Referral order placement decreases time to transfer to adult congenital heart disease care.

Abstract


Introduction
Current guidelines state that patients with moderate or great complexity pediatric congenital heart disease (CHD) should transfer to an accredited adult congenital heart disease (ACHD) center [1].In the United States (US), the recommended age for transfer to an accredited ACHD center is 18-21 years of age [2].These years, however, are a vulnerable time for patients with CHD who may be experiencing parallel physiological, psychological, and social changes that can potentially impact a successful transfer to accredited ACHD centers [3][4][5][6].Patients with CHD who experience fragmentation of care during the transition years have an increased risk for developing complications leading to hospitalization and procedural intervention [7][8][9][10][11].In fact, patients who experience a lapse in care during the transition years have about 3 times greater odds of requiring urgent cardiac intervention compared to those with no lapse [11].
Transfer rates of pediatric CHD patients to accredited ACHD centers are reported to be higher in countries with larger specialized CHD programs, closer location and a liation between pediatric CHD and ACHD services, and programs with structured transfer processes [12].In the US, between 11-39% of pediatric CHD patients transfer to ACHD centers [13][14][15][16][17][18][19][20], which is considerably lower than transfer rates in Europe [21,22] and Canada [23].In the US, structured transfer processes to accredited ACHD centers are less commonly used than in Europe [24,25].Not surprisingly, gaps in care during the transition years are more common in the US than other countries [6].Additionally, almost half of the US population lives over an hour away from tertiary centers [26].Some tertiary care centers resolve this issue by providing care for patients at outreach sites, but patients cared for at pediatric CHD outreach sites that do not offer transfer support may be at increased risk of experiencing geographic disparity in healthcare access, resulting in sub-optimal transfer outcomes [26-28].There are a variety of referral practices across clinical settings (e.g., timing of order placement), however, which may explain the low transfer rates in the U.S. Placement of a referral order at the last pediatric cardiology visit, regardless of location, may help alleviate some barriers to successful transfer.The purpose of this study is to examine time to successful transfer to an accredited ACHD center between those with a referral order placement at the last pediatric cardiology visit versus those with no and delayed referral order placement among pediatric patients with moderate or great complexity CHD.We hypothesized that placement of a referral order at the last pediatric cardiology visit would lead to higher occurrence of transfer to accredited ACHD centers and sooner time to transfer to an accredited ACHD center.

Study Design
In June 2022, we performed a retrospective medical record review of patients with moderate and great complexity CHD who were eligible to transfer to our tertiary center's a liated accredited ACHD center between March 2020 and March 2022.Our tertiary health care center has both pediatric CHD and accredited ACHD facilities and is the only accredited ACHD center in the state.However, the tertiary health care center is located at the northwestern corner of the state, creating di cult access for those living in remote areas (Fig. 1).Thus, patients with pediatric CHD are cared for at either the tertiary care center or at one of 11 outreach sites located across Oregon, which are staffed by pediatric cardiologists a liated with the tertiary care center.Patients were included in the analysis if they were under the care of a pediatric cardiologist (either at the tertiary care center or an outreach center), had a primary diagnosis of moderate or great complexity CHD (Supplement 1), were 17 years or older as of March 1, 2020 and had seen a pediatric cardiologist between January 2016 and March 2022.Patients with a history of heart transplant were excluded from this analysis.The study participants were divided into two groups: 1) patients who received referral order placement to the accredited ACHD center at the last pediatric cardiology visit and 2) patients who received no referral orders at the last pediatric cardiology visit (such as those who received contact information of an ACHD center to schedule the transfer appointment themselves, or those who had no documented transfer plan), or received a delayed referral order more than thirty days after the last pediatric cardiology visit.The study was approved by the Institutional Review Board.

