Les patients grands brûlés sont très vulnérables à la pneumonie et en particulier à la pneumonie sous ventilation assistée (PVA). La PVA reste une cause importante de morbidité et de mortalité, en dépit de l’amélioration des soins intensifs et de la chirurgie dans ces derniers temps. La longue durée de la ventilation, le type et l’extension des brûlures (les brûlures par inhalation en particulier) sont liés à la fréquence de la PVA. Autres facteurs de risque (nombre d’interventions, de ré-intubations et chirurgicales) sont également importants. Des stratégies efficaces de prévention doivent être respectées, et les protocoles doivent être mis en oeuvre pour faciliter le diagnostic et le traitement des la PVA. Les critères cliniques et radiologiques et la lavage bronchoalvéolaire doivent être utilisés pour déterminer le germe en question, et il faut avoir un seuil bas pour l’initiation précoce de la thérapie empirique, basée sur les profils de résistance en vigueur dans l’unité de brûlures. Les brûlures graves devraient être traitées dans des centres ayant accès à des ressources et à des connaissances multidisciplinaires.

A number of advances have been made in recent times with regard to fluid resuscitation protocols, dressings, infection control strategies and antimicrobials, surgical techniques, intensive care, and nutrition. There is now widespread recognition that specialist burns units or centres deliver the best care for these patients. As a result of these measures, mortality and morbidity rates have declined significantly over the last few decades.^3,4

Significant thermal injuries induce a state of immunosuppression; three-quarters of all severe burn related deaths are consequence of infection, most notably burn wound infections, sepsis, pneumonia and urinary tract infections, many of which are nosocomial.^4-8 There are a number of mechanisms for the development of pneumonia in the severely burnt. Pulmonary complications are common with inhalational injury, but burnt patients have more pulmonary complications even without direct lung injury. Atelectasis and hypostatic pneumonia are common due to altered ventilation and reduced lung expansion that may occur in patients with chest or abdominal burns. These patients may also have a high risk of aspirating, and respiratory physiotherapy with regular airway suctioning of upper airway secretions and expectoration of sputum may be critical to maintaining pulmonary function.^5,9

Patients who require prolonged ventilation are also at risk of developing ventilator-associated pneumonia (VAP). Prior to 2007, VAP could only be diagnosed after 48 h of mechanical ventilation. No minimum time of ventilation is now required to make the diagnosis.¹⁰ Despite this change in definition, there may still be some utility in considering patients in early or late groups, because causative organisms and their resistance patterns vary in relation to this.

In the paediatric intensive care setting, VAP is responsible for significant morbidity and mortality, and ranks as the second commonest hospital acquired infection. In fact, a large European trial in a variety of paediatric settings showed that VAP accounted for over half of hospital acquired infections in the PICU.¹¹ The prevalence of nosocomial pneumonia in the ICU ranges from 10 to 65% and mortality rates exceed 25%. Those who develop VAP are twice as likely to die compared to those without VAP, and spend longer in intensive care. The nosocomial bacteria that cause VAP tend to be more resistant.^11-15

Strategies to prevent VAP in burns patients have largely been adapted from studies focusing on other patient groups. Many of these have been incorporated into socalled ‘care bundles’ for ventilated patients. Table II lists some of the preventative strategies.

Daily interruptions and weaning protocols are now fundamental tenets of good adult ICU care. The role in children is less well established, but the principle of reducing ventilation duration to shortest possible is obviously paramount. ^9,11,1616 Adaptations to the endotracheal tube have been examined in an attempt to reduce the incidence of VAP. The NASCENT trial, for instance, showed a 35.9% risk reduction when a silver impregnated endotracheal tube was used.¹⁷ Techniques to continuously aspirate subglottic secretions have been shown to be effective and should be implemented.^17,18 At present intermittent open suctioning techniques are the norm in ICUs in developing countries.

Stress ulcer prophylaxis and early enteral nutrition are now well-established in both burns care and in the ICU scenario. However, they may also be responsible for raising the pH and gastric bacterial colonization, which predisposes to VAP. Gastritis is rare in the well-resuscitated child burn victim and many units may choose not to make use of stress ulcer prophylaxis. Sucrulphate may be a sensible alternative to other agents that raise gastric pH. In addition, post-pyloric feeding rather than gastric feeding, may reduce the incidence of VAP.^9,19 Placing the patient in the semi-recumbent position is effective and easily applied in the adult patient, but may prove difficult in the paediatric population, where there may be marked size differences between patients.⁹

Meta-analyses have shown a significant reduction in the incidence of pneumonia with the use of selective decontamination of the GIT.^20,21 Prophylactic antibiotics may also be justified in certain circumstances, but applied with an awareness of the prevailing resistance patterns.²² Chlorhexidine mouth-wash is a cost-effective and easily applied strategy for reducing VAP.^23,24 Contact precautions and hand hygiene are now well-established tenets of critical care, and adherence to these and other elements in the ventilator care bundles need to be regularly reviewed.²⁵

Burns patients frequently require blood transfusions during their course of treatment. A restrictive policy should be adhered to, as each unit of blood received is associated with a 13% increased risk of infection, presumably by exacerbating the state of immunosuppression.^26,27 Intense insulin therapy and tight glucose control, despite initial optimism, need to be re-evaluated in a major trial, and particularly in major paediatric burns. The consequences of hypoglycaemia are potentially catastrophic, and more significant than the potentiation of infection during periods of high normoglycaemia, or even hyperglycaemia.^28,29

The traditional 48 h of ventilation required to make the diagnosis has now been removed from the definition of VAP.¹⁰ The ventilator is recognized as a major risk factor for the development of pneumonia in the ICU. The spectrum of bacteria responsible (community acquired vs nosocomial) and the efficacy of preventative and therapeutic measures, make time an important component. VAP is a prominent cause of morbidity and mortality in burns patients, but patients who developed pneumonia prior to 48 h of ventilation were also at high risk, particularly if they had large body surface burn wounds. Of the organisms responsible for VAP, Acinetobacter baumannii infections, in particular, have proved to be a virulent cause of ventilator associated pneumonia in our setting.^6,9,11 Bronchoalveolar lavage, protected specimen brush or non-bronchoscopic lavage are the best means of obtaining reliable quantitative cultures. By combining one of these techniques with clinical features of pulmonary infection, the sensitivity and specificity may be optimized.^6,9,11,34-38

P. aeruginosa remains the most invasive pathogen in our patients, but Acinetobacter baumannii is now also a major problem, not least because of growing resistance. MRSA has not yet proved as significant a problem as experienced in other parts of the world. Because many of our patients are immunocompromised and malnourished prior to burn injury, several of our patients are particularly susceptible to fungal infections.

Many ICUs are still dependent on alternatives to invasive quantitative microbiological techniques like sputum cultures and tracheal aspirates. The results of these tests are relatively unreliable and treatment should be initiated and de-escalated with this is mind. If possible, empiric therapy should be limited to 48 h: this period of use should not increase resistance.^38,39 Antibiotics may safely be stopped once clinical features of infection have resolved. There is probably no difference in outcomes if an eight or fifteen day course of treatment is implemented, unless the organism responsible is P. aeruginosa or Acinetobacter, where a longer course is necessary to reduce recurrence.³⁹ Table IV lists some of the principles of antibiotic use for VAP.