Skip to main content
. 2022 Aug 23;72(722):e619–e626. doi: 10.3399/BJGP.2021.0725

Table 2.

Logistic regression model showing the predictors of SABA overprescribing (≥6 SABA in the previous 12 months) for adults with asthma

Variables Number Multivariate logistic modela,b

SABA <6 SABA ≥6 OR 95% CI P-value
Age
Adult 13 098 5970
Older adult (>60) 2686 2139 1.34 1.24 to 1.45 <0.01

Sex
Male 6392 3430
Female 9393 4679 0.85 0.90 to 0.80 <0.01

Ethnicity
White 6066 3297
Mixed 6082 3297 0.87 0.75 to 1.00 0.39
Asian or Asian British 1568 655 1.13 1.05 to 1.21 0.02
Black 778 352 0.94 0.84 to 1.06 0.15
Other 186 81 1.01 0.76 to 1.36 0.03
Not stated/unclassified 1104 427 0.84 0.74 to 0.96 0.03

Prescription type
Repeat 8671 6392
Acute and automatic 7074 1513 0.29 0.27 to 0.31 <0.01
Repeat dispensed 39 204 6.52 4.64 to 9.41 <0.01

Smoking
Never 10 512 4848
Current 2395 1524 1.51 1.38 to 1.63 <0.01
Ex-smoker 2840 1716 1.21 1.13 to 1.32 <0.01

Asthma step
Step 1 2485 670
Step 2 6218 3289 1.79 1.62 to 1.97 <0.01
Step 3 1864 1826 2.87 2.56 to 3.21 <0.01
Step 4+5 136 323 4.99 3.96 to 6.30 <0.01
Unknown 5082 2001 1.19 1.05 to 1.35 <0.01

Oral steroid
Zero 13 188 5955
1 1851 1139 1.35 1.21 to 1.45 <0.01
2 392 412 2.67 2.48 to 2.71 <0.01
≥3 354 603 2.90 2.46 to 3.31 <0.01

Physical comorbidities
0 5401 2044
1 6307 2802 1.07 1.00 to 1.16 0.05
2–3 3448 2591 1.58 1.47 to 1.73 <0.01
≥4 628 672 1.86 1.61 to 2.13 <0.01

Mental comorbidities
0 10 630 4795
1 2796 1612 1.19 1.10 to 1.28 <0.01
2 2204 1519 1.32 1.22 to 1.44 <0.01
3 154 183 2.29 1.80 to 2.91 <0.01
a

This model is adjusted by Index of Multiple Deprivation score.

b

This model is a mixed-effect multivariate logistic regression, which includes clustering based on the practice (Newham, Tower Hamlet and Waltham Forest). The adjusted interclass correlation coefficient of this variable is 0.006, which is negligible. CI = confidence interval. OR = odds ratio. SABA = short-acting beta-agonist.