Table 2. Classification of characteristics.
Characteristics shown in italic font do not exhibit a sufficient predictive power and are therefore not included in our subsequent analysis. We have divided the clinical treatment into four clinical stages to test our neural networks. Stage 1 corresponds to those patients who are in the process of receiving initial medical assessment; therefore, characteristics in categories 1,2, and 3a are assigned. For patients in stage 2, the COVID-19 status and pneumonia characteristics are added to the characteristics vector. When a decision has been made as to whether admit a patient into a hospital or send him/her back home, the hospitalization status characteristic is added at Stage 3. Finally, Stage 4 includes the intubation and ICU characteristics for those patients in critical condition.
| Category | Characteristics | |
|---|---|---|
| 1) Medical History | 1. Diabetes | 7. Obesity |
| 2. COPD | 8. Other chronic illnesses | |
| 3. Immunosuppressive drugs | *Pregnancy | |
| 4. Hypertension | *Asthma | |
| 5. Chronic renal failure | *Smoking | |
| 6. Cardiovascular diseases | ||
| 2) Demographic Data | 9. Gender | *Indigeneous |
| 10. State (birth) | *Indigeneous language | |
| 11. State (residence) | *Migrant | |
| 12. Age | *Foreigner | |
| 3) Recent Medical Info | Category 3a | 13. USMER designation |
| 14. Sector (medical facility) | ||
| 15. State (treatment) | ||
| 16. Days symptoms-treatment | ||
| *Exposure to positive patients | ||
| Category 3b | 17. COVID-19 status | |
| 18. COVID-19-related pneumonia | ||
| 19. Hospitalization status | ||
| 20. Intubation | ||
| 21. ICU | ||