Skip to main content

Table 3 Different multiple logistic regression models predict whether a donor will yield a high NT titer

From: Simple prediction of COVID-19 convalescent plasma units with high levels of neutralization antibodies

 

Log OR (95% CI)

OR (95% CI)

p-value

Model 1

   

N training = 491, N evaluation = 211

   

Gender (female)

0.14 [-0.49-0.78]

1.15 [0.61–2.17]

0.66

Age (years)

0.02 [0.0-0.04]

1.02 [1.00-1.04]

0.07

Body weight (kg)

-0.15 [-0.32-0.02]

0.86 [0.72–1.02]

0.09

Height (cm)

0.14 [-0.03-0.32]

1.16 [0.97–1.38]

0.11

Body mass index (kg/m2)

0.46 [-0.07-1]

1.59 [0.93–2.71]

0.09

Days after start of COVID-19 symptoms

0.02 [0.01–0.02]

1.02 [1.01–1.02]

< 0.001

COVID-19 symptoms duration (days)

0.01 [-0.02-0.03]

1.01 [0.98–1.03]

0.53

Model 2 (univariate logistic regression)

   

N training = 590, N evaluation = 254

   

Abbott quantitative SARS-CoV-2 Ab test (100 BAU/ml)

0.41 [0.33–0.50]

1.51 [1.38–1.64]

< 0.001

Model 3

   

N training = 590, N evaluation = 254

   

Abbott quantitative SARS-CoV-2 Ab test (100 BAU/ml)

0.37 [0.28–0.46]

1.44 [1.32–1.58]

< 0.001

Vaccination

0.97 [0.14–1.8]

2.64 [1.15–6.04]

0.02

Model 4

   

N training = 420, N evaluation = 180

   

Gender (female)

-0.53 [-1.43-0.36]

0.59 [0.24–1.44]

0.24

Age (years)

0.00 [-0.02-0.03]

1.00 [0.98–1.03]

0.77

Body weight (kg)

-0.23 [-0.47-0.02]

0.80 [0.62–1.02]

0.07

Height (cm)

0.19 [-0.06-0.44]

1.21 [0.94–1.56]

0.14

Body mass index (kg/m2)

0.7 [-0.05-1.45]

2.01 [0.95–4.27]

0.07

Days after start of COVID-19 symptoms

0.00 [-0.01-0.01]

1.00 [0.99–1.01]

0.66

COVID-19 symptoms duration (days)

0.00 [-0.03-0.04]

1.00 [0.97–1.04]

0.88

Abbott quantitative SARS-CoV-2 Ab test (100 BAU/ml)

0.32 [0.22–0.43]

1.38 [1.25–1.53]

< 0.001

Vaccination

1.73 [0.56–2.9]

5.66 [1.76–18.2]

0.004

  1. Notes: Model 1 was trained using data from 491 donors and its performance was evaluated on 211 donors. Model 2 was trained using data from 590 donors and its performance was evaluated on 254 donors. Model 3 was trained using data from 590 donors and its performance was evaluated on 254 donors. Model 4 was trained using data from 420 donors and its performance was evaluated on 180 donors