Anaesthesia for inguinal hernia repair in the newborn or ex-premature infant

Learning objectives.

By reading this article, you should be able to:

• •Explain the risks and benefits of general and regional anaesthesia for inguinal hernia repair in infants.
• •Identify which patients are at greatest risk of complications.
• •Select appropriate postoperative monitoring for ex-premature infants.
• •Discuss the challenges of laparoscopic hernia repair in infants.

Key points.

• •Inguinal hernia repair may be an elective or an emergency procedure.
• •Ex-premature infants and neonates are at increased risk of perioperative complications.
• •A comprehensive understanding of developmental physiology and pharmacology is essential.
• •Awake spinal anaesthesia without sedation reduces the risk of early apnoea.
• •Infants at risk of respiratory complications including apnoea should receive extended cardiorespiratory monitoring.

A hernia is defined as the protrusion of an organ or tissue through the wall of the cavity in which it is contained. Most inguinal hernias presenting in infancy are indirect. An indirect inguinal hernia arises as a consequence of herniation of bowel through the deep inguinal ring in the presence of a widely patent processus vaginalis.

The processus vaginalis is a diverticulum of peritoneum which, in boys, is pushed down the inguinal canal as the testes descend into the scrotum. A patent processus vaginalis provides an open communication between the peritoneal cavity and the scrotum. In the course of normal development, the processus vaginalis closes spontaneously by the age of 2 yr, with closure usually occurring in the first 2 months of life. In girls a corresponding protrusion of peritoneum extends from the uterus to the labia majora and is known as the diverticulum of Nuck. This normally closes at a gestational age of around 7 months.

The incidence of inguinal hernia is highest in the first year of life and there is a peak in the neonatal period: the incidence in newborn infants is 1–5%. Inguinal hernias are more common in premature infants with an incidence of around 11%. In extremely low birth weight (LBW) infants (i.e. those weighing less than 1000 g), inguinal hernia is very common with rates of around 40%.¹ The incidence of incarceration has been considered to be extremely high in premature infants, but this finding has not been consistently reproduced across all studies.1, 2 A recent large retrospective study showed that the overall incidence of inguinal hernia in children was 6.62% in males and 0.74% in females.²

The presentation and features of inguinal hernia in males and females reflect aspects of development in utero. Sixty percent of congenital inguinal hernia are right-sided, 30% are on the left. This pattern is a consequence of delayed descent of the right testis and a persistence of patency of the processus vaginalis on the right side for longer than the left during infancy in boys. Ten percent of hernias in term neonates are bilateral; this bilateral presentation is more common in premature and LBW infants, occurring in up to 50% of patients.1, 3

In addition to prematurity and low birth weight, inguinal hernia in infants may also be associated with other conditions:¹

• (i)urological anomalies—hypospadias, cryptorchidism, bladder extrophy;
• (ii)increased intra-abdominal pressure—presence of ventriculo-peritoneal shunts, ascites, peritoneal dialysis;
• (iii)abdominal wall defects—exomphalos and gastroschisis;
• (iv)family history of inguinal hernia;
• (v)other conditions (e.g. cystic fibrosis, mucopolysaccharidosis, Ehlers–Danlos syndrome, and Marfan syndrome).

Presentation

There is usually a history of a painless and intermittent swelling in the groin; the swelling may have been noted to be associated with straining. There may be no mass on examination or a reducible groin mass may be found. A hernia that cannot be reduced by manipulation is incarcerated. Bowel within the hernia can become erythematous and trapped within the hernia sac, and this can lead to bowel obstruction. In females the sac may contain an ovary. There may be vascular compromise of the entrapped contents of the hernia as a result of progressive swelling and oedema; at this stage, the hernia is described as strangulated. Bowel perforation may result and, rarely, bowel resection may be necessary. An infant with a strangulated hernia will be extremely unwell.

