Literature search

Thirty-two publications were identified for potential study inclusion. Two non-English publications were excluded, leaving 30 for analysis.

Prevalence: the dimension of the problem

CHD occurs in 4–6 per 1000 live births,20 ,51 ,52 and left-to-right shunt lesions represent the majority.1 ,3 ,20 ,51 ,52 The incidence may be higher if silent lesions are also included.51 ,52 A recent publication from the metropolitan Atlanta area reveals that the most common defects seen between 1998 and 2005 were muscular VSDs, perimembranous VSDs, and secundum ASDs, with a prevalence of 27.5, 10.6 and 10.3 per 10 000 births, respectively.20 Similar results were seen in the 2004 Danish registry.54

Many left-to-right shunts present as silent defects with little clinical significance1–3 but with important economic and psychosocial implications as they represent a significant portion of routine clinical practice.1–3 In fact, a recent study reveals that PDAs, VSDs, innocent murmurs and branch pulmonary artery stenosis during infancy account for 91% of diagnoses for patients presenting with a murmur at a tertiary centre, and the decision to discharge versus follow-up is usually not consistent.3 Additionally, another study reveals that cardiology referral for evaluation of a Still’s murmur does generate parental anxiety.46

Patent foramen ovale and small interatrial defects

PFOs and small ASDs are frequently encountered during routine CHD screening.3 The prevalence of secundum ASDs is approximately 10.3–10.8 per 10 000 births.20 ,46–48 There are no standardised criteria to classify the size of ASDs7–18 and establish follow-up. Table 1 depicts different thresholds that have been used to classify defect size.8–13 More recent publications have defined ASDs as having a diameter >4 mm, and a PFO is no longer considered a CHD.19 ,20

However, a recent survey highlights the confusion over ASD size and the differences between a PFO and an ASD.18 Most cardiologists use the size or the presence of a flap to differentiate between the two, whereas others use the echocardiographic width of the jet traversing the septal opening, the visibility of the opening, or its location within the atrial septum. Different interpretations may lead to confusion in the management of the defects,18 particularly in terms of follow-up (table 1). More recent publications recommend a more prudent approach given the natural history of these defects.18

Defects with diameters <3 mm generally close or become smaller early in life,8 ,9 ,13 while those with diameters 3–6 mm9 ,11––13 close spontaneously in at least 80% of cases. By contrast, spontaneous closure is improbable in defects with diameters >8 mm,8 ,9 ,11–13 occasionally increasing in diameter over time. The presence of an atrial flap is more frequently associated with spontaneous closure,16 while an associated PDA may have a negative prognostic significance.16 The influence of prematurity has been identified as positive by some10 and negative by others.12 Some authors suggest that infants with ASD <3 mm do not need follow-up as they all close by 18 months of age.9

The definition of spontaneous closure has also been problematic, particularly since complete closure occurs in most patients with a small number having a residual hole that is significantly smaller than seen at presentation. Interestingly, the reported closure rate in some studies9 ,11 ,12 is discordant with the reported 25% prevalence of PFOs from other studies.2 ,55

For bigger defects with diameters of 6–7 mm after 1 year of age, yearly follow-up is recommended until 4 or 5 years of age.7 The survey described previously reveals that most paediatric cardiologists see patients with secundum ASD in follow-up on a yearly basis for small defects and more often for larger ones (40–45% see the patients at 6-month intervals).18 Patient age and the presence of symptoms are also important determinants in the frequency of follow-up in this study. Another survey from the UK reveals that, for children over 3 years old with an ASD and a normal right heart, 60% of physicians recommend follow-up every 3 years, 25% recommend follow-up every 5 years, and many recommend discharge in late adolescence.2 By contrast, for children over 5 years old with an incidental PFO, 81% recommend immediate discharge accompanied by a discussion of the implications of scuba diving. Only a small number recommend follow-up every year (8%) or every 3 years (10%) (table 1).

