The impact of neonatologist performed echocardiography in an Italian neonatal unit

1 Introduction

Neonatologist performed echocardiography (NPE) is a wide-spreading practice in the neonatal units all over the world. NPE, as opposed to echocardiography performed by cardiologist, focuses on functional and hemodynamic changes of sick neonates in order to enrich the clinical evaluation, support the decision making, monitor the efficacy of therapy, and possibly improve the neonatal outcomes [1–4].

Recommendations for NPE state that the diagnosis of structural heart disease is beyond its goals [5–7]. Nevertheless, the neonatologists performing functional echocardiography should be confident in confirming normal heart anatomy; the cardiology consultation is then recommended in a reasonable time frame [5]. It is acknowledged that many neonatologists with specific training take on an additional role in diagnosis of congenital heart disease, mostly because they don’t have pediatric cardiology services readily available.

We conducted a 2-year retrospective study to review the echocardiographic activity performed by the neonatologist for functional and structural evaluations in an Italian neonatal unit.

2 Methods

This is a retrospective analysis of echocardiograms conducted by one neonatologist on newborns admitted to the neonatal unit and the postnatal ward of the Hospital San Pio of Benevento, Italy, between January 2019 and January 2021. As a local inborn and outborn referral center for neonatal care, our unit is a 12 bedded facility providing intensive and sub-intensive neonatal care for sick newborns. Healthy newborns stay together with their mothers throughout the birth hospitalization in the maternity unit (rooming-in) and they are also cared for by the neonatal team. There are two options if the cardiology consultation is deemed necessary; the cardiology service, which has just one cardiologist (NT) app-ropriately trained in pediatric cardiology, who, however, is not available on-call 24/7; and the regional pediatric cardiology center, housed 100 kilometers away from our unit. Therefore, it should come as no surprise that neonatologists started to perform echocardiography 10 years ago in our unit. Indi-cation-based echocardiograms are requested by the neonatologist in charge and mostly performed by one trained neonatologist. A study imaging protocol covering thoraco-abdominal situs, cardiac position, left or rightward orientation of the ventricular loop, atrio-ventricular and ventricular-arteries alignments, associated abnormalities of the cardiac chambers, septa, outflow tract and great vessels, venous connection, and possibly study of the coronary arteries, is implemented besides the functional evaluation. Whenever structural abnormalities are suspected or detected, patient referral to the cardiologist is mandatory.

The echocardiograms included in the study period were performed by the same neonatologist (AC) using either the LogiqP5 pro ultrasound (GE Healthcare Ultrasound Wauwatosa, WI, USA) or more re-cently the Arietta 850 machine (Hitachi, Tokyo, Japan), coupling with the sectorial multifrequency probe. Any further pediatric cardiology consultations were completely at the discretion of the neonatologist on an individual basis. Data collected for each echocardiogram was: gestational age, age of life in days at first scan, indication for performing the echocardiography, whether or not ECG was req-uested, findings, and planned management. Indications for echocardiography were categorized in fetal indication, genetic syndrome/malformation, neonatal ward admission (e.g. prematurity, respiratory distress syndrome, sepsis), re-examination by the neonatologist (recheck), maternal diabetes, brady-arrhythmia, cardiac murmur, family history of congenital heart disease. Screening indicated a miscellaneous group which included abnormal pulse oximetry screening at discharge, weak or absent femoral pulses at the clinical examination, and apparent life-threatening events. Management options were broadly categorized as medical treatment, recheck, and referral. Cardiology reviews, when available, were used to evaluate the rate of concordance between neonatologist and cardiologist.

The database was developed and analyzed (pandas.pydata.org). The quantitative data was expressed as median values and interquartile ranges and qualitative data as percentages and frequencies. A compa-rison of frequencies between groups (e.g. scans with and without cardiac anomalies) was made excluding echocardiograms with transitional and doubtful findings. Fisher’s exact test was used, with p≤0.05 considered statistically significant. The inter-rater agreement between neonatologist and cardiologist was measured using Cohen’s k coefficient [8].

3 Results

A total of 160 echocardiograms done by the neonatologist in 136 infants during the period investigated were reviewed retrospectively. The majority were full term neonates (median GA 38 weeks, IQR 36–40 weeks) with a median (IQR) age at the time of the first scan of 3 (2–7) days. ECG was requested in a minority of echocardiograms (17.2%). The most common indication for echocardiography was admission to the neonatal ward (31.9%of the studies), followed by fetal indication (15%), recheck (15%), and screening (13.8%). More than one-half of the scans (55%) were normal, whereas 27.5%showed pathological findings. Transitional abnormalities (mainly patent ductus arteriosus and/or tricuspid valve regurgitation) were found in 12.5%. Echocardiograms with doubtful findings were 5%, mostly related to the difficulty in distinguishing between patency of the oval foramen and small atrial septal defect. Among 52 scans with suspected or detected anomalies, septal defects accounted for 48.9%. The functional abnormalities reported were mainly persistent pulmonary hypertension of the neonates, hemodynamically significant patent ductus arteriosus of premature, heart failure in asphyxiated newborns, and hypertrophic cardiomyopathy in neonates of diabetic mothers (Fig. 1). Infants with abnormal findings were significantly more likely than those without fetal indication. Medical treatment (19.7%), recheck (40.8%), and cardiological referral (39.4%) were the management decisions resulting from 71 echocardiograms.

