This study looks at over 334,000 cases in the Qatar from a Middle Eastern Emergency Medical Service (EMS) between 2018 and 2022. These studies focused on which patients refused to be transported in an ambulance to the hospital looking at both receiving on scene treatment and without receiving on scene treatment. They utilized various machine learning algorithms such as logistic regression, decision trees, random forest, extra trees, and more to predict patient behavior during the COVID-19 pandemic. The random forest model was the most accurate (74.78%) with an F1 score of 0.74 and an ROC AUC of 0.81. They also found that during the pandemic, the percentage of patients who refused transport to the hospital increase from 24% pre-pandemic to 40% during the pandemic. This was due to the new fear that many developed of going to the hospital and contracting the virus at that time. This model also identified key predictors like age groups (14-59 years), nationality (Qatari, MENA, and South Asian populations), chief complaints (like abdominal pain), and provisional diagnoses (especially neurological issues). Overall, this study found that machine learning models can be used by EMS systems to identify low-risk patients who could be redirected to different healthcare facilities instead of using up the limited emergency resources, which is pivotal during public health emergencies such as a pandemic.