5. Results
5.5 Risk for revision and PJI after TJA in different CKD stages (Study 4)
Median follow-up time was 5.6 years (IQR 3.5-8.1 years). In total 1,778 patients died during up resulting in 2,111 joints to discontinue up. During the follow-up 677 (3.6%) joints underwent revision procedure. Of these 162 (24%) were due to PJI (incidence 0.9%). During the first postoperative year, 299 revision procedures (incidence 1.6%) were performed and 117 of these were due to PJI (incidence 0.6%).
5.5.1 Incidence of any revision and revision for PJI in different CKD stages
In hip replacement procedures, there were no noteworthy differences between the CKD groups in the incidence of any revision or revision for PJI. Further, there were no revision procedures in Coxa among patients with CKD 4-5. In knee replacement procedures, after two years, patients with CKD stage 1 and 4-5 had greater incidence of revision due to any cause. Also, after one year, patients with CKD 4-5 had greater incidence of revision for PJI. Altogether, there were only three revision operations in the CKD 4-5 group and two of them were revisions for PJI. Incidence of death was greater in patients with more advanced CKD stage in both hip and knee replacement recipients. (Table 18)
68
Table 19. Combinations of different comorbid conditions and risk of death
Morbidities 90-day 1-year 5-year
CK D
CD CH F
DM
Number survival
% OR for
mortality 95% CI survival
% OR for
mortality 95 % CI survival
% OR for
mortality 95% CI
- - - - 15,164 99.7 reference 99.0 reference 92.1 reference
+ - - - 1,805 99.1 3.38 1.89–6.05 97.3 2.75 1.98 – 3.83 78.6 3.19 2.80–3.63
- + - - 1,272 99.4 2.39 1.12–5.12 97.9 2.19 1.45 – 3.31 82.9 2.41 2.05–2.83
+ + - - 334 98.2 6.92 2.19–16.43 94.9 5.40 3.23 – 9.04 72.4 4.46 3.47–5.47
- - - - 16,162 99.7 reference 99.0 reference 91.7 reference
+ - - - 2,000 99.0 3.72 2.21–6.24 97.2 2.79 2.06 – 3.80 78.4 3.06 2.70–3.46
- - + - 274 99.3 2.58 0.62–10.67 96.0 4.06 2.18 – 7.56 73.8 3.94 2.99–5.20
+ - + - 139 99.3 2.54 0.35–18.54 93.5 6.71 3.36 – 13.42 66.9 5.49 3.82–7.88
- - - - 15,241 99.7 reference 98.9 reference 91.7 reference
+ - - - 1,914 99.2 2.67 1.48–4.79 97.5 2.36 1.70 – 3.28 78.7 3.00 2.64–3.40
- - - + 1,195 99.7 0.85 0.26–2.74 98.7 1.19 0.70 – 2.03 87.7 1.56 1.29–1.88
+ - - + 225 96.9 10.84 4.84–24.31 92.0 8.15 4.9 – 13.51 68.7 5.05 3.77–6.77
CKD = chronic kidney disease (estimated glomerular filtration rate < 60 mL/minute/1.72 m2); CD= coronary disease; CHF = congestive heart failure; DM = diabetes mellitus;
OR = odds ratio
5.4.4. Mortality of AKI patients (Study 1)
A Kaplan-Meier survival plot shows remarkably poorer survival for those patients who had postoperative AKI after joint replacement. At ten years, over 70% of the non-AKI patients were alive, while only 36% of AKI patients survived. (Figure 5)
Figure 5. Survival of AKI patients
69
5.5 Risk for revision and PJI after TJA in different CKD stages (Study 4)
Median follow-up time was 5.6 years (IQR 3.5-8.1 years). In total 1,778 patients died during up resulting in 2,111 joints to discontinue up. During the follow-up 677 (3.6%) joints underwent revision procedure. Of these 162 (24%) were due to PJI (incidence 0.9%). During the first postoperative year, 299 revision procedures (incidence 1.6%) were performed and 117 of these were due to PJI (incidence 0.6%).
