The results show that patients admitted to the burn center for suspected inhalation injury are at a higher risk of being intubated and mechanically ventilated if they are a current smoker or have quit smoking within the past 6 months, and/or if they have substance use. This information may provide future guidance for managing inhalation injuries. A positive toxicology screen on admission was associated with intubation. We speculate that the results of a positive urine toxicology screen on admission may predict patients at risk for further respiratory compromise. This knowledge may help to mobilize the appropriate resources. Our study revealed that benzodiazepine and cocaine use were risk factors for intubation.

Benzodiazepines are one of the most commonly prescribed substances that can decrease lung function [15, 16]. Long-acting subclasses of these medications can have half-lives that approach 250 h, extending the pharmacological effects on the lungs for weeks [15]. They relax pharyngeal skeletal muscles to narrow the airway, relax intercostal skeletal muscles to decrease lung expansion, and relax smooth muscles of the lungs to decrease minute ventilation, respiratory rate, and tidal volume [16]. The lung capillaries try to compensate by constricting and shunting blood flow to areas with higher perfusion. This process further impairs lung function by inducing hypoxic pulmonary vasoconstriction [17]. Although none of our patients tested positive for barbiturates, they have a similar mechanism of action to benzodiazepines. A toxicology screen may help by allowing healthcare professionals to monitor these patients more closely and provide earlier supportive care.

Urine tests were positive for cocaine for approximately 1 to 2 days. The relatively short half-life of 6 h allows us to assume these patients used this substance around the time of the event [18]. Pulmonary toxicity is uncommon, unless the substance has been inhaled and manifests as “crack lung” [19]. This can induce barotrauma, bronchoconstriction, noncardiogenic pulmonary edema, pneumothorax, pneumomediastinum, and worsening of preexisting lung disease [18]. If cocaine is chronically inhaled, reactive airway disease and tracheal stenosis can develop [20]. We must be aware cocaine administration can occur by many routes, and a patient with inhalation injury testing positive for cocaine metabolites may be at a higher risk for respiratory failure.

The two prospective studies by Silver et al. [5] and Davis et al. [2] both concluded an elevated blood alcohol content on admission is associated with poorer outcomes, larger burns, and more inhalational injury. Fluid requirements, duration of mechanical ventilation, ICU LOS, hospital LOS, and mean hospital charges were all increased in patients with alcohol abuse. Our study, by excluding the large burn patients, suggests that those outcomes are mainly caused by the burn and not inhalation injury.

Smoking is a well-known cause of lung injury and the leading cause of death from fire [21]. The chemicals inhaled by smoking cigarettes deposit in the lungs and induce inflammation. Electronic cigarettes are odorless methods of smoking that still induce lung injury [22]. Aerosolized nicotine has been found to cause inflammation by recruiting inflammatory markers that produce pulmonary edema [22, 23]. The absence of the scent of cigarettes should not eliminate the possibility of a smoker suffering an inhalation injury. Smoking is a risk factor that has historically predisposed patients to burn injuries and intubation [1, 8–10, 24].

Overall mortality in burn patients with inhalation injury has been reported at 41.5% compared to a value of 7.2% in patients not suffering inhalation injury. Age and TBSA of burns are predictors of increased mortality that are incorporated in this high reported mortality of 41.5% [25]. The 2% mortality rate we reported may be lower due to less than 5% TBSA and the expertise of management at a national burn center. Acute respiratory distress syndrome and rates of mechanical ventilation are higher in the literature with larger TBSA burns [26, 27].

Limitations of our study relate to the small sample size and retrospective study design. The sample size did not allow for a normal distribution; therefore, we had to use Fisher’s exact test instead of Pearson’s chi-square test for categorical data. Similarly, we had to use the Mann-Whitney U test instead of the two-sample independent t test for continuous data. The incidence of inhalation injuries has been reported as 2.2% in patients suffering less than 20% TBSA [28]. Our 115 inhalation injury patients with less than 5% TBSA burns over the time period of 8 years was a reasonable population number for our study. We analyzed data from a single-center, limiting the generalizability to the worldwide population. Our population may represent a higher level of alcohol, substance, and smoking use compared to other burn centers. Baltimore City reported 232 substance and alcohol intoxication deaths in 2009, 172 intoxication deaths in 2010, and 165 intoxication deaths in 2011 [29]. Opioid use was measured by codeine and morphine metabolites. Synthetic and semi-synthetic opioids do not routinely test positive by this screening method. This may explain why we did not see a difference in the opioid and non-opioid group of patients intubated.

This was the first study to evaluate sub-categories of substance use, alcohol use, and smoking use as risk factors for intubation in a population with less than 5% TBSA burns. By isolating a group of patients with just inhalation injury, it allowed us to determine risk factors and predictive outcomes specific to this location of burn.