthcare. Our findings also suggest that regular follow-up with an outpatient cardiologist is important for optimal healthcare utilization in children with CHD. The creation of clear, concordant follow-up guidelines based on CHD lesion for infants and children is therefore an essential step in CHD care in the pediatric population. Additionally, continued efforts to improve the accessibility of cardiology care, especially in rural populations are needed.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.