OPEN IN VIEWER

Fig. 1

Echocardiograms with abnormal or suspected findings. ASD, atrial septal defect; VSD, ventricular septal defect; PPHN, persistent pulmonary hypertension of the newborn; persistent sPAP, persistent systolic pulmonary arterial pressure; HCM, hypertrophic cardiomyopaty; TAPVC, total anomalous pulmonary venous connection; CAVC, common atrio-ventricular canal; D-TGA, D-transposition of the great arteries; AOCA, abnormal origin of the coronary arteries; ARSA, aberrant right subclavian artery; PAH, pulmonary artery hypertension.

Among 136 scanned infants, 108 were evaluated by the neonatologist alone, due to being considered normal in the first scan or later at the recheck, and/or managed for hemodynamic problems; 28 infants were eventually referred to cardiology consultation. The cardiology reviews were available for analysis in 15 patients only (about half of the referrals). Data is summarized in Table 1.

The study showed 66.7%of total agreement, 20%of partial agreement. The rate of concordance for ventricular septal defects (the most common defect confirmed by cardiologists) was 93.75%with a Cohen’s k coefficient computed at 0.86.

To our knowledge, this is the first study that des-cribes the activity of neonatologist performed echo-cardiography in an Italian neonatal unit.

In our hospital, the neonatologist with sound expertise regularly executes echocardiography for both functional and structural evaluations, following a rigorous study imaging protocol. Our unit has dedicated ultrasound machines that allow neonatologist to do echocardiography for a wide range of indications and in a timely manner. In this cohort we found a low incidence of cardiac abnormalities and there were no obvious strong associations between cardiac anomalies and indication type, globally. However, defining the indications in which echocardiography provides additional value to neonate management falls outside the scope of our review.

The median age of life in the first scan explains the high rate of re-examinations, given that the younger the age the higher the probability of transitional findings (e.g. tricuspid valve regurgitation, large shunt through the patent ductus arteriousus, or the interatrial septum). This put forward the question, whether it is more convenient to postpone the first echocardiographic scan, whenever indication and clinical condition are favorable. ECG was an underused tool by neonatologist, because it is allegedly perceived as not very informative for the diagnosis.

Due to the retrospective nature of the study, we missed collecting several patent ductus monitor studies in premature neonates who did not require treatment, as well as, echocardiograms to confirm tip location of central venous catheters, that would have added further value to the functional echo performed in our unit [9–11]. Moreover, the cardiology reports were available for comparison in around half of the referred infants only. The global rate of concordance between neonatologist and cardiologist was 86.7%, which is quite good. More specifically, the neonatologist recognized unexpected major congenital heart diseases and several minor lesions. The cardiologist detected additional heart anomalies in two neonates sent to pediatric cardiac consultation anyway. The two cases of over diagnosis were challenging congenital cardiopathies (e.g. total anomalous venous connection and anomalous origin of the coronary arteries). Finally, the cardiologist did not confirm a small apical ventricular septal defect in an infant evaluated six months later, but a spontaneous closure could have occurred overtime (Table 2).

The former debate about the ownership of neonatal echocardiography has been overcome by good concordance between well-trained neonatologists and cardiologists [12–15]. Diagnosis and sometimes follow-ups of congenital heart disease are well-recognized practices among neonatologists [6]. An Italian survey on NPE showed that structural echocardiography on stable neonates is undertaken by neonatologists alone in 46%of the centers [16]. A strict collaboration with pediatric cardiologists is recommended by guidelines [5–7]. This is essential to arrange the cardiology consultation, in case of doubts or anatomical abnormalities, and provide advice for managing very critical situations. Live or remote supervision by cardiologists improves congenital heart disease diagnosis [17]. Photographs and videos sent by mobile phone via instant messaging applications might be a promising tool for productive cooperation among specialists (personal experience).

In our cohort, a larger number of infants, having been assessed by the neonatologist alone, cannot be excluded for having underestimated congenital malformations. Nevertheless, it seems unreasonable, at least with regard to major cardiac anomalies, on the assumption that if the neonates are considered at low risk of critical congenital heart disease, like our population study, the rate of agreement between neonatologist and cardiologist is reported fairly high [18]. We do believe in the utility of predischarge echocardiography by neonatologist with advanced experience, whenever an indication is set. It may be twice as beneficial because it reduces parental anxiety [19] and filters out those infants who does not require to be referred to the overcrowded pediatric cardiology services.

Nonetheless, we are aware that undertaking the responsibility of screening our neonates for congenital malformations is a potential minefield because of the risk of missing minor lesions [20], therefore larger studies are needed to assess the risk-benefit balance of this approach.

In summary, NPE has a great potential for managing hemodynamically unstable newborns. Additional structural assessment by skilled neonatologists, using a rigorous segmental approach, may have a positive impact, as long as the neonatal population of study has a negligible risk of congenital heart malformations. We advocate the need of robust training and certification programmes, as well as, live and remote collaboration with pediatric cardiologists in order to optimize the benefits and reduce the risks of this practice.