Table 2.

Adjusted estimated regression coefficients in algorithm derivation

Coefficient	β-coefficient (SE)	P-value
Constant	−5.013	—

Factors considered for algorithm

Age group, years, reference group ≤60 years
61–80	0.192 (0.076)	0.01

BMI, kg/m², reference group 18.5–<25 kg/m²
Underweight: <18.5 kg/m²	0.597 (0.179)	0.001
Overweight: 25– <30 kg/m²	−0.210 (0.087)	0.02
Obese: ≥30 kg/m²	−0.103 (0.083)	0.21

Smoking status, reference group non-smoker
Current smoker	0.251 (0.089)	0.005
Ex-smoker	0.093 (0.078)	0.24

Blood eosinophil counts, reference group ≤400/µl
>400/µl	0.237 (0.084)	0.005

Rhinitis diagnosis and/or drugs	−0.212 (0.068)	0.002

Diabetes diagnosis and/or therapy	0.378 (0.071)	<0.001

IHD diagnosis	0.243 (0.116)	0.036

Anxiety and/or depression	0.192 (0.068)	0.005

History of anaphylaxis	0.790 (0.275)	0.004

GINA management step, reference group step 1–2
No therapy	0.568 (0.121)	<0.001
Step 3	0.101 (0.108)	0.35
Step 4–5	0.461 (0.078)	<0.001

GP consultations for LRTIs
1	0.313 (0.086)	<0.001
≥2	0.206 (0.122)	0.09

Acute oral steroids
1	0.551 (0.095)	<0.001
2	0.975 (0.120)	<0.001
≥3	1.141 (0.107)	<0.001

Paracetamol	0.204 (0.071)	0.004

Hospitalisation, HES data
≥1	1.877 (0.105)	<0.001

The pseudo R² for the equation was 0.0846. Example: for a 24-year-old non-smoker with a BMI of 28 kg/m², a history of rhinitis, and anxiety, who is receiving Step 3 GINA therapy and has had one course of prednisolone for asthma in the previous year, the at-risk probability = e^x/(1 + e^x) = e^−2.615/(1 + e^−2.615) = 0.01, where x = −4.591, calculated from −5.013 (constant) −0.210 (BMI) −0.212 (rhinitis) + 0.192 (anxiety) + 0.101 (asthma treatment) + 0.551 (asthma attacks). BMI = body mass index. GINA = Global Initiative for Asthma. HES = Hospital Episode Statistics. IHD = ischaemic heart disease. LTRI = lower respiratory tract infection. SE = standard error.