5.5.1 Incidence of any revision and revision for PJI in different CKD stages
In hip replacement procedures, there were no noteworthy differences between the CKD groups in the incidence of any revision or revision for PJI. Further, there were no revision procedures in Coxa among patients with CKD 4-5. In knee replacement procedures, after two years, patients with CKD stage 1 and 4-5 had greater incidence of revision due to any cause. Also, after one year, patients with CKD 4-5 had greater incidence of revision for PJI. Altogether, there were only three revision operations in the CKD 4-5 group and two of them were revisions for PJI. Incidence of death was greater in patients with more advanced CKD stage in both hip and knee replacement recipients. (Table 18)
70
5.5.2 Cumulative probability after considering death as a competing risk
Figure 6 illustrates the probabilities for any revision, revision for PJI and death according to the CIF, whereas Table 20 illustrates the probability of any revision according to the CIF at five and eight years after TJA. After hip replacement, the probability of any revision or revision for PJI was similar between CKD stages. Also, after knee replacement, the probability of revision for PJI did not differ between CKD stages. In contrast, after knee replacement, patients with CKD stage 1 had greater probability of any revision than patients with CKD 2 or 3. Fig 6b shows similar results across the entire follow-up. Patients with CKD 4-5 who underwent knee replacement had the highest probability of any revision and revision for PJI, but the results were not statistically significant. Probability of death increased gradually with every advancement of CKD stage. (Table 20, Figure 6)
Table 20. Cumulative probability (95% CI) of different outcomes according to the CIF
All-cause Revision PJI Death
5 years 8 years 5 years 8 years 5 years 8 years
Hips CKD 1 3.3 (2.6–4.1) 4.1 (3.3–5.0) 0.7 (0.4–1.0) 0.7 (0.4–1.0) 4.1 (3.3–4.9) 7.4 (6.0–8.7) CKD 2 3.8 (3.2–4.3) 4.5 (3.8–5.2) 0.8 (0.5–1.1) 1.1 (0.7–1.4) 7.8 (6.9–8.6) 16.9 (15.3–18.5) CKD 3 3.7 (2.4–5.1) 5.2 (3.3–7.0) 0.7 (0.1–1.3) 0.7 (0.1–1.3) 14.9 (12.2–17.6) 30.8 (26.3–35.2)
CKD 4–5 none none none none 29.2 (13.3–45.0) 49.6 (28.4–70.7)
Knees CKD 1 4.4 (3.7– 5.1) 5.5 (4.7– 6.3) 1.0 (0.7– 1.3) 1.2 (0.8– 1.5) 3.5 (2.9– 4.2) 6.7 (5.6–7.7) CKD 2 2.6 (2.2–3.0) 3.1 (2.6–3.6) 0.7 (0.5–0.9) 0.9 (0.7–1.2) 6.4 (5.7–7.0) 14.5 (13.3–15.7) CKD 3 2.0 (1.2–2.8) 2.8 (1.7–3.9) 0.5 (0.1–0.9) 0.5 (0.1–0.9) 15.4 (13.1–17.8) 30.5 (26.8–34.1) CKD 4–5 7.9 (0.0–16.6) 7.9 (0.0–16.6) 4.7 (0.0–11.2) 4.7 (0.0–11.2) 43.0 (25.7–60.4) 74.8 (51.2–98.4)
71
Figure 6. Probabilities of all-cause revisions, revisions for PJI, and deaths according to (the) CIF
70
5.5.2 Cumulative probability after considering death as a competing risk
