The multidisciplinary panel comprised ten specialist in Anaesthesia and/or Intensive Care Medicine and five plastic surgeons with special interest in the management of major burn patients. Panellists were located across three continents in six countries: the UK, the Republic of Ireland, the USA, Canada, Switzerland and Australia. All 15 panellists completed both rounds. Demographic characteristics and clinical expertise of the panel are provided in Table 1.

Table 1 Demographic characteristics and clinical expertise of the RAM expert panel members (n = 15)

The final survey included 140 statements on BII diagnosis and management: 74 were rated appropriate, 40 uncertain, and 26 inappropriate. Agreement was reached for all scenarios except six. Disagreement arose regarding the use of FOB to guide fluid requirements, tracheal tube size for initial intubation, airway pressure release ventilation for ARDS, and the use of nebulised heparin, N-acetylcysteine, and salbutamol in mild BII.

History and physical examination findings that lead to the suspicion of BII

In patients with a history of exposure to fire and smoke, indicators including exposure to fire within a closed space, prolonged exposure, loss of consciousness, cardiopulmonary resuscitation and known fatalities in the same incident were rated as appropriate for suspecting a diagnosis of BII. The presence of accelerants at the scene was rated uncertain as an indicator. Physical examination findings of supraglottic injury including facial and neck burns, singed facial and nasal hair, oedema, erythema and blistering of the oral cavity and oropharynx, and stridor, were rated appropriate. Indicators of subglottic injury including coughing, wheezing, hoarseness, dyspnoea, carbonaceous sputum, increased secretions, the use of accessory respiratory muscles and altered consciousness were rated appropriate (Fig. 2).

Fig. 2

Appropriateness of using history and examination findings as indicators of potential burn inhalation injury. For each statement, median scores were calculated. Statements with a median score of ≤ 3 being classed inappropriate (red background), > 3 and < 7 uncertain (amber background) and ≥ 7 appropriate (green background). Disagreement was not present for any statements. Panellists n = 15

Investigations

To aid diagnosis of BII in acute and subacute settings, appropriate measures included arterial blood gas measurement (including lactate), carboxyhaemoglobin measurement, conventional and video laryngoscopy, fiberoptic nasendoscopy and FOB. Radionuclide imaging with 133xenon, magnetic resonance imaging (MRI) or pulmonary function tests were rated inappropriate. Point-of-care lung ultrasound, chest radiograph and chest computed tomography were rated uncertain. Of interest, chest radiograph was rated uncertain in the acute setting, which may be a non-significant statistical anomaly, however it represents a standard of practice for any hospital admission patient with major injuries.

To assess BII severity and prognosis in the acute or subacute setting, appropriate measures included arterial blood gas measurement, carboxyhaemoglobin measurement, chest radiograph, chest computed tomography, video laryngoscopy and FOB. Radionuclide imaging with 133xenon was rated inappropriate. This remains an experimental tool at present, hindered by the practicalities of clinical availability and access. Prognostication investigations after the acute diagnostic process, such as point-of-care lung ultrasound, fiberoptic nasendoscopy, conventional laryngoscopy, magnetic resonance imaging or pulmonary functions tests were rated uncertain (Fig. 3).

Fig. 3

Appropriateness of investigations in the diagnosis and management of burn inhalation injury. For each statement, median scores were calculated. Statements with a median score of ≤ 3 being classed inappropriate (red background), > 3 and < 7 uncertain (amber background) and ≥ 7 appropriate (green background). Burn inhalation injury severity was defined according to Abbreviated Injury Score criteria as mild (grade 1), moderate (grade 2) and severe (grades 3–4). Disagreement was present for one statement (DI ≥ 1). BII, burn inhalation injury. Panellists n = 15. *Fiberoptic bronchoscopy, if intubated. ‡Denotes disagreement (DI ≥ 1)

FOB in the management of BII

FOB was rated appropriate in assessing mortality risk when used alongside AIS grading. Its use as an adjunct to guide fluid requirements generated disagreement and was rated uncertain (DI = 1.52). Uncertainty remained regarding its ability to predict duration of mechanical ventilation. FOB-initiated therapeutic lavage was rated uncertain for mild BII but appropriate in moderate and severe BII. Serial therapeutic lavage was rated inappropriate for mild BII, uncertain for moderate BII, and appropriate for severe BII. Serial visual assessment of the airways by FOB was uncertain for mild and moderate BII but appropriate for severe BII. The use of FOB to deliver therapeutic agents was rated uncertain across all BII severities (Fig. 3).

