Asthma is characterized by episodes of respiratory distress, called exacerbations. Exacerbations are currently treated on basis of their severity that involves correcting hypoxemia with oxygen and inhaling antagonists and corticosteroids to reverse airflow. There is a current push to treat exacerbations in a more personalized manner, in accordance with a patient's specific phenotypes. However, potential risk factors and the weight of their influence on the severity of exacerbation that is developed currently need to be identified. Data from those who were admitted to the Amsterdam University Medical Center with an asthma exacerbation were collected. Eligible patients were above the age of 18 at the time of admission and had been physician-verified to experience exacerbations. Study parameters (such as demographics, pre-existing conditions/other comorbidities, medication use, labratory findings, dosage of corticosteroids inhaled, lung function and spirometry values, etc.) were identified and measured; the severity of exacerbations was characterized by periphery oxygen saturation over inspired fraction of oxygen, aka SpO2/FiO2. Using Least Absolute Shrinkage and Selection Operator (LASSO) enabled the researchers to identify the risk factors with the highest predictive value for exacerbations. LASSO was first used on different subsets containing random selections of 80% of the patients and then random selections of 20% of patients before finally being trained on the dataset as a whole. The unpaired T-test, Mann-Whitney U test, and Chi squared test were used to assess normally distributed numeric variables, non-normally distributed numeric variables, and categorical variables. In the end, data from 377 patients was used, who had come to the emergency department with 644 exacerbations between all of them. Of the exacerbations, 43.3% needed hospital admission, 4.5% needed ICU admission, and 1.1% needed ventilation; nearly three-quarters of exacerbations involved inhalation of corticosteroids just before, and nearly half of exacerbations also involved the patient reporting flu-like symptoms as well. For those who were admitted to the hospital, the exacerbations where characterized by obstruction and lung function when measured by spirometry indicated a lower forceed expiratory volume. For those who needed admission to the ICU, the exacerbations had a higher blood neutrophil count, higher FiO2, higher partial pressure of CO2, and lower SpO2 and lower arterial blood gas pH. LASSO regression was applied to determine the most important risk factors and found that eosinophil and neutrophil count were selected for predictors of oxygenation efficiency, neutrophil count was an indicator for ICU admission, C-reactive protein (CRP) was an indicator for hospital admission, and infiltrate presence on chest X-Ray or CT scan was a statistically significant indicator for all outcomes except ICU admission. The area under curve values for prediction models were low when tested for both hospital and ICU admission - 0.632 and 0.695, respectively.