Figure 6 illustrates the probabilities for any revision, revision for PJI and death according to the CIF, whereas Table 20 illustrates the probability of any revision according to the CIF at five and eight years after TJA. After hip replacement, the probability of any revision or revision for PJI was similar between CKD stages. Also, after knee replacement, the probability of revision for PJI did not differ between CKD stages. In contrast, after knee replacement, patients with CKD stage 1 had greater probability of any revision than patients with CKD 2 or 3. Fig 6b shows similar results across the entire follow-up. Patients with CKD 4-5 who underwent knee replacement had the highest probability of any revision and revision for PJI, but the results were not statistically significant. Probability of death increased gradually with every advancement of CKD stage. (Table 20, Figure 6)
Table 20. Cumulative probability (95% CI) of different outcomes according to the CIF
All-cause Revision PJI Death
5 years 8 years 5 years 8 years 5 years 8 years
Hips CKD 1 3.3 (2.6–4.1) 4.1 (3.3–5.0) 0.7 (0.4–1.0) 0.7 (0.4–1.0) 4.1 (3.3–4.9) 7.4 (6.0–8.7) CKD 2 3.8 (3.2–4.3) 4.5 (3.8–5.2) 0.8 (0.5–1.1) 1.1 (0.7–1.4) 7.8 (6.9–8.6) 16.9 (15.3–18.5) CKD 3 3.7 (2.4–5.1) 5.2 (3.3–7.0) 0.7 (0.1–1.3) 0.7 (0.1–1.3) 14.9 (12.2–17.6) 30.8 (26.3–35.2)
CKD 4–5 none none none none 29.2 (13.3–45.0) 49.6 (28.4–70.7)
Knees CKD 1 4.4 (3.7– 5.1) 5.5 (4.7– 6.3) 1.0 (0.7– 1.3) 1.2 (0.8– 1.5) 3.5 (2.9– 4.2) 6.7 (5.6–7.7) CKD 2 2.6 (2.2–3.0) 3.1 (2.6–3.6) 0.7 (0.5–0.9) 0.9 (0.7–1.2) 6.4 (5.7–7.0) 14.5 (13.3–15.7) CKD 3 2.0 (1.2–2.8) 2.8 (1.7–3.9) 0.5 (0.1–0.9) 0.5 (0.1–0.9) 15.4 (13.1–17.8) 30.5 (26.8–34.1) CKD 4–5 7.9 (0.0–16.6) 7.9 (0.0–16.6) 4.7 (0.0–11.2) 4.7 (0.0–11.2) 43.0 (25.7–60.4) 74.8 (51.2–98.4)
71
Figure 6. Probabilities of all-cause revisions, revisions for PJI, and deaths according to (the) CIF
72
5.5.3 Results after considering confounders
Table 21 illustrates the probability of all-cause revision at five and eight years after stratification by diabetes and BMI. Again, in hip replacement patients there was a similar probability of all-cause revision in different CKD stages. In knee replacement patients, as in the unstratified analysis, only CKD 1 patients with no diabetes or BMI less than 30kg/m2 had greater probability for all-cause revision. (Table 21)
Table 21. Cumulative probability of all-cause revisions in CIF at 5 and 8 years according to diabetes and BMI
Joint Confounder Time CKD 1 CKD 2 CKD 3 CKD 4-5
Years Cumulative probability % (95% CI)
Knee DM- 5 4.5 (3.7–5.2) 2.5 (2.0–2.9) 2.2 (1.2–3.1) 12.1 (-1.3–25.6)
DM+ 5 4.4 (2.8–6.0) 2.8 (1.8–3.9) 1.6 (0.0–3.1) none
DM- 8 5.7 (4.7–6.6) 3.1 (2.6–3.6) 3.2 (1.8–4.5) 12.1 (-1.3–25.6)
DM+ 8 4.7 (3.0–6.4) 2.8 (1.8–3.9) 1.6 (0.0–3.1) none
Knee BMI <30 5 5.0 (4.0–6.0) 2.2 (1.7–2.7) 2.3 (1.1–3.4) 3.3 (-3.2–9.9) BMI ≥30 5 3.9 (2.9–4.8) 3.1 (2.4–3.8) 1.7 (0.5–2.8) 17.3 (-6.2–40.8) BMI <30 8 6.2 (5.0–7.4) 2.8 (2.2–3.4) 3.0 (1.5–4.5) 3.3 (-3.2–9.9) BMI ≥30 8 4.7 (3.6–5.8) 3.6 (2.8–4.4) 2.6 (0.9–4.3) none