Surgical management

The definitive treatment for any hernia is surgical repair: reduction followed by surgical closure of the patent processus vaginalis and repair of the floor of the inguinal canal. The timing of surgery is determined by the presentation and the co-morbidity, but the optimal timing is controversial.¹ An asymptomatic hernia may be scheduled electively but remains at risk of incarceration. The risks of anaesthesia in infancy must be balanced with the risk of incarceration, which is highest in the first 6 months of life, and surgical preference is usually to schedule surgery sooner rather than later.1, 3

An incarcerated hernia carries a risk of serious complications such as intestinal obstruction and strangulation of the hernial sac contents, and must be reduced. Manual reduction is usually attempted. After successful manual reduction, there is a risk of reincarceration. Open surgery after successful manual reduction may be technically more challenging because of localised swelling, and some surgeons will admit, observe, and delay (e.g. 24–48 h) definitive repair for a short interval after manual reduction. If manual reduction fails, then surgical reduction is indicated and should be undertaken as an emergency.

Some patients presenting with a unilateral hernia may be at risk of developing a second hernia on the contralateral side, and some surgeons will also explore the contralateral side during surgery.1, 3

Laparoscopic repair is an option for the management of inguinal hernia in infancy. Advantages of laparoscopic repair include better visualisation of the anatomy with better ability to explore the contralateral side and ligate if indicated. Minimally invasive surgery is also associated with improved cosmesis, reduced length of stay, and improved postoperative pain scores.3, 4 Anaesthetic considerations for laparoscopic surgery in infants are outlined below.

Age terminology

A premature infant is born before 37 completed weeks of gestation. An ex-premature infant is born before 37 completed weeks of gestation and is now older than 37 weeks of age. The American Association of Pediatrics (AAP) recommends the following definitions of age terminology:⁵

• (i)Gestational age (in completed weeks) is the time elapsed between the first day of the last normal menstrual period (LNMP) and the date of birth. In the setting of assisted reproductive technology, gestational age is calculated by adding 2 weeks to the conceptional age.
• (ii)Postmenstrual age is the sum of gestational age plus chronological age (weeks).
• (iii)Chronological age (days, weeks, months, or years) is the time elapsed from birth.
• (iv)Corrected age (weeks or months) is the chronological age minus the number of weeks born before 40 weeks of gestation; this term should be used only for children up to 3 yr of age who were born preterm.

Anaesthetic management

Conduct

A range of patients may present for elective hernia repair; the group is heterogeneous and includes premature infants, ex-premature infants, and term infants. A thorough assessment of both the developmental stage of the infant and any co-morbidity is essential. The timing of surgery should be carefully considered in each patient. Multidisciplinary care can be helpful in the more complex premature or ex-premature infants.

Table 1 highlights some of the factors to consider when anaesthetising any neonate.

Table 1.

