Table 4.

The Impact of Opium Use on the Severity of Pulmonary Involvement in CT scans of Hospitalized COVID-19 Patients

Outcomes	Model	OR (95% CI)	P-value
Ground Glass Opacities	Unadjusted	1.26 (0.50, 3.21)	0.627
	Full adjusted	0.66 (0.09, 4.62)	0.672
	Optimal model	0.76 (0.14, 4.18)	0.752
Posterior Lower Lobe Prediction	Unadjusted	0.75 (0.32, 1.77)	0.511
	Full adjusted	0.64 (0.24, 1.71)	0.374
	Optimal model	0.66 (0.37, 1.18)	0.163
Pure Consolidation	Unadjusted	0.74 (0.23, 2.40)	0.617
	Full adjusted	3.56 (0.29, 44.02)	0.323
	Optimal model	1.97 (0.62, 6.26)	0.251
Peripheral Subpleural distribution	Unadjusted	0.49 (0.21, 1.17)	0.108
	Full adjusted	1.31 (0.40, 4.24)	0.655
	Optimal model	1.34 (0.45, 3.94)	0.598
Pleural Effusion	Unadjusted	0.49 (0.06, 4.20)	0.519
	Full adjusted	–––	–––
	Optimal model	–––	–––
Multiple Lesions	Unadjusted	0.99 (0.42, 2.33)	0.980
	Full adjusted	0.66 (0.23, 1.86)	0.429
	Optimal model	0.69 (0.26, 1.81)	0.453
Crazy Paving Pattern	Unadjusted	2.35 (0.77, 7.15)	0.134
	Full adjusted	–––	–––
	Optimal model	5.45 (0.44, 474.33)	0.118
Air Bronco Gram	Unadjusted	0.33 (0.07, 1.52)	0.154
	Full adjusted	–––	–––
	Optimal model	–––	–––
Bilateral Involvement	Unadjusted	0.97 (0.43, 2.16)	0.935
	Full adjusted	0.89 (0.30, 2.68)	0.836
	Optimal model	0.90 (0.30, 2.66)	0.483

A full adjustment method was employed, incorporating all variables, and the optimal model was determined using a stepwise backward algorithm