Some newborn infants fail to adapt from fetal to extrauterine life and require immediate cardiopulmonary and cerebral resuscitation. The Apgar scoring system (Table 96.1) scored after 1 minute, remains the most widely accepted method of assessment. The best Apgar score is 10 and the worst is 0. There is an inverse relationship between the Apgar score and the degree of hypoxia and acidosis. It has been suggested that the 5-minute score is a guide to ultimate prognosis but this is questioned. Collection of sequential scores must not delay the institution of resuscitation.

Table 96.1 Apgar scoring system

BIRTH ASPHYXIA

The causes of birth asphyxia may be:

• placental failure – acute or chronic (e.g. toxaemia, diabetes and antepartum haemorrhage)

• drug depression due to maternal analgesics or sedatives administered immediately prior to delivery

• obstetric complications (e.g. difficult forceps, breech, caesarean section and prolapsed cord)

• fetal conditions (e.g. multiple births and prematurity)

• postnatal problems (e.g. respiratory distress immediately after birth from any cause)

MANAGEMENT

The principles of resuscitation are identical to those employed in other situations. Resuscitation must be started immediately after delivery. Although some cerebral insult may have occurred in utero or intrapartum, secondary insults from postpartum asphyxia must be avoided.

Babies suffering mild asphyxia immediately before birth with Apgar scores between 5 and 7 usually respond to stimulation and gentle suction to the nose, mouth and pharynx although oxygen therapy is occasionally required. Babies with moderate asphyxia (Apgar 3–4) usually respond to bag-and-mask ventilation with oxygen. Acid–base status should be determined and sodium bicarbonate administered to restore pH > 7.25. Severely asphyxiated infants (Apgar 0–2) need urgent cardiopulmonary resuscitation (CPR). After airway suction, bag-and-mask ventilation should be followed by rapid orotracheal intubation and positive-pressure ventilation with oxygen. Fear of complications of oxygen therapy must not mitigate against the administration of 100% oxygen at this stage. If the liquor contains thick meconium, it is vital that the pharynx and trachea be suctioned prior to the onset of respiration or application of positive pressure. Meconium aspiration syndrome is a preventable condition that is often difficult to manage.

Venous access via umbilical or peripheral veins must be established immediately and followed by the administration of 1–2 mmol/kg of sodium bicarbonate. Subsequent buffer therapy should be based on arterial acid–base status if available. Rapid or excessive infusions of hypertonic solutions (e.g. sodium bicarbonate or hypertonic dextrose) or volume expanders may precipitate intracranial haemorrhage, particularly in premature infants.

Asphyxiated infants are usually volume-depleted at birth and blood pressure should be immediately restored with colloids (10 ml/kg in the first instance). External cardiac massage and additional drug therapy (see below) should be employed, if necessary, to restore cardiac rhythm. Postresuscitative care to maintain cerebral perfusion pressure, correct abnormal serum biochemistry and control seizures is required to prevent secondary cerebral insults. Myocardial dysfunction may occur secondary to asphyxia. Dopamine or dobutamine (5–10 μg/kg per min) may prove useful.

The orotracheal tube should be changed to nasotracheal to enable secure fixation once a degree of stability is attained. Radiological confirmation of tube position should be made as soon as possible. The trachea is very short in the newborn and endobronchial intubation is a particular risk.

Approximate nose to mid-trachea (T2) distances are:

• 28 weeks’ gestation: 7 cm

• 33 weeks’ gestation: 9 cm

• term infant: 10.5 cm

CARDIORESPIRATORY ARREST IN CHILDREN

The vast majority of children lack intrinsic myocardial disease and when cardiac arrest occurs, it is usually the end result of hypoxaemia and acidosis. The most common predisposing causes are rapidly progressive upper-airway obstruction, near drowning, sudden infant death syndrome, pneumonia, sepsis, gastroenteritis and major trauma. Such children invariably arrest in asystole and this should be assumed if an electrocardiogram (ECG) is not immediately available.

Ventricular fibrillation (VF) may be anticipated in the following situations:

• congenital heart disease

• cardiomyopathies

• myocarditis

• poisoning (e.g. tricyclic antidepressant ingestion)

• hereditary prolongation of QT interval (Romano–Ward syndrome)

The International Liaison Committee on Resuscitation (ILCOR) has developed a consensus statement detailing the science underpinning paediatric and neonatal basic and advanced life support.² The management of paediatric cardiopulmonary resuscitation is discussed in detail elsewhere in this volume.

VASCULAR ACCESS

Venous access may be difficult, particularly in the collapsed, hypovolaemic or hypothermic child. The external jugular vein may be prominent when usual sites are inaccessible. Cannulation of central veins, apart from via the femoral and external jugular veins, is hazardous even in ideal situations and should not be attempted during cardiac arrest in small children and babies. Two may be considered: intraosseous access or endotracheal instillation.