Factors relevant to the provision of anaesthesia for neonates

System	Consideration	Precautions
Airway	Large head with prominent occiput. Short neck. Epiglottis is large, floppy, U shaped, and cephalad. Airway calibre is smaller and of relatively high resistance. Trachea is short, compliant and prone to malacia particularly in the ex-premature infant. Upper airway reflexes are pronounced. Ex-premature infants may have acquired sub-glottis stenosis.	Avoid flexion and hyperextension during mask ventilation as this may obstruct the airway. Avoid gastric insufflation during mask ventilation. Consider choice of technique for intubation (e.g. a straight blade may be useful to elevate the epiglottis at direct laryngoscopy). Awareness and anticipation of risks of (e.g. laryngospasm). Availability of a range of tracheal tubes sizes with careful selection of tube size particularly during laparoscopic surgery. Careful positioning of tracheal tube tip and avoidance of endobronchial intubation.
Respiratory system	The respiratory system, notably alveolar development, is immature. Respiratory control is immature. Respiratory mechanics are altered with predominantly diaphragmatic breathing. High ventilatory frequency and work of breathing hence during spontaneous respiration the infant is prone to fatigue. The diaphragm has less ‘fatigue resistant’ type 1 fibres. The chest wall is compliant with horizontal ribs. Functional residual capacity is low and is maintained by high ventilatory frequency and laryngeal braking (laryngeal adductors restrict expiration and maintain PEEP). The alveolar ventilation to functional residual capacity ratio is high. Ex-premature infants may have bronchopulmonary dysplasia. There is significant risk of postoperative apnoea in ex-premature infants. Anaesthesia ventilators do not perform as well as neonatal unit ventilators (e.g. flow sensors are located within the anaesthetic machine). This may lead to inaccuracies with, for example, tidal volume measurements. ETCO2 monitoring can be problematic particularly in premature infants.	During mechanical ventilation use a ‘protective’ or ‘open lung’ ventilation strategy to ensure adequate minute ventilation. Adjust ventilation to limit the tidal volume and peak inspiratory pressure in order to limit lung injury but ensure maintenance of functional residual capacity. Use positive end expiratory pressure. Consider the respiratory time constant when selecting inspiratory and expiratory times. Select an appropriate ventilatory frequency. Consider a longer expiratory time in infants with bronchopulmonary dysplasia. Ensure appropriate postoperative apnoea monitoring. Use equipment which minimises dead space. Ensure anaesthetic machine and breathing circuit check includes compensation for compressible loss (i.e. the compliance of tubing). Consider supplementary monitoring (e.g. use of transcutaneous CO2 monitoring), in a sick premature infant direct arterial monitoring may be indicated.
Cardiovascular system	Younger term neonates and premature neonates may have patent intra and extracardiac communications (e.g. patent foramen ovale and patent ductus arteriosus). Shunting may occur across these. Pulmonary vascular resistance is low compared with fetal values but higher than in older infants. Stiff non-compliant ventricles. Stroke volume relatively fixed. Rate-dependent cardiac output. High myocardial oxygen consumption.	Avoid rises in pulmonary vascular resistance (e.g. avoid hypoxia and hypercapnia). Consider monitoring pre- and post-ductal saturation if ductus arteriosus is patent. Maintain adequate preload, afterload and heart rate as fluctuations are poorly tolerated especially in the premature infant.
Renal system	The full complement of nephrons is present at birth in term infants, but renal vascular resistance is high, this decreases from birth. As the resistance decreases, an increasing proportion of CO is directed to the kidneys. Low GFR and immature tubular function. Immature kidney may not be able to excrete a water load and there is a risk of hyponatraemia.	Limit fasting time to 2 h for clear fluids and consider if a dextrose infusion is necessary Minimise transepidermal fluid loss in premature infants and adjust fluid balance to include increased evaporative insensible losses. Replacement fluid should usually be a balanced salt solution adjusted to maintain intravascular volume, any deficit and ongoing losses.
Hepatic and metabolic	Limited glycogen stores make neonates vulnerable to hypoglycaemia Immature hepatic enzyme systems result in altered pharmacokinetics (e.g. CYP isoenzymes).	Monitor blood glucose levels. Administer dextrose in premature infants (5–8 mg kg−1 min−1). Early return to feeding. Glucuronidation is immature in neonates so caution should be exercised with, for example, paracetamol, morphine, and propofol and doses adjusted to reflect decreased hepatic clearance.
Temperature regulation	Extremely prone to hypothermia.	Warm the environment, inspiratory respiratory gases and intravenous fluids. Use heating devices and monitor core temperature. Keep baby dry.
Immature oxidative systems	At risk of oxygen toxicity as a result of cellular activity of reactive oxygen species.	Use preductal oxygen saturation as a guide and titrate inspired oxygen. Avoid hyperoxia, aim for saturation level of 91–95% in premature infants.

Prematurity has an impact on many of the developing organ systems. The major comorbidities associated with prematurity are listed in Table 2. The risks of these comorbidities increase with decreasing gestational age. Extreme prematurity (defined as <28 weeks gestational age) has been extensively studied.6, 7, 8 Many of the findings arising from studies of extremely premature infants have a bearing on the care of all premature and ex-premature patients (e.g. persisting impairment of respiratory function, neurocognitive morbidity, and the vulnerability to oxidative stress).⁸

Table 2.

