The association between metal allergy, total hip arthroplasty, and revision

Figures & data

Table 1.  Characteristics of patients undergoing primary total hip replacement and revision between 1995 and 2007.

	DHAR ^a(n=70,698) % (n)		Cases ^b (n=356) % (n)		OR (95% CI) ^c	p-value ^c
Gender
Female	58.8	(41, 136)	67	(238)	0.68 (0.55–0.85)	0.001
Age-groups
10–49	5.8	(4,145)	1.1	(4)	5.45 (2.03–14.61)	0.001
50–59	13	(9,442)	4.5	(16)	3.26 (1.97–5.38)
60–69	28	(20,060)	23	(81)	1.34 (1.04–1.71)
70–79	35	(24,821)	32	(113)	1.15 (0.92–1.44)
>80	18	(12,586)	40	(142)	0.32 (0.26–0.40)
Primary diagnosis
Primary osteoarthritis	76	(53,944)	70	(249)	1.35 (1.08–1.70)	0.05
Sequelae of trauma	14	(10,183)	19	(68)	0.71 (0.54–0.92)
Avascular necrosis	2.4	(1,726)	3.4	(12)	0.71 (0.40–1.27)
Rheumatoid arthritis	2.1	(1,487)	3.1	(11)	0.67 (0.37–1.22)
Paediatric conditions	3.4	(2,385)	2.5	(9)	1.34 (0.69–2.60)
Other diagnoses	1.9	(1,329)	2.0	(7)	0.95 (0.45–2.01)
Cause of revision
Aseptic loosening	7.4	(5,284)	7.9	(28)	0.94 (0.64–1.39)	0.2
Osteolysis	0.2	(133)	0.3	(1)	0.67 (0.09–4.77)
Deep infection	1.3	(908)	1.1	(4)	1.14 (0.42–3.06)
Fracture or luxation	2.8	(1,986)	1.1	(4)	2.53 (0.94–6.79)
Other	1.3	(925)	0.6	(2)	2.33 (0.58–9.39)
Femoral head material
Metal	76	(53,618)	82	(290)	0.70 (0.54–0.91)	0.2
Ceramics	16	(11,292)	13	(46)	1.27 (0.93–1.74)
Not exchanged (at revision)	0.004	(29)	0	–
Other or missing	8.6	(6 115)	5.6	(20)	1.58 (1.01–2.49)
Acetabular liner material
Polyethylene	40	(28,425)	31	(111)	1.47 (1.18–1.84)	0.3
Ceramics	2.6	(1,862)	0.8	(3)	3.17 (1.02–9.87)
Metal	1.6	(1,132)	1.4	(5)	1.14 (0.47–2.75)
Other or missing ^d	56	(39,635)	67	(237)	0.63 (0.51–0.79)
Articulations
Metal-on-polyethylene	28	(19,933)	23	(83)	1.28 (1.00–1.64)	0.5
Ceramic-on-ceramic or
ceramic-on-polyethylene	12	(8,211)	7.0	(25)	1.73 (1.15–2.60)
Metal-on-metal	1.1	(748)	1.1	(4)	0.94 (0.35–2.52)
Missing ^b	59	(42,161)	68	(244)	0.67 (0.54–0.84)

^a DHAR: Patients from the Danish Hip Arthroplasty Register (DHAR) 1996–2007.

^b Cases, defined as dermatitis patients from the DHAR who were also examined for metal allergy at Gentofte

Hospital.

^c Comparison of patients from the Danish Hip Arthroplasty Registry (DHAR) and patients from the DHAR

who were also examined for metal allergy (Chi-square test). Fisher's exact test was used when the expected

frequency in one of the cells was lower than 5.

^d The proportion of missing data was high, as the variable was introduced in recent years.

OR: odds ratio.

CI: confidence interval.

Table 2.  The prevalence of metal allergies in cases (n = 356) and matched controls (n = 712) from the patch-test database at Gentofte Hospital, Denmark

	Cases (n = 356) % (n)	Controls (n = 712) % (n)	p-value ^a
Metal allergy ^b	14 (51)	14 (97)	1.0
Nickel allergy	10 (36)	9.8 (70)	0.8
Cobalt allergy	5.3 (19)	5.2 (37)	0.5
Chromium allergy	3.1 (11)	3.9 (28)	0.4

^a Comparison of cases and controls matched by sex and by age in

years (conditional logistic regression analysis).

^b Metal allergy was defined as a positive patch-test reaction to nickel, cobalt or chromium.

Table 3.  The prevalence of metal allergies in cases (n = 356) from the patch-test database, stratified by the number of revisions

	Cases operated once (n = 273) % (n)	Cases revised once (n = 70) % (n)	Cases revised two or three times (n = 13) % (n)	OR (95% CI) ^a	p-value ^a
Metal allergy ^b	16 (43)	11 (8)	0	1.38 (0.68–2.80)	0.4
Nickel allergy	11 (29)	10 (7)	0	1.06 (0.48–2.32)	0.9
Cobalt allergy	5.5 (15)	5.7 (4)	0	0.96 (0.32–2.81)	1.0
Chromium allergy	2.9 (8)	4.3 (3)	0	0.68 (0.18–2.51)	0.7

^a Comparison of cases without revision and those with one revision (Chi-squared test). Fisher's exact test was used when the expected frequency in one of the cells was lower than 5.

^b Metal allergy was defined as a positive patch-test reaction to nickel, cobalt, or chromium.

OR: odds ratio.

CI: confidence interval.

Table 4.  The prevalence of metal allergies and the type of material inserted during hip operation in patients who were patch-tested prior to the operation (n = 292) and in patients who were patch-tested after the operation (n = 64)

	Cases patch-tested
	before operation (n = 292) % (n)	after operation (n = 64) % (n)	OR (95% CI) ^a	p-value ^a
Metal allergy ^b	16 (47)	6.3 (4)	2.87 (1.00–8.30)	0.4
Nickel allergy	11 (33)	4.7 (3)	2.59 (0.77–8.73)	0.1
Cobalt allergy	6.5 (19)	0 (0)	–	0.03
Chromium allergy	3.1 (9)	3.1 (2)	0.98 (0.21–4.67)	1.0
Cause of revision
Aseptic loosening	7.5 (22)	9.4 (6)	0.79 (0.31–2.03)	0.2
Osteolysis	0	1.5 (1)	–
Deep infection	1.0 (3)	1.5 (1)	0.65 (0.07–6.39)
Fracture or luxation	1.3 (4)	0	–
Other	0.3 (1)	0	–
Femoral head material
Metal	81 (236)	84 (54)	0.78 (0.37–1.63)	0.6
Ceramics	13 (38)	13 (8)	1.05 (0.46–2.37)
Not exchanged (at revision)	0	0	–
Other or missing	6.2 (18)	3.1 (2)	2.04 (0.46-9.01)
Acetabular liner material
Polyethylene	35 (103)	13 (8)	3.81 (1.75–8.31)	0.001
Metal	1.7 (5)	0	–
Ceramics	1.0 (3)	0	–
Other or missing	62 (162)	88 (56)	0.18 (0.08–0.39)

^a Comparison of cases and controls (Chi-squared test). Fisher's exact test was used when the expected frequency in one of the cells was lower than 5.

^b Metal allergy was defined as a positive patch-test reaction to nickel, cobalt or chromium.

OR: odds ratio.

CI: confidence interval.