Impact of home monitoring program on interstage mortality after the Norwood procedure

Objective While early outcome after the Norwood operation for hypoplastic left heart syndrome has improved, interstage mortality until bidirectional cavopulmonary shunt (BCPS) remains a concern. Our aim was to institute a home monitoring program to (HMP) decrease interstage mortality. Methods Among 264 patients who survived Norwood procedure and were discharged before BCPS, 80 patients were included in the HMP and compared to the remaining 184 patients regarding interstage mortality. In patients with HMP, events during the interstage period were evaluated. Results Interstage mortality was 8% (n = 21), and was significantly lower in patients with HMP (2.5%, n = 2), compared to those without (10.3%, n = 19, p = 0.031). Patients with interstage mortality had significantly lower birth weight (p < 0.001) compared to those without. Lower birth weight (p < 0.001), extra corporeal membrane oxygenation support (p = 0.002), and lack of HMP (p = 0.048) were risk factors for interstage mortality. Most frequent event during home monitoring was low saturation (<70%) in 14 patients (18%), followed by infection in 6 (7.5%), stagnated weight gain in 5 (6.3%), hypoxic shock in 3 (3.8%) and arrhythmias in 2 (2.5%). An unexpected readmission was needed in 24 patients (30%). In those patients, age (p = 0.001) and weight at BCPS (p = 0.007) were significantly lower compared to those without readmission, but the survival after BCPS was comparable between the groups. Conclusions Interstage HMP permits timely intervention and led to an important decrease in interstage mortality. One-third of the patients with home monitoring program needed re-admission and demonstrated the need for earlier stage 2 palliation.


Introduction
The neonatal Norwood procedure for hypoplastic left heart syndrome (HLHS) and its variant has undergone constant improvement and modification to archive best possible results (1,2).Nevertheless, the crucial phase does not end when patients are discharged from hospital but spans the interstage period.Various research has indicated that interstage mortality may range from 2% to as much as 20% (3)(4)(5).Risk factors for adverse events during that period have been studied throughout the years.Several factors, such as low birth weight, anatomical subtypes, or extra-cardiac anomalies, are associated with higher interstage mortality rates.However, due to their inherent characteristics, these factors are mostly beyond the reach of operative interventions (4,6,7).Therefore, patients affected by these factors require close monitoring (6).
Frequent complications that may arise include excessive hypoxemia, hypovolemia, change in systemic and pulmonary vascular resistance or progressive myocardial dysfunction (5,8,9).Particularly, stenosis of modified Blalock-Taussig shunt (MBTS) or right ventricle to pulmonary artery conduit (RVPAC) may disturb the balance between systemic and pulmonary circulations (9).Respiratory illnesses might also be a risk for interstage complications.Hence, our aim was to identify potentially critical courses before patients would present with fulminant complications.To extend the monitoring of key parameters beyond hospital discharge, we launched a home monitoring program (HMP) according to the report of Ghanayem, et al. (10).HMP aims at detecting critical changes, such as drop of oxygen saturation and decrease in weight gain.In addition, HMP serves to support parents and custodians in the care of critically ill children.Many studies have presented positive results in reducing interstage mortality by implementation of similarly structured programmes (9,10).Rudd et al. have reported excellent interstage survival rates of 98% following the introduction of their HMP (11).
This study intends to evaluate the following two points: (1) whether introduction of HMP reduced the interstage mortality after the Norwood procedure.(2) How HMP detected the problems of patients, and how HMP influenced the outcomes after bidirectional cavopulmonary shunt (BCPS).

Ethical statement
This study was approved by the Institutional Review Board of the Technical University of Munich (approved number of 305/20 S-KH on 2 June, 2020).Because of its retrospective nature, the need for individual patient consent was waived.

Patients
Between January 2001 and December 2020, 335 consecutive patients with HLHS and its variants underwent the neonatal Norwood procedure at the German Heart Centre of Munich.Patients who were subsequently discharged from hospital with shunt-dependent physiology were included in this study (n = 264).Patients who died in hospital and those who underwent BCPS following S1P without being discharged in the interim were excluded from the study (Figure 1).Pre-and postoperative data was obtained by the review of the patients' medical records.Since 2013, patients with single ventricle physiology were included in a home monitoring program (n = 80) which allowed the assessment of adverse events prior to BCPS and if so, therefrom resulting interstage mortality.

