Table 3.

Univariate and multivariable binary logistic regression models for the background variables analyzed fitted for the number of purchased doctor‐prescribed OCS

	N all	N (%) outcome	OR (95% CI)	Adjusted OR (95% CI)
Gender
Female	88	25 (28.4)	1
Male	52	16 (30.8)	1.12 (0.53–2.37)
Age
≥47 years	70	16 (22.9)	1
<47 years	70	25 (35.7)	1.88 (0.89–3.94)
Previous ESS
<3	90	21 (23.3)	1	1
≥3	50	20 (40.0)	2.19 (1.04–4.63)	2.15 (0.98–4.70)
LM score
<16/24	49	11 (22.4)	1
≥16/24	82	29 (35.4)	1.89 (0.84–4.25)
NP score
<4/8	21	6 (28.6)	1
≥4/8	104	32 (30.8)	1.11 (0.40–3.13)
NP eosinophilia
<30%	17	3 (17.6)	1
≥30%	54	16 (29.6)	1.97 (0.50–7.79)
Baseline OCS courses/year
<1	60	11 (18.34)	1	1
≥1	74	29 (39.2)	2.87 (1.29–6.41)	2.78 (1.23–6.26)
Baseline ab courses/year
<4	76	18 (23.7)	1
≥4	50	20 (40.0)	2.15 (0.99–4.66)
Baseline total ethmoidectomy
Yes	22	7 (31.8)	1
No	118	34 (28.8)	0.87 (0.33–2.32)
Baseline start with ATAD
Yes	80	27 (33.8)	1
No	60	14 (23.3)	0.60 (0.28–1.27)

Note: The search for prescription data was performed from nation‐wide electronic prescription database during 2016–2020. The outcome measurement was >1 OCS courses per year or continuous OCS or biological therapy started during the follow‐up. Only the N‐ERD subjects with available e‐prescription data (n = 140) were included. Bold values denote statistical significance at the p < .05 level. The second model is a multivariable model adjusted by the background variables that were associated with follow‐up corticosteroid or antibiotic use at p < .05 level in the first model.

Abbreviations: ab, antibiotic; ATAD, ASA treatment after desensitization; CI, confidence interval; NERD, nonsteroidal anti‐inflammatory drug exacerbated respiratory disease; OCS, oral corticosteroid; OR, odds ratio.