Assessment of Features of Carbon Monoxide and Cyanide Inhalation Poisoning in Patients With Flame Burns at Kenyatta National Hospital

Abstract

Background & Justification of the Study Patients with flame burns invariably inhale smoke but about 60% of those presenting to hospital sustain heat or smoke inhalation that is significant to contribute to their morbidity and mortality. A direct systemic effect of smoke inhalation is caused by inhalation of toxic gases during the combustion of organic and inorganic substances. The two most common gases that contribute to morbidity and mortality of flame burnt patients are carbon monoxide (CO) and cyanide (CN). These two systemic inhaled toxicants cause many of the fire and smoke related acute deaths and should be suspected in every fire victim. The median mortality among admitted fire victims occurs at day 5 of admission. The bulk of early deaths are attributable to burn shock and inhalational injuries. It is probable that among the causes of these early deaths is inhaled smoke containing toxicants such as CO and CN whose measurement was not being done in flame-burnt patients presenting to KNH. This gap was the basis of this study. Aim & Objectives To assess for features of carbon monoxide and cyanide inhalational poisoning in flame burnt patients presenting to Kenyatta National Hospital. Methodology This was a prospective descriptive study. Eighty two consecutive non-pediatric consenting patients presenting with acute (<24 hrs) flame burns were recruited into the study between April 2017 to February 2018. Features of possible toxic smoke inhalational poisoning were assessed and carboxyhemoglobin levels determined by a Masimo SETR Radical 57TM CO-oximeter system at admission. Serum lactate levels were also determined by Siemens RapidPointR500 blood gas analysis machine which was used as a surrogate marker for cyanide toxicity. Data was analysed with Statistical Package for Social Sciences (SPSS) for Windows Version 21 with p value of <0.05 considered significant. Study Shortcomings Lactate levels were done only on 48% of the patients due to frequent unavailability of the service at the concerned laboratory. 2 ©Mackutwa E.N., Postgraduate Dissertation 2018, UoN Results Eighty two subjects were recruited and after data cleaning, analysis was done on eighty. Most of the subjects resided in estates considered low class. Fifty six percent were males and 44% were females. Kerosene stove explosions contributed to most of the mechanisms of burns, at 28.7%. Over 53% of the material burnt by the fire was worn clothes only. Commonest complaints that could point towards toxic inhalation was Confusion (n=23) and headache (n=21). Seventy percent (n=56) of the respondents had singed nasal vibrissae and 25% had upper airway hyperemia or edema which pointed towards likelihood of significant inhalational injury. The mean TBSA was 30.9% and average SPCO% was 5.48%. Average Arterial lactate was 2.91mmol/L. Overall mortality was 38.7% out of which 22% occurred in less than 24 hours after admission. Factors found to have positive correlation with death included upper airway edema/hyperemia (p=0.024), %TBSA (p=0.01), lactate (p=0.01) and Base deficit (p=0.01); but not SPCO% (p=0.708). Discussion Rampant use of open flames for household cooking and lighting by most of the subjects who resided in relatively low class houses where precaution to fire outbreaks may not be taken may have contributed to the flame injuries. The, mean SPCO% of 5.48% fell below levels that would cause symptoms hence may not have accounted for the confusion and headache seen in over one third of the subjects neither could it account for the morbidity/mortality because it was below toxic levels. In those whom lactate levels were measured, only 7 had lactate levels of >5mmol/L. However, this level was far below 10mmol/L, which is the threshold to suspect cyanide poisoning in acute burns. Therefore, the 22% of all deaths that occurred before 24 hours of admission may be attributed to well-known factors that cause acute deaths of burn patients such as burn shock and inhalational injury. There may have been little or no contribution by CO and CN inhalation. Conclusion and Recommendations A large proportion of flame burnt patients presenting acutely to KNH have symptoms that are seen in CO or CN inhalational poisoning. However, due to low SpCO% and low lactate levels in majority of them, these symptoms are considered nonspecific for toxic inhalational poisoning. Therefore CO and CN may not be a significant factor in their poor outcomes as hypothesized in this study. Nonetheless, many of these patients have significant inhalational injury which has a strong positive correlation to their deaths and hence would still require early high flow oxygen administration and probable tracheal intubation.