Major morbidities associated with extreme prematurity

Respiratory distress syndrome (RDS)

Bronchopulmonary dysplasia (BPD)

Patent ductus arteriosus (PDA)

Retinopathy of prematurity (ROP)

Necrotising enterocolitis (NEC)

Sepsis

Intraventricular haemorrhage (IVH)

Periventricular leucomalacia (PVL)

Choice of technique: general or awake regional anaesthesia

The choice of anaesthetic agent and technique must be informed by patient and surgical factors. There are few comparative studies that examine different general anaesthetic agents. One study did compare recovery after maintenance of anaesthesia with sevoflurane and desflurane, and found no difference in postoperative respiratory events.⁹ General inhalation anaesthesia with shorter-acting agents such as sevoflurane, supplemented with caudal or ilioinguinal blockade, is widely utilised for open inguinal hernia repair. Neuromuscular blockers and opioids should be used with caution because neonates and infants are particularly vulnerable to the respiratory depressant action of opioids and the effects of residual neuromuscular block. Neonates are more likely to develop respiratory complications, although, as described below, studies have not confirmed that the use of either opioids or neuromuscular blockers are risk factors for the occurrence of postoperative apnoea. Neonates exhibit extensive interindividual variability in pharmacokinetics and pharmacodynamics. This is pertinent to drugs such as morphine, which should be carefully titrated to effect when used.

Airway support must be tailored to the needs of the neonate. Placement of a tracheal tube and artificial ventilation is usual and appropriate for premature neonates, those with respiratory co-morbidity (e.g. bronchopulmonary dysplasia), or those undergoing laparoscopic repair, whereas a supraglottic airway device may be considered for the healthy term neonate undergoing surgery later.

Regional anaesthesia has a significant role in the management of infant hernia repair. There has been renewed interest in awake spinal anaesthesia in the light of concerns about the potential neurotoxicity of anaesthetic agents. Neurotoxicity is further discussed below. Neuraxial techniques used to facilitate inguinal hernia repair include spinal, caudal, and epidural blocks, usually as a single bolus injection or sometimes as an infusion via a catheter. They may be used as a sole technique (e.g. awake spinal), in combination (e.g. awake spinal plus caudal), or to supplement general anaesthesia. Ilioinguinal blockade and field blockade may also be used to supplement spinal or general anaesthesia. The duration of the block may limit the use of single shot spinal anaesthesia; surgery should be completed within approximately 60 min. It is also noteworthy that when used in isolation, the rate of failure and subsequent conversion from spinal to general anaesthesia is significant. Adjuvants can be used to prolong the duration of blockade, but the use of these should be considered carefully. Some adjuvants such as clonidine have sedative effects and could predispose to apnoea, although there is little robust evidence to provide guidance. The benefits of neuraxial anaesthesia include a reduction in respiratory complications (see below), but this is not sustained if sedatives are administered. Complication rates after neuraxial block in neonates are higher than in older children but severe complications, for example meningitis and neurological injury, are rare.10, 11

During a neonatal spinal anaesthetic, the infant is placed in the sitting or lateral decubitus position and lumbar puncture is performed at the interspace between L4 and L5; this can be identified using the intercristal line. In term neonates the conus normally lies at the L1–L2 level, but extends to L3–L4 in premature babies, and the dural sac usually terminates between S2 and S4. CSF volume is greater in neonates.¹¹

Full aseptic precautions should be taken and a sterile field established. Chlorhexidine 0.5% solution should be used for skin asepsis and allowed to dry before skin puncture. Spinal needles, 22–25G 25 mm are used; pencil point needles and needles with introducers are available.¹¹

Local anaesthetic drugs that have been used include bupivacaine, levobupivacaine, ropivacaine, tetracaine, and lidocaine, although the evidence comparing their efficacy is limited. There are also few systematic studies that have examined toxicity to the developing spinal cord. Suggested doses are isobaric bupivacaine, ropivacaine, or levobupivacaine 0.5% 1–1.2 mg kg⁻¹. Adjuvants used in conjunction with neuraxial blockade include fentanyl, morphine, dexmedetomidine, clonidine, and ketamine.10, 11

It is essential that all members of the team are sensitive to the needs of the infant during awake spinal anaesthesia, and care and attention should be given to the environment and lighting. The baby may benefit from soothing, and oral sucrose may be administered using a pacifier.¹¹