Small interventricular defects

Despite the fact that VSDs are the most common CHD,20 there are no clear criteria pertaining to management (table 2). Some authors have defined small VSDs as having diameters <3 mm by echocardiography,26 but other definitions have been published using other modalities.27 Normal pulmonary artery pressures and left heart size from echocardiography and a normal electrocardiogram are other criteria used to define small defects.2 Many studies have revealed a benign natural history for most VSDs.10 ,21–23 ,25 ,27 ,32–36 Muscular VSDs tend to close in 80% cases early in life,21 ,22 ,25 ,32–36 a frequency that is at least twice the frequency of spontaneous closure for perimembranous VSDs.25 Spontaneous closure may occur up to age 6 years25 or beyond school age for perimembranous VSDs.23 Interestingly, the presence of partially occluding accessory tissue is not predictive of spontaneous closure.23

At present, there are no standards for follow-up of small VSDs, though more frequent follow-up has recently been recommended. The UK survey previously described reveals that, for a thriving child over 6 months old with a small VSD, 67% advocate follow-up every 3 years, whereas 15% recommend annual follow-up.2 Most agree that these patients should be discharged as adolescents.

Published indications for VSD closure originate from an era with high operative mortality (up to 7.5%) and incidence of significant residual shunts (up to 20%). However, surgical correction for a small VSD in the current era is classified a low-risk procedure with only a 2% complication rate.56 Improved surgical techniques have led to revised recommendations for closure of small VSDs.29 ,30 Additionally, data on long-term sequelae for small VSDs have been discordant. Endocarditis occurs in 1.8–15%,27 ,31 aortic regurgitation occurs in 2–20%,27 and arrhythmias occur in up to 10% of cases.27 ,31 Differences in patient populations, classification systems, and management may explain some of these contrasting results.

Varied natural histories may also explain the heterogeneity in management for these patients. Some recommend surgery for all conal VSDs, VSDs with aortic valve prolapse (even without regurgitation), and VSDs associated with a harsh murmur persisting beyond age 5 years.30 By contrast, others suggest that surgery is not necessary if the pulmonary-to-systemic flow ratio is less than 2, there is no left ventricular volume overload, the pulmonary arterial pressure is not elevated, and there is no aortic regurgitation or associated symptoms27 (table 2).

Small patent ductus arteriosus

PDAs account for 10% of all CHDs with an incidence of 1 per 2000 term births.44 ,45 Small PDAs represent 3% of these cases. Up to 0.5% of patients with an innocent murmur have a small PDA.41 The definition of a small PDA has not been well established. Criteria to differentiate a small PDA with a characteristic murmur but no haemodynamic consequences from a very small and silent PDA have been proposed,45–47 but clear echocardiographic criteria have not been established (table 3).

The age at diagnosis may be important. In a recent Atlanta CHD registry, PDAs are not considered CHD if the diagnosis is made in premature infants (gestational age <36 weeks) or in newborns (<6 weeks in age).19 Whether and how these children should be followed has not been discussed. In fact, there are no clear guidelines on the follow-up of small PDAs. In the same UK survey, for children over 3 years old with a tiny PDA, normal pulses, no heave, no ductal murmur, normal left heart size and normal pulmonary artery pressures, 46% recommend discharge, 35% recommend follow-up every 3 years and 14% recommend follow-up every 5 years.2 Two opposing recommendations emerge with free texting in the survey: one group favouring discharge and the other percutaneous closure.

Closure of large and haemodynamically significant PDAs is standard of care and can be performed safely and effectively with surgical or percutaneous methods. Small silent PDAs have been controversial with a clear division between intervention43 ,45 and conservative management.46 ,49 ,50 The remote risk of developing infective endocarditis (IE)34 should be weighed against the potential complications associated with intervention.37

The risk/benefit balance has been interpreted differently by various authors.44–50 A recent review states that only three cases of IE have been described in the past decade for silent PDAs,39 ,46–52 thereby estimating the risk of IE at around 0.01% per year (0.1% life-time risk over 85–90 years) using a conservative approach (a more liberal approach decreases the risk to 0.001% per year). They highlight the difficulty with accurate estimation of IE risk since prior studies do not take PDA size into consideration. Major adverse events for percutaneous PDA closure occur in 0.5–1.5%. Additionally, incomplete closure probably results in a higher IE risk than in similarly sized PDAs. Conservative estimates of the incidence of IE and percutaneous closure complications result in similar risk for closure versus non-closure, whereas more liberal estimates highlight the disadvantages of intervention.

Lastly, routine PDA closure represents a significant economic impact, with costs estimated at US$33 billion for IE prevention.39 ,49 Most recent works do not favour intervention for small PDAs,39 ,49 ,50 but closure is still recommended for all patients regardless of PDA size in some centres.45 ,47