Table 5.

Factors Affecting the Recovery after Patching Procedure in Traumatic Tympanic Membrane Perforation (by Univariable and Multivariable Logistic Regression Analysis)

	Univariable Logistic Regression				Multivariable Logistic Regression
	B	SE	P	OR (95% CI)	B	SE	P	OR (95% CI)
Patch material (HADA = 1)	1.310	1.200	0.275	3.706 (0.353–38.927)	0.963	1.289	.455	2.619 (0.209–32.758)
Sex (female = 1)	0.211	1.052	0.841	1.235 (0.157–9.708)
Age (years)	–0.011	0.028	0.690	0.989 (0.937–1.044)
Affected side (left = 1)	0.201	1.209	0.868	1.222 (0.114–13.065)
Perforation size (mm)	–0.597	0.665	0.369	0.550 (0.150–2.026)
Duration of perforation (days)	–0.163	0.108	0.132	0.850 (0.688–1.050)	–0.163	0.117	.165	0.850 (0.676–1.069)

Univariable and multivariable logistic regression analyses were performed. The variables that were relatively significant (P < .3) in univariable analysis were involved in the multivariable analysis. The dependent variable was recovery or not (failure = 0, recovery = 1). The explanatory variables were patch material (paper = 0, HADA = 1), sex (male = 0, female = 1), age, affected side (right = 0, left = 1), perforation size, and duration of perforation.

B, unstandardized regression coefficient; SE, standard error; OR, odds ratio.