Within the past decade, there has been increasing concern about the potential for neurotoxicity, and subsequent neurological morbidity, as a result of exposure to general anaesthesia in neonates. This is after persuasive preclinical evidence emerged detailing neuronal apoptosis as a result of exposure to general anaesthetic agents, and also demonstration of long-term cognitive deficits in rodent models. However, several large multi-centre clinical trials have failed to provide conclusive evidence in human studies. The GAS study is an international, randomised, and multi-centre trial looking at the long-term effects of anaesthesia on the developing brain in infants.12, 13 Subjects recruited were ex-premature and term infants between 26 and 60 weeks postmenstrual age undergoing inguinal hernia repair. Those with an existing risk of adverse neurodevelopmental outcome were excluded. Subjects were randomised to receive general or regional anaesthesia. Regional anaesthesia consisted of spinal or caudal anaesthesia alone or spinal plus caudal or ilioinguinal block. General anaesthesia consisted of sevoflurane for induction and maintenance, and was supplemented with caudal or ilioinguinal blockade. Neuromuscular blockers were permitted, but opioid analgesia was not. At 2 yr of age, the study did not find any significant difference in psychomotor scores between the two groups. A single brief exposure to general anaesthesia is not felt to contribute to significant neurotoxicity at present.¹⁴

Laparoscopic surgery

Increased intra-abdominal pressure can restrict diaphragmatic excursion and this is exacerbated by head down positioning. Tidal volume and functional residual capacity are reduced. Ventilator settings may need to be adjusted to maintain adequate minute ventilation.

Pnuemoperitoneum is achieved with CO₂ insufflation, and intraperitoneal pressures should not exceed 10 mm Hg. Insufflation can lead to significant CO₂ absorption into the circulation, and this can contribute up to 20% of the exhaled CO₂.¹⁵ Ventilation will need to be adjusted accordingly.

In order to optimise ventilation and control arterial CO₂, a close-fitting or cuffed tracheal tube is necessary. Use of neuromuscular blockers may result in lower intra-abdominal pressure. Sidestream capnography, even with very low dead space systems, will often significantly underestimate arterial CO₂. In the smaller more premature infants, intermittent sampling from an arterial catheter or a continuous transcutaneous CO₂ monitor may better guide adjustments to ventilation.

Pneumoperitoneum also leads to compression of the abdominal vessels, and as a result of this an increase in systemic vascular resistance. The compression also leads to a decreased venous return and reflex tachycardia. This is poorly tolerated in the presence of intravascular fluid depletion.

Pulmonary vascular resistance decreases at birth as the circulation transitions but initially remains relatively high. The ductus arteriosus normally closes within 48 h in a well term baby, but there is a significant incidence of patent ductus arteriosus in premature infants. The presence of patent communications (e.g. ductus arteriosus or foramen ovale), can lead to shunting across the heart. Early neonates are at risk of pulmonary hypertensive crises with right to left shunting; this can be precipitated by hypoxia or acidosis and will be detected by a decrease in lower limb oxygen saturation relative to right upper limb oxygen saturation. Management should be directed at reduction of pulmonary vascular resistance. In the event of a gas embolus during laparoscopic surgery, the presence of intracardiac communications adds to the risk of adverse neurological sequelae as gas may traverse the communication between the atria.

Postoperative care

A retrospective analysis from 2011 analysed almost 300 infants undergoing herniorrhaphy and found that four variables predicted an increase in the duration of PACU stay; these were postmenstrual age <45 weeks, prematurity, general anaesthesia, and postoperative opioid administration (nalbuphine). In contrast, they found that both use of regional anaesthesia alone and the use of intraoperative regional anaesthesia (ilioinguinal and iliohypogastric block) predicted a decrease in the duration of PACU stay.¹⁶

Risk of postoperative apnoea

Postoperative respiratory complication rates are higher in infants, and this is further exaggerated in the ex-premature infant. Ex-premature infants are at increased risk of apnoea after inguinal hernia repair. The incidence of apnoea is inversely related to both postmenstrual age (the most important risk factor) and gestational age.13, 17

Most apnoeas occur in infants <44 weeks postmenstrual age and often resolve without intervention, but some need interventions such as tactile stimulation, administration of oxygen, bag and mask ventilation, continuous positive airway pressure, and cardiopulmonary resuscitation. Anaesthesia may unmask impaired and immature respiratory function and expose vulnerability in both central ventilatory control mechanisms, and the developmental immaturity of the chest wall and diaphragm and lungs.

