A healthy body activates the immune response to target invading pathogens (i.e., viruses, bacteria, fungi, and parasites) and avoid further systemic infection. The activation of immunological mechanisms includes several components of the immune system, such as innate and acquired immunity. Once any component of the immune response to infections is abnormally altered or dysregulated, resulting in a failure to clear infection, sepsis will develop through a pro-inflammatory immunological mechanism.
As a severe medical problem, sepsis has imposed a significant socioeconomic burden on patients and physicians in both pediatric and adult intensive care units. Sepsis is characterized as a life-threatening condition with a series of pathophysiological symptoms such as hypertension, leukocytosis/leukopenia, and hyper/hypothermia, according to the third international consensus definitions of sepsis. All these symptoms are consequences of the dysregulated immune response to an infection, which induces systemic inflammation and response syndrome (SIRS).
Additionally, late in the progression of SIRS, the decreased resistance is responsible for the induction of disseminated intravascular coagulation, multi-organ dysfunction syndrome (MODS) and eventually leads to the death of the patients. Some scholars postulate that septic shock is a subtype of sepsis characterized by cellular and metabolic abnormalities and particularly profound cardiopulmonary circulation dysfunction, which leads to a greater risk of death than sepsis alone. Severe sepsis is another stage of disease development that involves an elevated plasma lactate level or even lactic acidosis, oliguria, acute respiratory distress syndrome (ARDS) and mental disorder of the patients.
Acute respiratory distress syndrome (ARDS) is a devastating complication of severe sepsis. Sepsis and ARDS have similar underlying mechanisms, characterized by inflammation and endothelial dysfunction. In addition, severe sepsis is the most common etiology of ARDS, and patients with sepsis-induced ARDS have higher case fatality rates than patients with other risk factors of ARDS. Furthermore, the severe inflammatory responses induced by sepsis also increases vascular permeability, leading to acute pulmonary edema and resulting in acute respiratory distress syndrome (ARDS). Apparently, we can predict the early onset of ARDS and by administrating drugs prevent its further spread. Thus, this article presents a comprehensive analysis that highlights the relationship between sepsis and ARDS by evaluating the P/F ratio by Hospital admission time therefore suggesting a direction for treatment of sepsis induced ARDS.
The diagnostic criteria for ARDS utilize the PaO2/fraction of inspired oxygen (FiO2) [P/F] ratio to assess the degree of hypoxemia (abnormally low concentration of Oxygen in the blood). The acute hypoxemic resp failure arises from widespread diffuse injury to the alveolar - capillary membrane (Alveolar refers to the alveoli, the millions of tiny air sacs that are scattered throughout the lungs. The capillaries are very tiny blood vessels that connect the alveoli to larger blood vessels. When a person breathes in air, oxygen travels to the lungs and into the alveoli) which is termed as ARDS. The diagnosis of ARDS is based on the following criteria: -
1. Acute Onset –
Sudden, rapid, or unanticipated development of sepsis or its symptoms.
2. Bilateral Lung Infiltrates on chest radiography of a non-cardiac origin-
Bilateral interstitial is a serious infection that can inflame and scar the lungs. It's one of many types of interstitial lung diseases, which affect the tissue around the tiny air sacs in lungs.
3. PaO2/FiO2 (P/F) ratio-
The ratio of PaO2 to the fraction of inspired oxygen (PaO2/FIO2) is commonly used to determine the severity of acute lung injury and acute respiratory distress syndrome (ARDS).
DATA SOURCE FOR ANALYSIS
We obtained the clinical data from three geographically distinct U.S. hospital systems with three different electronic medical record systems: Beth Israel Deaconess Medical Center (hospital system A), Emory University Hospital (hospital system B), and a third, unidentified hospital system (hospital system C). These data were collected over the past decade with approval from the appropriate Institutional Review Boards. Data and labels for 40,336 patients from hospital systems A and B were posted publicly for download and data and labels for 24,819 patients from hospital systems A, B, and C were sequestered as hidden test sets.
The Challenge data consisted of a combination of hourly vital sign summaries, lab values, and static patient descriptions. In particular, the data contained 40 clinical variables: 8 vital sign variables, 26 laboratory variables, and 6 demographic variables; Table 1 describes these variables. Altogether, these data included over 2.5 million hourly time windows and 15 million data points. Data extracted from the Electronic Medical Record (EMR) underwent a series of preprocessing steps prior to formal analysis and model development.
Table 1: - Clinical time series data used for Analysis: vital signs (rows 1-8), laboratory values (rows 9-34), demographics (rows 35-40), and outcome (row 41).
To analyze the ARDS, we had to calculate the P/F ratio, but from the given clinical data set we had SaO2 and O2Sat, therefore it was important to first do the conversion from SpO2 (%) to PaO2(mmHg).
To calculate the P/F Ratio we first created a Calculated Field – PaO2mmHg
For the conversion we had to take 3 different categories of SpO2(%) (A.K.A O2Sat) shown below: -
SpO2 (%) to PaO2(mmHg)
1. 100 to 90
For every single drop in % of SpO2 the PaO2 will decrease by 4 mmHg
For every single drop in % of SpO2 the PaO2 will decrease by 1.5mmHg
3. Below 80
For every single drop in % of SpO2 the PaO2 will be decreased by 50% mmHg
Table 2: -Shows the 3 categories of SpO2(O2Sat) in % and their conversion calculation in mmHg
Conversion Calculation: -PaO2mmHg
Fig1: - Calculated field Pao2 mmHg
Once the SpO2 (aka O2Sat) was converted to PaO2 mmHg. We will create another calculated field to calculate the P/F ratio –To obtain the Ratio we will divide the Calculated Field PaO2 mmHg by the Fraction of Inspired Oxygen FiO2
P/F Ratio = [PaO2 mmHg]/ [FiO2]
Fig2: - Shows the Calculated Field P/F Ratio
Now, after having the P/F ratio, we will plot this against the hospital admission time , gender wise where blue indicates females and orange indicates males, also further filtered on sepsis checker (calculated field for checking values on sepsis and Non sepsis) for only sepsis patients and P/F ratios filtering non null values only.
Fig 3: - Showcasing here P/F Ratio vs. Hosp Adm Time by gender in tableau. The data is filtered on Sepsis_Checker for only sepsis patients and P/F Ratio filtering non-Null values only.
Normal Ranges: -
P/F > = 400 indicates normal
P/F < 400 - Hypoxemia
P/F < 300 - Respiratory Failure
P/F < 250 - Severe Respiratory Failure
P/F < 200 - Critical Respiratory Failure
From the data visualization, it indicates that the females from the time of admission to 500th hour of admission was seen to have severe to critical ARDS condition.
Thus, the P/F ratio is a powerful objective tool to identify acute hypoxemic respiratory failure when supplemental oxygen has already been administered and no room air ABG (arterial blood gas) is available, or pulse oximetry readings are unreliable.
We can conclude that the P/F ratio measured within the first 24 hours after hospital admission, is an independent indicator of ARDS development among sepsis patients at risk.