Airway management

Initial intubation with a standard cuffed endotracheal tube of internal diameter ≥ 8.0 mm was rated appropriate for BII patients at risk of airway compromise. Intubation with tracheal tubes < 8.0 mm generated disagreement and was rated uncertain (DI = 1.19). Repeated extubation and re-intubation between theatre trips for BII patients requiring surgery was rated inappropriate. Instead, early tracheostomy (within seven days of intubation) or late tracheostomy (after eight days or more of intubation) were rated appropriate (Fig. 4).

Fig. 4

Appropriateness of airway management strategies for burn inhalation injury. For each statement, median scores were calculated. Statements with median score of < 3.5 were classed as inappropriate (red background), ≥ 3.5 and < 6.5 as uncertain (amber background) and ≥ 6.5 as appropriate (green background). Disagreement was present for one statement (DI ≥ 1). BII = burn inhalation injury. Panellists n = 10. *Exceeding 7 days. ‡Denotes disagreement (DI ≥ 1). §In-between theatre visits

Systemic toxicity in BII

For treatment of carbon monoxide intoxication, high fractional inspired oxygen therapy was rated appropriate. Patient transfer for hyperbaric oxygen therapy outside of a burns service was rated inappropriate. The theoretical availability of on-site hyperbaric oxygen therapy at a burns service was rated uncertain. Indicators such as exposure to fire and smoke within a closed space, raised serum lactate (≥ 8 mmol/L), unexplained cardiac dysfunction, altered consciousness, seizures, cardiac arrest and respiratory arrest were rated appropriate for suspecting hydrogen cyanide intoxication. Syncope as an indicator was rated uncertain. For treatment of hydrogen cyanide intoxication, high fractional inspired oxygen therapy and hydroxocobalamin were rated appropriate. Sodium thiosulphate, dicobalt edetate and amyl and sodium nitrite were rated uncertain. In the absence of laboratory confirmation of hydrogen cyanide intoxication, administering hydroxocobalamin promptly was rated appropriate for patients with hyperlactataemia ≥ 8 mmol/L, but inappropriate for those with normal lactate levels (< 2 mmol/L). It was rated inappropriate to delay administration of hydroxocobalamin to await laboratory confirmation of hydrogen cyanide toxicity for patients with hyperlactataemia. It remained uncertain whether hydroxocobalamin should be administered promptly for patients with only mild hyperlactataemia, or if administration should be delayed until laboratory confirmation or higher clinical suspicion of intoxication (Fig. 5).

Fig. 5

Appropriateness of diagnostic and management strategies for burn inhalation injury associated systemic toxicity. For each statement, median scores were calculated. Statements with a median score of ≤ 3 being classed inappropriate (red background), > 3 and < 7 uncertain (amber background) and ≥ 7 appropriate (green background). Clinical suspicion of hydrogen cyanide toxicity was defined as low (normal blood lactate and the absence of potentially suspicious features), moderate (moderate lactatemia below 8 mmol/L and few potentially suspicious features) and high (hyperlactataemia ≥ 8 mmol/L and potentially suspicious features including anion gap lactic metabolic acidosis, altered consciousness, unexplained cardiac dysfunction). High serum lactate was defined as ≥ 8 mmol/L. Disagreement was not present for any statements. BII, burn inhalation injury. Panellists n = 15. *Until a higher clinical suspicion or laboratory confirmation is available