It is important to recognise that apnoea may arise as a result of other co-existing pathology, for example neurological abnormalities, patent ductus arteriosus, intracerebral haemorrhage, metabolic disturbances such as hypoglycaemia, or evolving systemic illness such as sepsis. The aetiology of any apnoea must be carefully considered, and appropriate investigations should be undertaken.

There can be considerable variation in practice in the management of ex-premature infants at risk of postoperative apnoea, and there is no clear consensus about the timing of discharge and the nature and duration of postoperative monitoring.12, 16, 17, 18, 19, 20 Consideration should always be given to individual patient factors, the risks associated with delayed hernia repair, local institutional policies, and availability of resources for extended cardiorespiratory monitoring in the postoperative period.

Evidence base: postoperative apnoea

The occurrence of postoperative apnoea in the ex-premature infant initially came under scrutiny in the 1980s.¹⁷ One of the core issues explored has been identification of those most at risk with consideration of whether anaesthetic technique could modify that risk. Other aspects considered include the definition of a clinically significant apnoea, the role of overnight admission, the optimal type of monitoring, and the role of intensive care in the postoperative phase.13, 17, 18, 19, 20, 21

Studies of postoperative apnoea have been limited by the heterogeneity of the ex-premature group of patients with any one institution usually caring for small numbers. Sample size in individual studies has remained an issue, despite attempts to offset this by pooling data and recruiting at multiple centres. Different studies have used different definitions of apnoea (15–20 s) and bradycardia (rates of 80–100 min⁻¹). In addition, terms such as postconceptual age (PCA) and postmenstrual age have been problematic and used inconsistently, although some of the work was undertaken before the AAP's publication of guidance on terminology. Many studies have excluded patients with significant co-morbidity.13, 17

One consistent finding has been that the incidence of apnoea varies with the type of monitoring used. There is no standardised set of neonatal postoperative monitors. The most sensitive techniques for detection of apnoea, such as those incorporating continuous polysomnographic-type electronic recording (e.g. impedance pneumography, end-tidal carbon dioxide, and nasal thermistry) are unlikely to be practicable or available. Neonatal respiratory monitors that use an abdominal sensor to detect abdominal movement, or under mattress movement sensors are sometimes used but may not be available in all institutions. Oxygen saturation, ECG, and nursing observation are usually more readily available. Where duration of cardiorespiratory monitoring has been specified in the literature, it has been for a minimum of 6 or 12 apnoea-free hours in accordance with estimated risk.13, 17, 18, 19

One of the most significant studies was undertaken by Cote and colleagues using pooled data from several studies.¹⁷ Logistic regression was used to generate models examining the risk of apnoea at different gestational and postconceptual ages (this study precedes the AAP policy statement about terminology and uses the term PCA, and postmenstrual age is estimated as PCA plus 2 weeks). Multivariate models were also used to determine whether the risk of apnoea was influenced by other potential risk factors (i.e. birth weight, history of respiratory distress syndrome, bronchopulmonary dysplasia, neonatal apnoea, necrotising enterocolitis, ongoing apnoea, anaemia, and use of opioids or non-depolarising neuromuscular blockers). Not all patients had data available for each potential risk factor.