Operative techniques
The Norwood procedure is performed under standard cardiopulmonary bypass (CPB) with hypothermic circulatory arrest.Selective cerebral perfusion has been performed since 2009.The details of the operative techniques are described in our previous reports (12,13).The choice of shunt was made at the discretion of patients' surgeon and cardiologist.In patients who received RVPAC, a 5.0 mm ringed Gore-Tex graft was used in most patients.In patients who underwent MBTS, most patients received a 3.5 mm non-ringed Gore-Tex graft.All patients were postoperatively admitted to the intensive care unit (ICU) with chest open.Delayed sternal closure was performed usually on the second postoperative day.

Home monitoring program
In 2013, an interstage HMP was established at our center, which allowed close monitoring and regular check-ups of the patients.The HMP team consists of a nurse case manager and a cardiologist (J.L.), who offers additional support.During the implementation phase, the personnel involved were exclusively sourced from the existing resources within the Division of Pediatric Cardiology.Over time, there emerged a need for the administration to allocate time towards team communication, acquire equipment, and maintain the program.The suppliers of durable medical equipment supplied pulse oximeters, primarily receiving reimbursement from patients' insurance after the HMP team submitted a letter of medical necessity.Patients' caretakers were instructed in measuring and documenting the patient's daily percutaneous oxygen saturation (SpO2), heart rate, weight gain and feeding protocol at least twice a week.Surveillance criteria were measured using pulse oximeters and sensitive digital infant scales.Parents were informed that resting SpO2 less than 70%, weight loss of 30 g or failure to gain 20 g of weight for 3 days were to be considered critical and that in the respective case a call to the hospital would have to be made (12).Furthermore, parents received the usual discharge information to recognize warning signs for respiratory or gastrointestinal illness, respiratory distress, such as tachypnea and nasal flaring, or changes in perfusion.If problems arose, any necessary action was decided upon depending on the severity of the symptoms.

Identification of factors affecting mortality
Interstage mortality, defined as mortality between hospital discharge and BCPS, was analyzed using Cox-regression model.Pre-, intra-and postoperative variables were analyzed as risk factors.

Follow-up data
With the exception of the Hospital Management Program (HMP), all patients received outpatient follow-up care from pediatric cardiologists.The survey concluded for patients who experienced mortality by recording the time of their death as the endpoint.The cause of death was not ascertainable for patients who experienced death outside of a healthcare facility.
The subsequent data were consistently monitored and revised utilizing our institutional database system specifically designed for single-ventricle cases.
In our imaging laboratories, the grading of aortic valve regurgitation (AVVR) was performed using transthoracic echocardiography, which follows established principles that entail the semiquantitative assessment of AVVR through the utilization of Doppler color flow.This evaluation is based on the assessment of the ratio between the area of the regurgitant jet and the area of the atrium.A regurgitant color jet area-toatrial area of less than 30% signifies mild aortic valve regurgitation, while a range of 30%-50% indicates moderate regurgitation, and a value exceeding 50% indicates severe aortic valve regurgitation.

Statistical analysis
Categorical variables are presented as absolute numbers and percentages.Continuous variables are expressed as medians with interquartile ranges (IQR).An independent Student´s t-test was used to compare normally distributed variables.The Mann-Whitney U test was used for variables that were not normally distributed.Survival after hospital discharge and survival after BCPS was estimated by the Kaplan-Meier method.Risk factors for mortality were identified using uni-and multi-variate Cox regression models.Weight fo age z-score (WAZ) was calculated using the WHO Anthro software 3.2.Data analysis was performed using SPSS version 25.0 for Windows (IBM, Ehningen, Germany) and R statistical software 4.2.1 (R Foundation for Statistical Computing).

