Changing perspectives in the prehospital management of patients with severe burns

Prophylactic endotracheal intubation

Controversy has emerged recently over whether PETI is appropriate in the initial emergency management of suspected IBI. Compounding this, it still appears that no appropriate, evidence-based guidelines are available (Costa Santos et al, 2015). The most recent relevant consensus statement released by the Faculty of Pre-hospital Care features no advice (Allison and Porter, 2004), and the latest guidance from the Joint Royal Colleges Ambulance Liaison Committee (JRCALC) (2019) offers only basic advice. The same is true of the British Burns Association (BBA) (2018), the National Institute for Health and Care Excellence (NICE) (2018) and, locally, the South Western Ambulance Service NHS Foundation Trust (2018).

IBI has been found to significantly increase mortality in burns patients (Chen et al, 2012; 2014). Oedema occurs when the upper airway is directly exposed to superheated gases and toxic fumes (Toussaint and Singer, 2014). Therefore, inaction in this situation can be catastrophic as the airway may swell and be lost (Ahmed et al, 2014). Because of this, the safest approach has traditionally been early PETI (Toussaint and Singer, 2014). This follows historical research that evidences lower mortality rates in patients with IBI in whom PETI is undertaken (Phillips and Cope, 1962; Bartlett et al, 1976; Venus et al, 1981). These studies are, however, dated and use unreliable methodology. Consequently, no concrete inferences should be made.

On the contrary, two recent literature reviews, which appear to show a shift in attitudes, argue that performing PETI on all patients with suspected IBI is too simplistic an approach (Latenser, 2009; Walker et al, 2015). Mosier and Pham (2009) explain that endotracheal intubation (ETI) and mechanical ventilation (MV) bypass many of the respiratory tract's defences against infection. Evidence suggests that this action increases a patient's risk of mortality as they are at greater danger of developing ventilator-associated lung injury or ventilator-associated pneumonia (Soni and Williams, 2008; Villar and Slutsky, 2010; Hunter, 2012).

Despite this, Mackie et al (2011) found that the number of burns patients who received MV in a specialist burns centre (SBC) increased significantly between 1987 and 2006. This was despite there being no notable change in patient and burn characteristics over the same period. This research retrospectively reviewed the medical records of 258 patients with burns >30% TBSA. Although not generalisable to the overall patient population, these data provide an insight into the increasing application of PETI despite a clear understanding of its risks in the medical community. The researchers hypothesise that this may be because new ATLS and ABLS teaching was introduced mid review, which emphasised a low threshold for ETI in patients with facial burns (Mackie et al, 2011).

Numerous recent, single-centre studies with a similar retrospective review design, in various localities, have found that a large proportion of patients with suspected IBI admitted to SBCs are unnecessarily intubated. This has also been attributed to previous ATLS and ABLS airway management teaching. Most notably, Eastman et al (2010) reviewed patients admitted to a single centre between 1982 and 2005 in Texas. A large sample size was achieved over this period and the researchers found that 41.4% of 1272 intubated burns patients admitted were extubated after less than 2 days. The researchers conclude that such a short period of ETI, combined with no requirement for reintubation, suggests that ETI was inappropriate (Eastman et al, 2010).

Comparably, Romanski et al (2016) found that 40.1% of the 416 patients included in their review were extubated within 48 hours of admission, with none being reintubated. Although Orozco-Peláeza (2018) highlights that the researchers had not specified which patients were extubated early as their levels of consciousness had improved, the original indication for their ETI may represent bias. A much smaller study by Costa Santos et al (2012) found similar rates of extubation within 48 hours. Importantly, the researchers concluded that facial burns were a poor predictor of IBI.

Using a different methodology, Amani et al (2012) studied a small number of patients who had suspected IBI from fires caused by ignition of home oxygen therapy. Only 12% of the 32 patients who were intubated before transport to an SBC had an IBI later confirmed by bronchoscopy. There was no statistical difference in TBSA between intubated and non-intubated patients.

The research that supports the requirement for a higher threshold for PETI is observational in design, single-centre and, in some cases, uses small samples. Including more participants would improve reliability and external validity (Biau et al, 2008). The chosen methodology, although relatively simple to undertake, presents a risk of selection bias (Eastman et al, 2010). Because of these limitations, it could be argued that no concrete causal link between the variables studied can be concluded (Boyko, 2013).

Randomised controlled trials (RCTs) may be the only way to provide truly robust evidence (Sackett et al, 1996). However, future research of this type may not be possible, as randomly allocating patients to a treatment already demonstrated to be inferior is unethical (Ashcroft and ter Meulen, 2004). However, because of the breadth of evidence, its varied settings and a significant absence of current evidence contradicting it, a change in practice seems warranted and safe.

Despite the need for a higher threshold, there will still be patients with suspected IBI who require PETI. In the absence of clear guidelines, practitioners are left to make this decision based on clinical knowledge and experience (Ahmed et al, 2014). The International Society for Burn Injuries (ISBI) (2016) has emphasised the need for a screening tool that is both highly sensitive and specific. Eastman et al (2010) argue that signs such as facial burns, singed nasal hairs and soot in the naso/oropharynx, previously regarded as typical of IBI, are neither. Burns surgeons have been questioning the validity of these signs for some time now (Madnani et al, 2006; Mlcack et al, 2007). Following their research, Romanski et al (2016) developed guidance advocating that intervention should be considered only when shortness of breath, wheezing, stridor, hoarseness or a decreased level of consciousness are present. ISBI (2016) concurs but admits that the evidence base for this list is weak.