Data Collection
We collected baseline data from the last pediatric cardiology o ce visit, including sociodemographic variables (i.e.age, sex, language, insurance provider, last pediatric cardiology visit location and residence either within or outside the metropolitan area) and clinical variables (i.e.primary cardiac diagnosis, date of the last pediatric cardiology visit, date of the rst ACHD center visit).We then classi ed patients' cardiac anatomy by the 2018 American Heart Association (AHA) / American College of Cardiology (ACC) guideline for the management of adults with CHD ACHD Anatomic and Physiological (AP) classi cation system [1].The AHA/ACC 2018 guidelines de ne moderate complexity CHD as non-cyanotic defects that are either repaired or unrepaired and moderate in nature such as atrio-ventricular septal defects, repaired tetralogy of Fallot.Great complexity CHD anatomy is de ned as cyanotic defects which are unrepaired or palliated and single ventricle, such a hypoplastic left heart, and transposition of the great arteries [1].(Supplemental table 1).
Then we followed up with patients to collect transfer outcomes.Patients were categorized as either 1) successful transfer (i.e., completed an initial visit with an ACHD provider), or 2) unsuccessful transfer (i.e., did not complete an initial visit with an ACHD provider).We further categorized unsuccessful transfer as either 1) Retained in pediatric care (i.e.planned continuation of pediatric cardiology care during the study period), 2) Overdue for transfer (i.e.past the due date to see an ACHD provider and not meeting criteria for being lost to follow-up), or 3) Lost to follow-up (i.e.absent from care beyond six-months of the planned transfer date and were issued two phone calls and a letter from the ACHD team, or those who had a threeyear gap in pediatric CHD care [6, 14,18].

Data analysis
Standard descriptive statistics of frequency, central tendency, and dispersion were used to describe the sample.Comparative statistics (chi-square test, Fisher's exact test and Student's T test) were used to compare differences in sociodemographic and clinical variables between those who received referral order placement at the last pediatric cardiology visit versus those who did not.We generated time-totransfer curves using Kaplan-Meier graphs and used the log-rank Mantel-cox test to compare the probability of successful transfer between those who received referral order placement at the last pediatric cardiology visits versus those who did not.We also explored whether CHD complexity, location of last pediatric cardiology visit, living location within or outside of the metro area, age and sex impacted time to transfer.Cox proportional hazard regression modeling was used to analyze time to transfer, adjusting for factors signi cant in bivariate testing as well as age and sex.Predictors are reported as hazard ratios (HR) with corresponding 95% con dence intervals (CI).Stata/MP version 17 (StataCorp, College Station, TX) and SPSS version 28 (IBM, Armonk, NY) were used for data analysis.

Results
A total of 65 patients met inclusion and exclusion criteria.About half were female, and the majority were Non-Hispanic White (Table 1).In this sample, 21 (32.3%)patients received a referral order at the last pediatric cardiology visit.Of those who did not receive an order at the last visit, 11 (16.9%) had an order placed more than 30 days later.Signi cantly more patients received a referral order placement at the last pediatric cardiology visit at the tertiary care center (30.8%) compared with outreach sites (1.5%), p > 0.001.In this sample, 47.7% successfully transferred to the accredited ACHD center, and those with a referral order at the last visit had more successful transfers than those who did not (p < 0.001).There was a signi cant difference in the time to transfer to the accredited ACHD center between those who received a referral order at the last pediatric cardiology visit compared with those who did not (logrank 38.84, p > 0.001; Fig. 2).The median time to transfer was 12.8 months (Interquartile range: 9-18.9) for those with a referral order placement at the last pediatric visit and 28.4 months (Interquartile range:18-41.73)for those without.We also observed signi cant differences in time to transfer between those who were seen at the tertiary care center versus those seen at outreach sites (log-rank 8.52, p = 0.004), those living inside the metro area compared with those lived outside of the metro area (log-rank 4.54, p = 0.033), and those with greater complexity CHD versus those with moderate complexity CHD (logrank 5.87, p = 0.015); these variables were then moved into the multivariate regression model also accounting for sex and age.In the multivariate Cox regression model (Table 2), those with a referral order placed at the last pediatric cardiology visit were signi cantly more like to transfer sooner than those without (HR 6.0; 95% CI 2.2-16.2,p > 0.001), adjusting for age, sex, CHD clinic site location, living location, and CHD complexity.Older age was also signi cantly associated with a sooner time to transfer (Table 2).

