Skip to main content

Table 1 Variables included in the prognostic model are shown for the derivation cohort

From: Development of a prognostic model of COVID-19 severity: a population-based cohort study in Iceland

Predictor Derivation cohort
n = 4756
Median and interquartile range for continuous or n and % for categorical variables
(n missing and %)
Age, years 40, 28–54
(0, 0%)
Sex, male 2455, 51.6%
(0, 0%)
Body mass index 26.0, 23.1–29.45
(1,441, 30.3%)
Current smoking 395, 9.1%
(494, 8.5%)
Diabetes 136, 3.0%
(242, 5.1%)
Hypertension 569, 12.6%
(225, 0.7%)
Heart disease 282, 6.2%
(235, 4.9%)
Chronic kidney disease 25, 0.6%
(246, 5.2%)
Pulmonary disease 245, 5.4%
(239, 5.0%)
Cancer 114, 2.5%
(243, 5.1%)
Flu-like symptoms 3781, 80.8%
(77, 2.2%)
Upper respiratory symptoms 2727, 59.5%
(169, 3.6%)
Lower respiratory symptoms 1206, 26.6%
(226, 4.8%)
Gastrointestinal symptoms 1018, 22.6%
(244, 5.1%)
Clinical score = moderate or high severity 470, 10.6%
(306, 6.4%)
Telehealth only 4143, 87.1%
(0, 0%)
Urgent care visit 375, 7.9%
(0, 0%)
Hospitalization 188, 4.0%
(0, 0%)
Intensive care unit admission or death 50, 1.1%
(0, 0%)
  1. Continuous variables are summarized as medians and interquartile ranges (IQR). The number of cases behind each categorical variable are presented along with the percentage. For each of the variables, the number and proportion of cases with missing data are displayed within parenthesis. Two candidate predictor variables (chronic kidney disease [n = 25] and clinical score = high severity [n = 74]) were not included in the final model due to small sample sizes