Airway intervention should be carried out only when there are signs of impending airway or respiratory compromise to ensure high specificity. Clinicians must recognise, however, that no history or patient presentation will constitute a definitive diagnosis (Ikonomidis, 2012; Dries and Endorf, 2013; Oscier, 2014). Furthermore, the absence of these signs should not serve as reassurance that IBI is not present (Cancio, 2009).

Cooling

Although there is a lack of guidance on the management of IBI, there appears to be a consensus over the use of prehospital cooling of burn injuries. The use of cool tap water has been demonstrated to be more effective at reducing skin surface temperature than Burnshield and Burnaid dressings (Cho and Choi, 2016). Consequently, the BBA (2018), JRCALC (2019) and NICE (2017) advocate that burns should be treated immediately with continuous running cool tap water over the injured area for 20 minutes (Cuttle et al, 2010). Cooling attempts may induce hypothermia, however—especially in children, the elderly and adult patients who have sustained severe burns. In adults with severe burns, damage to large parts of the integumentary structure can lead to thermoregulatory compromise and thus an inability to maintain core temperature (Bourke and Dunne, 2013; Hostler et al, 2013).

Hostler et al (2013) conducted a retrospective cohort study to investigate the prevalence and effect of hypothermia on mortality among patients with burns treated at five SBCs in Pennsylvania between 2000 and 2011. The study found that 40% of the 12 097 patients reviewed were hypothermic upon admission and that hypothermia is independently associated with mortality. The study could not, however, establish the cause of the hypothermia. Although the study was not experimental in design, the large sample size in this observational research meant that the effects of confounding variables could be evaluated using statistical tests. The same was not true for a similar yet smaller study from a single SBC in New York state by Singer et al (2010), which made comparable findings. Unlike Hostler et al (2013), though, this research concluded that prehospital cooling was not the cause of the hypothermia.

Conversely, Singer et al (2010) found that temporary, mild hypothermia increased survivability in a small number of experimental rats that had received 40% TBSA scalds. Although the research used random allocation and a control group, animal studies are regarded as poor predictors of human reactions (Bracken, 2009). Research by Weaver et al (2014) found that the risk factors for hypothermia in burns patients are older age, burns of 20% TBSA or more, and exposure to the elements. The researchers argue that these factors are easily identifiable and manageable by prehospital clinicians. These data may not be generalisable as the inclusion criteria covered only patients who had been transported by emergency medical services directly to an SBC. Furthermore, despite the data being prospectively collected, the research was, again, observational.

The prevalence and risks of hypothermia in patients with burns are clear. Consequently, guidance has long advocated the ‘cool the burn, warm the patient’ mantra (Allison and Porter, 2004). However, patients may have such severe burns that first aid cooling needs to cover large areas of the body, which makes warming the patient impossible. When deciding whether warming the patient or cooling the burn should take priority, it is important to explore the evidence that informs cooling. Cooling a burn injury reduces the production of inflammatory mediators and promotes the maintenance of viability in the zone of stasis, potentially preventing the progression of tissue damage (Bourke and Dunne, 2013). Although a recent RCT has shown the use of cooling agents to be effective at lowering the skin temperature (Cho and Choi, 2017), there is little human evidence suggesting this treatment has any consequences of burn injury or influence on patient outcome (Baldwin et al, 2012).

Cuttle et al (2010) explain that a huge amount of controversy remains over the efficacy of first aid treatments for analgesia and minimising burn injuries as this is informed primarily by anecdotal evidence and animal studies (Rose, 1936; Jackson, 1953; Brown et al, 1956; Martin et al, 1958; Shulman, 1960; King et al, 1962; Ofeigsson, 1965; Poy et al, 1965; Zimmerman and Krizek, 1984; Venter et al, 2007; Yuan et al, 2007; Bartlett et al, 2008; Rajan et al, 2009).

There is an increasing awareness of the unreliability of animal studies and their potential to cause harm when used to predict clinical outcomes in humans (Beauchamp and DeGrazia, 2015). Despite this, Wright et al (2015) argue that the sheer abundance of animal study evidence justifies cooling. Furthermore, the evidence from these animal studies should not be overlooked as conducting similar research on human participants may not be possible for ethical reasons. In contrast, Muehlberger et al (2010) suggest that, because of the lack of robust human research, cooling in the emergency setting has only a negligible effect on the prognosis of burns patients and they suggest that the swift administration of analgesics should take priority instead.

In recent years, Wood et al (2016) conducted unprecedented research to investigate the relationship between water cooling first aid on short-term, post-burn outcomes in humans. This retrospective cohort study reviewed burns patient records over 3 years. The study found that water cooling did reduce the risk of intensive care unit admission and the need for wound repair surgery. A reduction in mortality, however, was not demonstrated. The multicentre design of this observational study enabled a greater degree of external validity. Potential bias from different local practices was evaluated using statistical tests made possible by the large sample size of 2320. The methodology employed constitutes a significant limitation, however, so the results must be interpreted within this context.

The author feels that it would be prudent to put a greater emphasis on hypothermia prevention rather than burn injury cooling as there is more compelling evidence suggesting that the former has a far greater influence on mortality than the latter. This may, however, lead to a longer time in hospital and a greater requirement for surgical intervention. If cooling does not occur immediately, the BBA (2018) advises that it can still be effective up to 3 hours post injury. This recommendation is, however, based on findings from porcine studies (Bartlett et al, 2008; Cuttle et al, 2010).