Discussion
In this study, we analyzed the impact of referral order placement at the last pediatric cardiology visit on transfer outcomes and time to transfer to an accredited ACHD center in a sample of 65 transition-aged patients with moderate and great complexity CHD.We found signi cant differences in time to transfer to the accredited ACHD center in those who received a referral order at the last pediatric cardiology visit, adjusting for other demographic and clinical variables.Patients who received a referral order at the last pediatric cardiology visit were almost six times as likely to transfer to the accredited ACHD center than those who received no or a delayed referral order.
The transfer occurrence to the accredited ACHD center in this study (44.7%) is higher than previously reported in the US [13][14][15][16][17][18][19][20].In our study 17% of participants were lost to follow-up and 5% were overdue for transfer to an ACHD center.Patients who had a referral order placed at the last pediatric cardiology visit had a lower occurrence of lost to follow-up (1.5%) compared to those with no referral order (15%).
But overall, our lost to follow-up rates were lower than previously reported in meta-analysis by Moons and colleagues who found that patients with CHD in the US experience higher rates of discontinuity in care during the transition years compared to those in Canada and Europe [6].Finally, retention in pediatric cardiology care was 31% in our study.This is comparable with US studies that show a range of 21-69% retention in pediatric cardiology care [13,[15][16][17][18][19].
Our ndings are in line with previous studies which have linked formal referral to ACHD centers with successful transfer outcomes [13,29], and extends these ndings by identifying which part of the referral process aids successful transfer to ACHD centers.At our ACHD center, referral order placement triggers a proactive outreach by a scheduling specialist, and it seems to be most effective when the referral order placement is done concurrently with the last pediatric cardiology visit.These contextual factors may explain the mechanism by which referral order placement facilitates sooner time to transfer.
Distance to the referral center has shown to either have no effect on transfer outcomes [13,16,17,21] or to negatively impact transfer outcomes [15,23,29].In this study, living inside the metro area seemed to shorten the time to transfer to the accredited ACHD center, but not when other variables were adjusted for in the model.Given that many patients with CHD live far away from ACHD centers, referral order placement may be one possible way to reduce geographic disparities.
In our study, the median time to transfer to the accredited ACHD center with referral order placement at the last visit was just under 13 months.This time frame is comparable to a study including a similar cohort of patients who underwent a robust resource-intensive transition program intervention in Texas (mean time to transfer was 13 months) [30].Comparing the two studies, our patients, who received a referral order alone, waited a similar amount of time to transfer to an ACHD center as the more intensive intervention.Thus, the simple placement of a referral order may demonstrate successful outcomes on par with more intensive interventions.
In clinical practice, placement of a referral order at the last pediatric cardiology visit may not require resource intensive processes.As placement of a referral order may lead to successful transfer outcomes and shorter time to transfer to accredited ACHD centers, it should be considered as a component of routine transfer of care practices.Providing clinicians with education about the downstream impact of the cascade of events that are triggered by a referral order placement as well as creating standard processes around referral order placement may help drive this practice.
Limitations of this study include the small sample size, a racially homogenous sample, a single site academic center and limitations based on retrospective record review.We did not collect data on patient and parent readiness for transfer, social, environmental, and structural factors related to race and racism, or other socioeconomic barriers impacting care (e.g., access to transportation), due to limitations of retrospective record review.This study requires validation from larger samples that are representative of the broader US population.Also, there is a need for a better understanding of the barriers of implementing structured transfer processes in the US, which are more common in other countries.Lastly there is a need to design interventions to improve the transfer process.

Conclusion
Our study highlights that referral order placement at the last pediatric cardiology visit improves transfer outcomes and time to transfer to accredited ACHD centers from pediatric CHD centers.These ndings are important as CHD patients who experience lapses in care during the transition years are at risk for developing complications.

Declarations
Compliance with Ethical Standards

Figures
Figures

Figure 1 Location
Figure 1

Table 1
Socio-demographics, Clinical Characteristics, and Transfer Outcomes Great complexity CHD included: congenital atresia of pulmonary valve, congenital tricuspid atresia, double inlet left ventricle, double outlet right ventricle, hypoplastic left heart syndrome, interruption of aortic arch, and transposition of great vessels a Moderate complexity CHD included: atrioventricular canal, ostium primum and tetralogy of Fallot.bGreat complexity CHD included: congenital atresia of pulmonary valve, congenital tricuspid atresia, double inlet left ventricle, double outlet right ventricle, hypoplastic left heart syndrome, interruption of aortic arch, and transposition of great vessels a Moderate complexity CHD included: atrioventricular canal, ostium primum and tetralogy of Fallot.b

Table 2
Multivariate Cox Proportional Hazard Regression for Occurrence of Transfer to the Accredited ACHD Center