Table 2.

(A) Multivariate linear regression model for LogMAR gain post YAGPC. (B) Multivariate binary logistic regression model for repeat YAGPC.

Independent variable	B coefficient (95% CI)	p value
(A)
Ethnicity (REF Black)
White	−0.024 (−0.059 to 0.011)	0.171
SA	−0.037 (−0.085 to 0.012)	0.137
Operator Grade (REF ANP)
SAS	−0.017 (−0.062 to 0.029)	0.469
Resident	0.057 (0.009 to 0.106)	0.020
Fellow	−0.018 (−0.070 to 0.033)	0.485
Consultant	−0.034 (−0.078 to 0.010)	0.128
Ocular Co-morbidity	−0.068 (−0.104 to −0.032)	<0.001
Age	−0.003 (−0.004 to −0.002)	<0.001

Independent variable	B coefficient	Odds ratio (95% CI)	p value
(B)
Ethnicity (REF White)
Black	0.406	1.501 (0.950 to 2.372)	0.082
SA	0.035	1.035 (0.517 to 2.075)	0.922
Operator Grade (REF ANP)
SAS	0.819	2.268 (1.090 to 4.720)	0.029
Resident	0.885	2.423 (1.103 to 5.321)	0.027
Fellow	0.808	2.243 (1.032 to 4.876)	0.041
Consultant	0.894	2.446 (1.217 to 4.916)	0.012
Ocular Co-morbidity	0.425	1.530 (0.948 to 2.468)	0.082
Age	−0.042	0.959 (0.946 to 0.971)	<0.001

SA South Asian, REF reference variable.

Significance defined as p < 0.05. Significant values in bold.

(A) LogMAR gain: No difference between ethnicities and LogMAR gain. No difference between ANP and other operator grades for LogMAR gain in VA other than residents who had significantly higher LogMAR gain than ANP (p = 0.020). Ocular co-morbidities and increased Age were both strongly correlated to reduced LogMAR gain (p < 0.001 in both instances).

(B) Repeat YAGPC: No difference in ethnicity and repeat YAGPC rate. SAS (p = 0.029), residents (p = 0027), fellows (p = 0.041), and consultants (p = 0.012) all had significantly higher repeat YAGPC rate compared to ANP. Increased age had reduced YAGPC rate (p < 0.001).