Hip DM- 5 3.4 (2.6–4.1) 3.6 (3.0–4.2) 4.5 (2.9–6.1) none
DM+ 5 3.2 (1.4–5.1) 4.6 (2.9–6.4) 1.2 (-0.5–2.9) none
DM- 8 4.0 (3.1–4.9) 4.3 (3.6–5.1) 5.9 (3.7–8.0) none
DM+ 8 4.7 (2.0–7.5) 5.3 (3.1–7.4) 2.5 (-0.5–5.5) none
Hip BMI <30 5 3.1 (2.3–3.9) 3.5 (2.8–4.1) 3.9 (2.2–5.5) none
BMI ≥30 5 3.8 (2.4–5.2) 4.5 (3.3–5.7) 3.5 (1.2–5.7) none BMI <30 8 3.5 (2.6–4.4) 4.1 (3.3–4.9) 5.1 (2.9–7.3) none Hip BMI ≥30 8 5.6 (3.6–7.6) 5.5 (3.9–7.0) 5.2 (1.9–8.5) none
73
In the multivariable Fine and Gray model, CKD stage was not associated with PJI (Table 22)
Table 22. Subdistribution hazards of revision for PJI
Knees Hips
Risk factor Subdistribution HR Subdistribution HR
CKD 1 reference reference
CKD 2 0.854 (0.56-1.30) 1.23 (0.69-2.19) CKD 3 0.533 (0.23-1.26) 1.06 (0.42-2.68)
CKD 4-5 * no patients
Analysis adjusted with diabetes and BMI * Excluded from the analysis due to unproportional hazards
72
5.5.3 Results after considering confounders
Table 21 illustrates the probability of all-cause revision at five and eight years after stratification by diabetes and BMI. Again, in hip replacement patients there was a similar probability of all-cause revision in different CKD stages. In knee replacement patients, as in the unstratified analysis, only CKD 1 patients with no diabetes or BMI less than 30kg/m2 had greater probability for all-cause revision. (Table 21)
Table 21. Cumulative probability of all-cause revisions in CIF at 5 and 8 years according to diabetes and BMI
Joint Confounder Time CKD 1 CKD 2 CKD 3 CKD 4-5
Years Cumulative probability % (95% CI)
Knee DM- 5 4.5 (3.7–5.2) 2.5 (2.0–2.9) 2.2 (1.2–3.1) 12.1 (-1.3–25.6)
DM+ 5 4.4 (2.8–6.0) 2.8 (1.8–3.9) 1.6 (0.0–3.1) none
DM- 8 5.7 (4.7–6.6) 3.1 (2.6–3.6) 3.2 (1.8–4.5) 12.1 (-1.3–25.6)
DM+ 8 4.7 (3.0–6.4) 2.8 (1.8–3.9) 1.6 (0.0–3.1) none
Knee BMI <30 5 5.0 (4.0–6.0) 2.2 (1.7–2.7) 2.3 (1.1–3.4) 3.3 (-3.2–9.9) BMI ≥30 5 3.9 (2.9–4.8) 3.1 (2.4–3.8) 1.7 (0.5–2.8) 17.3 (-6.2–40.8) BMI <30 8 6.2 (5.0–7.4) 2.8 (2.2–3.4) 3.0 (1.5–4.5) 3.3 (-3.2–9.9) BMI ≥30 8 4.7 (3.6–5.8) 3.6 (2.8–4.4) 2.6 (0.9–4.3) none
Hip DM- 5 3.4 (2.6–4.1) 3.6 (3.0–4.2) 4.5 (2.9–6.1) none
DM+ 5 3.2 (1.4–5.1) 4.6 (2.9–6.4) 1.2 (-0.5–2.9) none
DM- 8 4.0 (3.1–4.9) 4.3 (3.6–5.1) 5.9 (3.7–8.0) none
DM+ 8 4.7 (2.0–7.5) 5.3 (3.1–7.4) 2.5 (-0.5–5.5) none
Hip BMI <30 5 3.1 (2.3–3.9) 3.5 (2.8–4.1) 3.9 (2.2–5.5) none
BMI ≥30 5 3.8 (2.4–5.2) 4.5 (3.3–5.7) 3.5 (1.2–5.7) none BMI <30 8 3.5 (2.6–4.4) 4.1 (3.3–4.9) 5.1 (2.9–7.3) none Hip BMI ≥30 8 5.6 (3.6–7.6) 5.5 (3.9–7.0) 5.2 (1.9–8.5) none
73
In the multivariable Fine and Gray model, CKD stage was not associated with PJI (Table 22)
Table 22. Subdistribution hazards of revision for PJI
Knees Hips
Risk factor Subdistribution HR Subdistribution HR
CKD 1 reference reference
CKD 2 0.854 (0.56-1.30) 1.23 (0.69-2.19) CKD 3 0.533 (0.23-1.26) 1.06 (0.42-2.68)
CKD 4-5 * no patients
Analysis adjusted with diabetes and BMI * Excluded from the analysis due to unproportional hazards
74