Table 1 Thirty-nine variables used to train the predictive model.
Variables | |
|---|---|
Age | Vital signs on presentation |
Gender | Systolic blood pressure |
Bed-ridden | Diastolic blood pressure |
Mental change | Pulse rate |
Nursing home residence | Respiratory rate |
Body temperature | |
Medical device | Peripheral oxygen saturation |
Urinary catheter/Cystostomy | |
Complete blood cell counts | |
Infection type | Leukocyte count |
Hospital-acquired | Hemoglobin |
Platelet | |
Comorbid conditions | |
Diabetes mellitus | |
Hypertension | Other laboratory findings |
Cardiovascular disease | C reactive protein |
Chronic renal failure | Glucose |
Cerebrovascular disease | Blood urea nitrogen |
Malignancy | Creatinine |
Past history | Arterial blood gas analysis |
Hospitalization within 3 months | pH |
Antibiotics use within 3 months, | PCO2 |
UTI within a year | PO2 |
HCO3− | |
Assessment models | Arterial oxygen saturation |
qSOFA | Lactic acid |
Hypothermia | |
Leukopenia | |
Medication | |
Antacid, H2 blocker or proton pump inhibitor | |