INTRAOSSEOUS ACCESS ³

This involves infusion into bone marrow – a non-collapsible venous system in direct communication with the circulation. In dog experiments, drugs infused in this way reach the central circulation as rapidly as those injected into a central vein. The effect of equal doses of adrenaline (epinephrine) given by intraosseous and i.v. routes is identical in shocked dogs. The technique is simple and quick to learn.

Intraosseous infusion has been successfully used in CPR.⁴

Potential complications include:

• osteomyelitis: conventional venous access should be established after resuscitation

• compartment syndromes requiring limb amputation: these have resulted from needle misplacement in muscle or administration of fluids into muscle through another bone perforation ⁵

Careful placement is essential, and close observation of the limb and early removal once alternative access is achieved are recommended.

ENDOTRACHEAL INSTILLATION

Endotracheal instillation is, at best, a poor substitute for the above methods. Do not use bicarbonate and calcium as they will cause direct lung damage and cannot be administered by this route.

Epinephrine, lidocaine and atropine may be diluted in saline (newborn 1 ml, infants and preschool children 3 ml, older children 5 ml) and delivered into the bronchi through a catheter passed down the endotracheal tube.

ARTERIAL CANNULATION AND PRESSURE MONITORING

Arterial cannulation is routine practice in paediatric intensive care, even in infants weighing < 1 kg. It is indicated in all critically ill infants for continuous blood pressure monitoring and accurate blood gas sampling. In the neonate, difficult ‘stabs’ may lead to significant errors in PaO₂ and PaCO₂. Umbilical arterial catheters are commonly used in newborn infants, although users must be aware of the potential vaso-occlusive complications (e.g. lower limb ischaemia, renal thrombosis, necrotising enterocolitis and rarely, paraplegia).

Peripheral arteries used include radial, ulnar, brachial, femoral, posterior tibial and dorsalis pedis. Radial and ulnar or posterior tibial and dorsalis pedis vessels must never be cannulated sequentially in the same limb. The safety of brachial and femoral cannulation lies in the presence of rich collateral vessels around the elbow and hip joints. Arterial lines are kept patent by continuous flushing with 1–2 ml/hour of heparinised normal saline (5 units heparin/ml flushing solution) or 5% dextrose. Complications include distal ischaemia, infection, retrograde embolisation and haemorrhage. Retrograde embolisation occurring with flushing is a particular risk in the small infant, depending on the length and volume of the vessel and volume and speed of injection. In a 1.5 kg infant, as little as 0.5 ml of fluid injected rapidly into the right radial artery will reach the cerebral circulation. Haemorrhage from accidental disconnection can be significant because of the relatively small blood volume. Meticulous fixation is therefore required.

CENTRAL VENOUS CANNULATION AND PRESSURE MONITORING

All routes of central venous cannulation are applicable to infants and children and must be within the skills of paediatric intensivists. The femoral route is sometimes the safest in emergency situations. Catheters utilising the Seldinger technique have greatly increased successful placement. Ultrasonography is useful to determine the exact location of veins. Multilumen catheters are recommended when infusing multiple drugs and for parental nutrition. Complications including catheter-related sepsis are the same as in adults. This risk can be reduced by adherence to a bundle of measures.⁶ The need for prolonged venous access may warrant regular catheter changes, or surgical implantation of a central venous device (e.g. Infusaport, Broviac or Hickman catheter). Umbilical venous cannulation and passage of the catheter through the sinus venosus to the right atrium is useful in neonatal emergencies.

PULMONARY ARTERY PRESSURE MONITORING

Pulmonary artery (PA) pressure monitoring using flow-directed 4 and 5 FG catheters is feasible even in neonates; however, it is invasive, technically more difficult and has greater risks than in the adult. It is rarely required in the neonate, as the pulmonary circulation and response to therapy can usually be gauged indirectly from the magnitude of right-to-left shunting. Systemic levels of PA pressure are usually found in neonates with severe lung disease.

The major indication for PA pressure monitoring is following surgery for congenital heart disease (e.g. repair of ventricular septal defect with pulmonary hypertension, truncus arteriosus and obstructed total anomalous pulmonary venous drainage). A catheter is inserted directly into the PA or via the right ventricular outflow tract at the time of surgery. It is most useful in guiding weaning from mechanical ventilation.

LEFT ATRIAL PRESSURE MONITORING

Left atrial pressure monitoring is often useful after open heart surgery for congenital heart disease, and is achieved by means of a catheter placed directly in the left atrium during surgery.

