Table 3.

Association rod fracture risk with surgical and instrumentation variables

Confounders	Subgroups	Rod(s) fracture		Odds ratio (95 % Cl: min; max)	P(χ 2)
		Yes	No
Osteotomy by type	Pedicle subtraction	6	31	7.2 (0.8; 62.7)	0.1
	Smith-Peterson	1	37
Osteotomy by location	Thoracolumbar junction	3	17	2.3 (0.5; 11.1)	0.3
	Other	4	51
Use of navigation	Yes	7	62	N/A	>0.5
	No	0	5
Use of cement	Yes	0	16	N/A	0.3
	No	7	48
Screw/rod manufacturer	Device Company 1	3	29	0.9 (0.2; 4.3)	>0.5
	Other	4	34
Pre-contoured rods	Yes	1	24	0.3 (0.03; 3.8)	0.4
	No	3	14
Material of rods	Titanium	5	40	1.2 (0.06; 25.9)	>0.5
	Other	0	4
Diameter of rods	6 mm	3	29	0.7 (0.1; 3.8)	>0.5
	Other	3	20
Type of screws	Polyaxial	5	36	N/A	N/A
	Monoaxial	0	0
Screw density	Incompletea	4	27	2.0 (0.4; 9.8)	0.4
	Complete	3	41
Connectors	Standard	6	48	2.5 (0.3; 22.1)	0.4
	Other	1	20
Interbody support	Yes	4	45	0.7 (0.1; 3.2)	>0.5
	No	3	22
Sagittal rod contour	>60°	5	19	10.0 (1.1; 95.1)	0.04
	≤60°	2	49
Crosslinks	≥2	3	13	3.2 (0.6; 15.9)	0.2
	0–1	4	55
Domino and/or parallel connectorsb	Yes	4	8	10.0 (1.9; 53.1)	0.01
	No	3	60
Number of fused levelsc	≥8	6	41	3.8 (0.4; 33.4)	0.2
	8	1	26
Number of crossing junctions	2	6	27	9.1 (1.0; 80.0)	0.05
	0–1	1	41
Fusion to sacrum	Yes	7	38	N/A	0.1
	No	0	30
Fusion to pelvis	Yes	4	20	3.2 (0.7; 15.6)	0.2
	No	3	48

^aMissing one or more pedicle screws at available locations along construct. ^bPresence of domino and/or parallel-connectors at date of rod fracture. ^cIncluding levels when connecting to prior instrumentation