Fig 1.
Patient population and exclusions from the study.
Table 1.
ICD codes included in analysis.
Table 2.
Characteristics of final groups of primary hip replacement (n = 403,881) for analysis by bearing type.
Fig 2.
Cumulative risk of a new diagnosis of any type of cancer following primary hip replacement by bearing type with numbers at risk shown.
Table 3.
Hazard Rate Ratios (HRR) for being diagnosed with any type of cancer according to bearing type following primary hip replacement in men (n = 163,111; 14,825 were subsequently diagnosed with cancer) and women (n = 240,765; 17,040 were subsequently diagnosed with cancer).
Table 4.
Sub-hazard Rate Ratios (SHR) for being diagnosed with any type of cancer according to bearing type following primary hip replacement in men (n = 163,111; 14,825 were subsequently diagnosed with cancer; 9,524 ‘competing’ deaths) and women (n = 240,765; 17,040 were subsequently diagnosed with cancer; 13,999 ‘competing’ deaths).
Table 5.
Hazard Rate Ratios (HRR) with: (i) 1:1 matching for being diagnosed with any type of cancer following stemmed MoM in men (a: n = 17,450; 1,879 were subsequently diagnosed with cancer) and women (c: n = 16,930; 1,437 were subsequently diagnosed with cancer), and MoM hip resurfacings in men (b: n = 23,722; 1,529 were subsequently diagnosed with cancer) and women (d: n = 10,672; 704 were subsequently diagnosed with cancer).
(ii) 1:4 matching for being diagnosed with any type of cancer following stemmed MoM in men (a: n = 41,000; 4,462 were subsequently diagnosed with cancer) and women (c: n = 41,815; 3,704 were subsequently diagnosed with cancer), and MoM hip resurfacings in men (b: n = 38,350; 2,684 were subsequently diagnosed with cancer) and women (d: n = 20,425; 1,375 were subsequently diagnosed with cancer).
Table 6.
Cox proportional hazards regression models of the relative effects on time to cancer diagnosis of the bearing types of a second (contralateral) primary hip replacement, for patients who received stemmed non-MoM THR (other) at first primary in men (n = 141,179; 13,076 were subsequently diagnosed with cancer) and women (n = 259,858; 15,958 were subsequently diagnosed with cancer).
Table 7.
Cox proportional hazards regression models of the risk of being diagnosed with a haematological cancer following primary hip replacement by bearing type for men (n = 163,111; 2,109 were subsequently diagnosed with a haematological cancer) and women (n = 240,765; 2,953 were subsequently diagnosed with a haematological cancer).
Table 8.
Cox proportional hazards regression models of the risk of being diagnosed with malignant melanoma following primary hip replacement by bearing type for men (n = 163,111; 576 were subsequently diagnosed with malignant melanoma) and women (n = 240,765; 596 were subsequently diagnosed with malignant melanoma).
Table 9.
Cox proportional hazards regression models of the risk of being diagnosed with a urinary tract cancer following primary hip replacement by bearing type for men (n = 163,111; 1,849 were subsequently diagnosed with a urinary tract cancer) and women (n = 240,765; 1,078 were subsequently diagnosed with a urinary tract cancer).
Table 10.
Cox proportional hazards regression models of the risk of being diagnosed with prostate cancer following primary hip replacement by bearing type for men (n = 163,111; 4,061 were subsequently diagnosed with prostate cancer).