Efficacy and cost-effectiveness of universal pre-operative iron studies in total hip and knee arthroplasty

Table 7 Factors affecting blood transfusion/blood loss in TKA

Total knee arthroplasty data	OR/estimate (95% CI)	Comparison P value	Global P value
Bivariate analysis^a
Previous iron transfusion	13 (3, 56)		0.0007
Multivariable analysis for outcome: blood transfusion^a
Late pre-operative haemoglobin	0.80 (0.70, 0.91)^a		0.0010
Multivariable analysis for outcome: blood loss
Type of anaesthesia			0.0111
Spinal/GA vs GA	−111.4 (−185.6, −37.4)	0.0032
Spinal vs GA	−71.3 (−135.2, −7.5)	0.0286
Anticoagulants			<.0001
0 vs 2^b	−274 (−417, −130)	0.0002
0 vs 3^b	−232.7 (−403.0, −62.5)	0.0074
Multivariable analysis for outcome: haemoglobin drop
BMI	−0.02 (−0.04, 0.0)		0.0120
Gender (male vs female)	−0.35 (−0.56, −0.15)		0.0007
Anticoagulants			0.0288
0 vs 2^b	−0.67 (−1.26, −0.08)	0.0249
1 vs 2^b	−1.44 (−2.41, −0.48)	0.0034
Late pre-operative haemoglobin	0.02 (0.01, 0.03)		<.0001

^aBinary logistic regression – odds ratio (95% confidence interval); modelling the probability that blood transfusion = ‘Yes’
^bAnti coag (0 = Clexane, 1 = aspirin, 2 = warfarin, 3 = all others [rivaroxaban, apixaban])

ISSN: 1749-799X