Keywords
Body Mass Index, Septic Shock, Survival
Disciplines
Critical Care
Abstract
Background: Obesity is one of the most widespread epidemics of our time. In fact, currently 65.7% of US adults age 20 and older are overweight, while 30.6% are obese. It has been well-established that obesity has numerous adverse effects on long-term health, however the specific effect on patients treated for sepsis and septic shock is unclear. Body Mass Index (BMI) is a measure of total body fat content and surrogate marker for obesity. In our study, we aimed to identify if BMI was an independent risk factor for poor survival or increased length of stay (LOS) in patients with sepsis.
Methods: We retrospectively selected patients with diagnostic codes of sepsis and septic shock who were admitted to the ICU over three years. These patients were further separated into groups of alive and deceased. Based on their perceived association with mortality in sepsis, numerous variables were investigated, such as BMI, LOS, age, cirrhosis, chronic kidney disease (CKD), lactate, age, multiple organ dysfunction syndrome (MODS), and APACHE II scores. Specifically, BMI was classified into sub-groups, including underweight (BMI30).
The alive and deceased groups were initially compared for any significant differences with univariate analysis. Thereafter, the significant variables were analyzed using multivariate analysis to assess whether any were able to independently predict mortality in sepsis.
Results: Our study selected 293 patients with sepsis, including 185 alive and 108 deceased. Interestingly, our univariate analysis revealed that underweight and obese patients exhibited slightly less mortality in sepsis compared to normal and overweight patients. However, these results did not reach statistical significance, with a p-value of 0.30; this was confirmed in multivariate analysis, which resulted in a p-value of 0.08. Additionally, underweight, overweight, and obese patients had a slightly decreased median LOS in the ICU and hospital compared to patients with normal BMI. Nevertheless, these results were not significant either, with ICU LOS p-value of 0.22 and hospital LOS p-value of 0.45.
Univariate analysis identified certain variables that reached statistical significance, including cirrhosis (p2 (p=0.03), median lactate (p=0.05), age (p>.01), and APACHE II scores (p>0.01). Multivariate analysis of these variables established that only the presence of cirrhosis (p=0.03), age (p
Conclusion: The data suggests that normal BMI in patients with sepsis may result in increased mortality and LOS both in the ICU and hospital, though this was not statistically significant. Other variables that were significant independent predictors for mortality in sepsis were cirrhosis, mean age, and mean APACHE II score. As the obesity epidemic continues to rise, further inquiry into the association of BMI and mortality in sepsis is needed.