CARDIAC OUTPUT MEASUREMENT

Cardiac output determination can be performed in small children by dye dilution, thermodilution or Doppler techniques. Unfortunately, errors have been substantial, and the first two techniques are invasive, intermittent and only repeatable within finite limits. Interpretation of results is made difficult or impossible in the presence of intracardiac shunts or valvular incompetence. The dye curve may, in fact, be useful in demonstrating residual intracardiac shunts. Calculation of derived variables is only as accurate as the flow measurement.

A recently developed system (PiCCO, Pulsion Medical Systems AG, Munich, Germany) provides continuous cardiac monitoring even in small infants. Cardiac output is determined both intermittently by transpulmonary thermodilution ⁷ and continuously through arterial pulse contour analysis.⁸ The system also derives cardiac preload volume, an index of left ventricular contractility, and an estimate of intrathoracic blood volume and extravascular lung water.

TEMPERATURE MONITORING

Temperature monitoring is important in neonatal and paediatric intensive care. Prevention of cold stress requires accurate measurement of core and skin temperature. Core–toe–ambient temperature gradients provide a sensitive, albeit indirect index of cardiac output and peripheral perfusion and are useful in managing fever. Toe temperature normally lies between core temperature (tympanic or oesophageal) and ambient temperature. The toe–core temperature gradient increases with low cardiac output or vasoconstriction from any cause. Toe temperature is a useful guide to the efficacy of vasodilator therapy.

PULSE OXIMETRY

Pulse oximetry provides continuous non-invasive measurement of arteriolar saturation (SaO₂) and provides a rapid indication of hypoxaemia. Accurate information is given:

• when the oxyhaemoglobin dissociation curve is shifted to the left (e.g. fetal haemoglobin and alkalosis) or to the right (e.g. sickle-cell disease and acidosis)

• in the presence of carboxyhaemoglobin (functional saturation is accurate)

• with moderately severe desaturation (e.g. cyanotic heart disease)

• with anaemia (haemoglobin concentrations above 5 g/dl)

• when skin is pigmented

Errors occur with extreme hypoperfusion, excessive movement and rapidly changing ambient light. A range of sensors is now available to monitor children of all ages.

TRANSCUTANEOUS PO₂ and PCO₂ Monitoring

Oxygen and carbon dioxide diffuse through well-perfused skin from the superficial capillary network and can be measured using modified polarographic and glass electrodes respectively. The electrodes are heated to 43–45°C to arterialise the capillary blood and maximise capillary blood flow. Under optimal conditions, there is good correlation between arterial and transcutaneous gas tensions. Hence continuous monitoring of blood gas tensions is possible in a non-invasive way. The PtcO₂–PaO₂ gradient and the output of the heating element have been used as indices of microcirculation. The accuracy of these devices is mainly confined to the neonatal period.

DRUG INFUSIONS

All drugs used in cardiovascular and respiratory support are administered according to body weight; accurate delivery is crucial. Accurate drug infusions require accurate devices, of which syringe pumps are the most useful. Potentially lethal errors in calculating drug dilutions are minimised by the use of dose/dilution/infusion rate guidelines (Table 96.2).⁹

Table 96.2 Calculation of drug infusion dilutions:

1. Select desired drug dosage to be delivered in μg/kg per min

2. Select infusion rate of syringe pump in ml/h (from centre of table)

3. Calculate number of milligrams of drug to be mixed in 50-ml syringe e.g.: 10-kg child, 0.1–2 μg/kg per min, infusion 1–20 ml/h: put 0.3 ml/kg (= 3 mg) in 50 ml

PAIN RELIEF AND SEDATION IN CHILDREN

Management of pain and agitation in children has received inadequate attention and has tended to be underestimated and underrated. Infants and children are often unable or unwilling to complain of pain. In the past, some believed that the neonate could not perceive pain. It is now clear that even neonates possess all the anatomical and neurochemical systems necessary for pain perception and exhibit physiological and behavioural responses to pain.¹⁰ Stress responses associated with pain and agitation may increase morbidity and mortality in critically ill patients. Analgesia can be provided by narcotic infusions, local blocks and regional techniques in children of all ages. Painful procedures in the ICU must always be accompanied by appropriate analgesia. The addition of sedative agents such as benzodiazepines can reduce agitation, minimise harmful stress responses and result in narcotic sparing.