Pooled data analysis found significant variability in the incidence of postoperative apnoea between institutions (ranging between 5% and 49%). The incidence of apnoea doubled when using continuous monitoring devices as opposed to bedside monitors and observation. The incidence of apnoea was found to be strongly and inversely related to PCA and gestational age. Anaemia was found to be the only independent risk factor, and a relationship with the other variables considered was not found. When intervention with bag mask ventilation was required, it usually—but not exclusively—occurred in PACU and not in the ward. The concluding recommendations were guarded and suggested that ‘given the limitations of this combined analysis each physician and each institution must decide what is an acceptable risk for apnoea’. After publication of this work, the thresholds adopted for overnight admission were usually between 54 and 60 weeks PCA. The thresholds were derived from the upper limits of the 95% confidence intervals (CIs) at which the probability of apnoea decreased below 1%. In this model, infants with either anaemia or those who experienced apnoea in the PACU were excluded:

• (i)At 54 weeks PCA, the probability of apnoea (upper limit of 95% CI) decreased below 1% in infants born at 35 weeks gestational age.
• (ii)At 56 weeks PCA, the probability of apnoea (upper limit of 95% CI) decreased below 1% in infants born at 32 weeks gestational age.
• (iii)At 60 weeks PCA, the probability of apnoea (upper limit of 95% CI) decreased below 1% in infants born at 28 weeks gestational age.

Extrapolating from logistic regression models that included anaemic infants or infants who experienced apnoea in the PACU, the probability of apnoea (upper limit of 95% CI) did not decrease below 1% in infants who were born at 28, 32, or 35 weeks gestational age. The analysis was not projected beyond 60 weeks PCA.

A Cochrane review has examined the role of prophylactic methylxanthine (caffeine) to prevent postoperative apnoea after general anaesthesia in preterm infants. The review was based on a small number of subjects from three trials. The review found some evidence that caffeine given at the time of surgery reduces apnoea, bradycardia, and cyanosis after anaesthesia but was cautious about recommendations for a change in practice.²²

As described above, the more recent GAS study examined the influence of general anaesthesia on neurodevelopment as the primary outcome measure; a secondary aim was to compare the incidence of apnoea in patients undergoing inguinal hernia repair.¹³ The incidence in ex-premature infants (awake regional and general anaesthesia) was 6.1% and 0.3% in term infants. Less early apnoea (<30 min) occurred with regional anaesthesia, and similar rates of late apnoea occurred after both regional and general anaesthesia. Occasional life-threatening apnoea occurred in PACU after both regional and general anaesthesia. Overall, 9 of 642 patients required positive pressure ventilation or cardiopulmonary resuscitation to treat apnoea. Two infants from that group of nine experienced multiple apnoeas, which started at 6–7 h after surgery; both of these infants had received regional anaesthesia. This underscores the importance of close and extended cardiorespiratory monitoring. Notably, although early apnoea was found to be a strong predictor of late apnoea, it is not a sensitive measure and late apnoea may occur in the absence of an early episode.

The study found that 19% of patients in the regional anaesthesia group needed conversion to general anaesthesia. No association with anaemia was found. Associations identified were prematurity (the strongest association), lower gestational age at birth, lower weight, lower postmenstrual age, a history of recent apnoea, a history of methylxanthine use, a history of ever being ventilated with a tracheal tube, and a history of ever needing oxygen support. Risk factors identified for both early and late apnoea were similar.

Ideally, the evidence base would provide a robust predictive model of the need for overnight admission in the ex-premature infant undergoing minor surgery, and also support more selective use of ICU resources. As this is not the case, many centres continue a policy of overnight admission with extended cardiorespiratory monitoring for ex-premature infants undergoing minor surgery such as inguinal hernia repair until a postmenstrual age of 56–60 weeks. The risk of apnoea in term neonates is not clear as there are limited data, but term infants are usually considered at risk below the age of 44 weeks postmenstrual age.

Declaration of interest

None declared.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Biographies

Nargis Ahmad MA (Cantab) FRCA is a consultant at Great Ormond Street Hospital and an honorary Senior Lecturer at University College London Institute of Child Health. She is a member of the Association of Paediatric Anaesthetists of Great Britain and Ireland (APAGBI) Education & Training Committee and Chair of the London Neonatal Anaesthesia Network.

Sarah Greenaway BSc FRCA PGCME is a clinical research fellow at Great Ormond Street and the Royal London Hospitals. Her research interests include transfusion and coagulopathy in neonates. She is the new project coordinator on the APAGBI Paediatric Anaesthesia Trainee Research Network.

Matrix codes: 1A01, 1A02, 2D02, 3D00