Invasive mechanical ventilation in BII

For BII patients requiring mechanical ventilation both with and without ARDS, lung protective strategies (tidal volume < 6 mL/kg ideal body weight, plateau pressure < 30 cmH20) were rated appropriate. Conventional non-lung protective ventilatory strategies, high-frequency oscillatory ventilation (HFOV) were rated inappropriate as was high-frequency percussive ventilation (HFPV), a ventilatory strategy used for effecting better airway clearance. Airway pressure release ventilation (APRV) was rated appropriate for BII patients with ARDS but uncertain for those without ARDS. Prone positioning, recruitment manoeuvres, inhaled prostacyclin analogues, inhaled nitric oxide, neuromuscular blocking agents and referral for venovenous extracorporeal membrane oxygenation (vvECMO) were rated appropriate for BII patients with refractory hypoxaemia. Referral for vvECMO was rated appropriate, even if it required transferring the patient from a burns service without vvECMO capabilities to an alternative ICU. The practicalities of separating the burns surgical and nursing teams from the ECMO ICU team are a challenge to optimal patient care, if not co-located (Fig. 6).

Fig. 6

Appropriateness of ventilation strategies for burn inhalation injury and/or acute respiratory distress syndrome. For each statement, median scores were calculated. Statements with median score of < 3.5 were classed as inappropriate (red background), ≥ 3.5 and < 6.5 as uncertain (amber background) and ≥ 6.5 as appropriate (green background). Lung protective ventilatory strategies were defined as tidal volume < 6 mL/kg ideal body weight, plateau pressure < 30 cmH20. Disagreement was present for one statement (DI ≥ 1). BII, burn inhalation injury. ARDS, acute respiratory distress syndrome. Panellists n = 10. ‡Denotes disagreement (DI ≥ 1)

Pharmacotherapy in BII

5000 IU nebulised heparin was rated uncertain for patients with mild BII (DI = 1.52), but appropriate for moderate and severe BII. 10,000 IU nebulised heparin was rated inappropriate for mild BII and uncertain for moderate and severe BII. Nebulised N-acetylcysteine was rated uncertain for mild BII (DI = 1.70), appropriate for moderate BII and uncertain for severe BII. Nebulised sodium bicarbonate was rated uncertain regardless of BII severity. Nebulised salbutamol was rated uncertain for mild BII (DI = 1.36) and appropriate for moderate and severe BII. Nebulised racemic epinephrine was rated inappropriate for mild BII and uncertain for moderate and severe BII. Systemic prophylactic antibiotics and corticosteroids were rated inappropriate regardless of BII severity (Fig. 7).

Fig. 7

Appropriateness of pharmacological therapies for varying severities of burn inhalation injury. For each statement, median scores were calculated. Statements with a median score of ≤ 3 being classed inappropriate (red background), > 3 and < 7 uncertain (amber background) and ≥ 7 appropriate (green background). Burn inhalation injury severity was defined according to Abbreviated Injury Score criteria as mild (grade 1), moderate (grade 2) and severe (grades 3–4). Disagreement was present for three statements (DI ≥ 1). BII, burn inhalation injury. IU, international units. Panellists n = 15. ‡Denotes disagreement (DI ≥ 1)

A complete table of the second-round questionnaire, median appropriateness ratings and disagreement index values for each statement are available in Additional file 4: Table S3. Detailed figures illustrating median scores and variation of the statement judgments are available as Additional file 5: Fig. S2, Additional file 6: Fig. S3, Additional file 7: Fig. S4, Additional file 8: Fig. S5, Additional file 9: Fig. S6, Additional file 10: Fig. S7. Key results and panel recommendations are summarised in Fig. 8.

Fig. 8

Summary of expert panel recommendations for burn inhalation injury. Parts of the figure were drawn by using pictures from Servier Medical Art. Servier Medical Art by Servier is licensed under a Creative Commons Attribution 3.0 Unported License (https://creativecommons.org/licenses/by/3.0/)