NEONATAL AND PAEDIATRIC EMERGENCY TRANSPORT

Care of critically ill neonates and children necessitates the use of specialised retrieval services linked to neonatal and paediatric ICUs. Retrieval services should offer facilities for specialist consultation in addition to secondary transport. Careful audit of such services is necessary to improve patient outcome.¹¹ The aim of retrieval services is to extend the intensive care facility to peripheral hospitals, allowing stabilisation by experienced personnel prior to rapid transport to a regional centre in the most appropriate vehicle (see Chapter 4). Special considerations such as thermoregulation and oxygen monitoring must be provided for in neonatal transport. Well-designed neonatal emergency transport services have resulted in significant reductions in morbidity and mortality.

OUTCOME OF PAEDIATRIC INTENSIVE CARE

Depending on admission criteria, mortality in paediatric ICUs ranges from 5 to 15%. If patients with pre-existing severe disabilities are excluded, the majority of survivors have a normal or near-normal life expectancy. A number of scoring systems have been developed or modified for paediatric application to predict ICU mortality. These scoring systems allow comparison between different ICUs, internal audits, stratification of patients for research purposes and analysis of cost benefit. The paediatric risk of mortality (PRISM) score ^12,¹³ and the paediatric index of mortality (PIM) score¹⁴ are applicable to a wide range of critically ill infants and children. Although PRISM performs marginally better, PIM is easier to collect and hence less prone to errors in data collection. PIM also has the advantage that it predicts mortality based on admission parameters whereas PRISM is based on the worst variables in the first 24hours. As many paediatric ICU deaths occur in the first 24hours, PRISM is often recording the dying process rather than predicting it. Specialised scores have been developed for specific problems, e.g. the modified injury severity scale (MISS) and paediatric trauma score (PTS) for paediatric trauma, and the modified Glasgow Coma Scale (GCS) for neurological insults. Numerous scoring systems have been developed for meningococcaemia, the best validated being the Glasgow meningococcal septicaemia prognostic score (GMSPS).¹⁵

Compared with adult intensive care, children with equivalent therapeutic intervention scores (TISS) have a lower in-hospital and 1-month mortality.¹⁶ In addition, non-survivors do not consume a disproportionate amount of resources. While multiple organ failure increases mortality, the prognosis is considerably better than for adults.¹⁷ There is evidence that mortality is lower in specialist paediatric ICUs,¹⁸ and that paediatric ICUs with a larger workload have better outcomes than those looking after fewer children.¹⁹ General hospitals should therefore have facilities for urgent resuscitation of children prior to early transport to a specialised paediatric ICU. Unless unavoidable, critically ill children, particularly those requiring mechanical intervention, should not be cared for in an adult ICU for longer than 24 hours. The American Academy of Pediatrics, the Society of Critical Care Medicine, the British Paediatric Association and the Australian National Health and Medical Research Council have all stated that children should receive intensive care in specialist paediatric units.

REFERENCES

1 Kinsella JP, Neish SR, Dunbar ID, et al. Clinical responses to prolonged treatment of persistent pulmonary hypertension of the newborn with low doses of inhaled nitric oxide. J Pediatr. 1993;123:103-108.

2 The International Liaison Committee on Resuscitation. The International Liaison Committee on Resuscitation (ILCOR) Consensus on Science with treatment recommendations for pediatric and neonatal patients: pediatric basic and advanced life support. Pediatrics. 2006;117:955-977.

3 Rosetti VA, Thompson BM, Miller J, et al. Intraosseous infusion: an alternative route of pediatric intravascular access. Ann Emerg Med. 1985;14:885-888.

4 Saccheti AD, Linkenheimer R, Liberman M, et al. Intraosseous drug administration: successful resuscitation from asystole. Pediatr Emerg Care. 1989;5:97-98.

5 Moscati R, Moore GP. Compartment syndrome with resultant amputation following intraosseous infusion. Am J Emerg Med. 1990;8:470-471.

6 Bernholtz SM, Pronovost PJ, Lipsett PA, et al. Eliminating catheter-related bloodstream infection in the intensive care unit. Crit Care Med. 2004;32:2014-2020.

7 McLuckie A, Murdoch IA, Marsh MJ, et al. A comparison of pulmonary and femoral artery thermodilution cardiac indices in paediatric intensive care patients. Acta Paediatr. 1996;85:336-338.

8 Goedje O, Hoeke K, Lichtwarek-Aschoff M, et al. Continuous cardiac output by femoral arterial thermodilution calibrated pulse contour analysis: comparison with pulmonary arterial thermodilution. Crit Care Med. 1999;27:2407-2412.

9 Shann F. Continuous drug infusions in children: a table for simplifying calculations. Crit Care Med. 1983;11:462-463.

10 Anand KJS, Hickey PR. Pain and its effects in the human neonate and fetus. N Engl J Med. 1987;317:1321-1329.

11 Henning R, McNamara V. Difficulties encountered in transport of the critically ill child. Pediatr Emerg Care. 1991;7:133-137.