Interstage mortality
A total of 21 cases of interstage mortality were observed during the period between hospital discharge and the next stage of the BCPS procedure.The cause of deaths were 12 cases of   2).Nevertheless, a cohort comprising 71 individuals waere excluded from receiving HMP treatment due to either their lack of residency in Germany or their inability to consistently provide updates on their health status for diverse reasons.The interstage mortality of patients who were included into the HMP (n = 2, 2.5%) was lower than the mortality of patients without HMP (n = 19, 11.6%, p = 0.031).Risk factors for interstage death are shown in Table 3. Lower birth weight (p < 0.001), weight at Norwood (p = 0.016), longer ICU stay (p = 0.005), need for ECMO support (p = 0.002) and lack of HMP (p = 0.048) were identified as risk factors using univariable analysis.Multivariable analysis identified lower birth weight (p = 0.003) as an independent risk factor.When we selected 151 patients who had undergone the Norwood procedure since 2013 and performed the same analysis, weight at Norwood (p = 0.008), need for ECMO (p < 0.001) and lack of HMP (p < 0.001) were identified as independent risk factors (Supplementary Table S1).

Characteristics at the time of BCPS
Characteristics at the time of BCPS in patients with and without HMP are shown in Table 5A.Age at BCPS (p = 0.015), and Z score of weight at BCPS (p = 0.010) were significantly lower in patients with HMP.Mortality after BCPS was similar between the groups (p = 0.500).In patients with HMP, the patient characteristics at the time of BCPS with and without readmission are shown in Table 5B.Age at BCPS (p = 0.001) and weight at BCPS (p = 0.007) were lower in patients who needed readmission.However, mortality after BCPS was similar between patients who needed readmission and those who did not (p = 0.614).

Comment
Interstage mortality was lower in patients with HMP compared to those without.Lower birth weight, ECMO support after the Norwood procedure, and lack of HMP were risks for interstage mortality.Most frequent events during home monitoring were Survival after hospital discharge in patients with and without HMP.low percutaneous oxygen saturation below 70%, infection stagnated weight gain.One-third of the patients needed unintended readmission.Age and weight at BCPS were lower in patients who needed readmission compared to those who did not.However, survival after BCPS was similar between the groups with and without readmission.

Interstage mortality after the Norwood procedure
Interstage mortality accounts for an increasing proportion of mortality after the Norwood procedure.Most of interstage mortality occurred in patients with shunt dependent physiology between stage I and stage II palliations.Incidence of interstage mortality is reported to be up to 12% in a multi-center trial (4).Previous studies have identified several pre-and postoperative variables as risk factors for interstage mortality (13)(14)(15)(16).

