. Author manuscript; available in PMC: 2018 Apr 1.

Published in final edited form as: J Arthroplasty. 2016 Nov 15;32(4):1153–1158.e1. doi: 10.1016/j.arth.2016.11.007

Table 2.

Multiple Logistic regression coefficients for four key variables for predicting nonresponse from Dowsey and colleagues and from the OAI data

Variable		Data from Dowsey et al			Data from OAI
Variable	Coefficient (se)	Odds Ratio	P value	Coefficient (se)	Odds Ratio	P value
Intercept	−1.53 (0.50)		0.002	0.57 (0.45)		0.20
BMI ≥40	1.25 (0.29)	3.48	<0.001	0.04 (0.91)	1.04	0.95
K&L Grade ≤3	0.95 (0.25)	2.59	<0.001	0.98 (0.36)	2.66	0.007
WOMAC Total	−0.21 (0.09)	0.81	0.02	−0.83 (0.15)	0.44	<0.001
MCS (Mild)	−0.08 (0.35)	0.93	0.83	−1.05 (0.62)	0.35	0.09
MCS (Moderate)	0.55 (0.31)	1.74	0.075	1.24 (0.87)	3.46	0.15
MCS (Severe)	1.19 (0.42)	3.30	0.005	-	-	-

OAI, Osteoarthritis Initiative; SE, standard error; BMI, body mass index; K&L, Kellgren and Lawrence knee osteoarthritis grade; WOMAC, Western Ontario and McMaster Universities Arthritis Index, MCS, mental component score of the Short Form Health Survey.