Home monitoring program
In 2013, our facility launched an interstage HMP, which enabled more thorough patient monitoring as well as more frequent checkins.The unintended benefits were earlier recognition of inadequate growth and the elimination of growth failure that is commonly reported in infants with shunt-dependent heart disease (11).During the interstage period, 30% of patients were re-admitted for observation without cases of cardiac shock.Interstage mortality was reduced significantly from 12.4% to 2.2% after the introduction of the HMP.Two mortalities under HMP occurred in its early era (one patient in 2013 and another in 2014).Since 2015, we experienced no interstage mortality in patients who were supported by HMP.When compared to previous single-and multicenter reports, this represents a significant drop in overall mortality as well as morbidity rate between stages (9,(17)(18)(19).
In addition, our findings demonstrate that the introduction of HMP decreases interstage mortality in patients undergoing the Norwood procedure.The mortality rate decreased significantly throughout the staged intervention, cutting down from 12.4% to 2.2%.After the start of HMP, a significantly higher number of patients was unable to be discharged prior to BCPS (12/57 vs. 8/105, p = 0.022).This directly resulted from the implementation of HMP.At the time of BCPS, the children who had been part of the HMP were significantly younger and significantly lighter than those who were part of the remainder.At that time, their ages ranged from 67 to 299 days.After BCPS, early survival rate was not affected in any discernible way.
HMP identifies physiologic disturbances before decompensation and provides an early warning system for prompt interventions.This resulted in a significant reduction of interstage mortality in our population.31% (n = 14) of the patients needed interstage treatment because of HMP concerns and of these, 8 patients required early surgical procedures.
In accordance with previous publications, we matched our study participants to a historic control group.Therefore, additional Freedom from readmission after hospital discharge.confounding factors, such as developments in care over the past 19 years, may account for the reported outcomes (9,10,20).However, we believe that surgical modifications and the improvement of care protocols are more likely to influence the early postoperative period.The most important factors affecting the interstage period, including imbalanced systemic and pulmonary circulation, congestive heart failure, the risk of shunt stenosis, and recurrent aortic arch stenosis, are comparable amongst the study groups (20).
Refining of the HMP and future prospective Hansen et al. implicated that the thorough training of parents before discharge is crucial to improve their understanding of potential problems their child might encounter at home and to emphasize the importance of using the monitor continuously in situations where they are not directly interacting with the child (9).The HMP began with desired thresholds for SpO2 and weight change.The programme expanded to include circulatory insufficiency and failure to thrive based on initial findings, patient outcomes and parental feedback.Several programming enhancements were implemented concurrently to increase the efficiency of this home surveillance technology.Before discharge, dietary targets, standard equipment, systematic data-recording, educational materials, and room-in parent care were set.Imaging and lab data are collected during interstage clinic visits only when clinically necessary.Time required for pre-S2P echocardiography and cardiac catheterization varies by patients.The effectiveness of the HMP has been improved, in our opinion, by conducting consultations over the phone.The parents took advantage of additional counselling by phone, and the consistency of therapy gave the families additional confidence.
These findings do not indicate whether improved survival is due to more precise measurement of physiologic variables, improved care coordination, or both.The HMP is a complex care paradigm that improves communication between families and healthcare providers.These newborns are cared for at home by a multidisciplinary team of clinicians who work closely together.In our experience, there was a lower threshold for contact between parents and HMP NPs as a result of the surveillance calls.Furthermore, coordination between community cardiologists and the surgical site was improved, as was the transition from inpatient to outpatient treatment.When a local cardiologist or pediatrician is unavailable, the HMP provides an additional resource for the family or emergency medical personnel.The Norwood procedure with HMP is a model with far-reaching implications for the care of children with chronic conditions outside of the hospital, enabling multidisciplinary care coordination across pediatric disciplines.

Study limitations
All studies are based on observational, non-randomized data from a single center, limiting the study's ability to be generalized.We examined the study group to a historical control group, so any confounding variables, such as advances in supportive care over 19 years, could potentially account for the observed result.In addition, recent patients in the control group were in part coming from abroad, some of them from regions, where medical service is not as developed as in Germany.Thirdly, some patients were in a too bad condition to be discharged until BCPS and could not go into HMP.We had insufficient information regarding critical events in the control group throughout the interstage period, and only parents or local physicians reported the cause of death.None of the infants were subjected to postmortem examinations.

Conclusions
The home monitoring programme, which allowed for timely intervention, primarily early stage 2 palliation, reduced interstage mortality significantly.A greater number of patients remained hospitalized until the bidirectional cavopulmonary shunt was installed as a result of adhering to the surveillance criteria prior to discharge.In the surveillance group, earlier BCPS had no negative effect on early survival after bidirectional cavopulmonary shunt.

FIGURE 1 Flow
FIGURE 1Flow chart of patients undergoing the Norwood procedure.

TABLE 1
Patient characteristics.

TABLE 2
Operative-and Post-operative variables.-cardiac-output, 4 sudden deaths, 2 progressive hypoxias, and 3 unclear causes of death.Since the implementation of the Hospital Management Program (HMP) in our institution in 2013, a total of 80 patients have been enrolled (Figure HMP, home monitoring program; AA, aortic atresia; MA, mitral atresia; AS, aortic stenosis; MS, mitral stenosis.The bold values in the analysis indicate the most statistically significant results.low

TABLE 3
Analysis of candidate risk factors for interstage mortality following the Norwood procedure.

TABLE 4
Events of Home Monitoring Program (HMP).

TABLE 5
Characteristics at time of BCPS.The bold values in the analysis indicate the most statistically significant results.