Systematic Review and Meta-analysis of COVID-19 and Kidney Transplant Recipients, the South West London Kidney Transplant Network Experience

Abstract

Introduction: There is paucity of literature comparing outcomes of kidney transplant patients with COVID-19 to that of dialysis and waitlisted patients. This report describes our data, provides comparative analysis, together with a meta-analysis of published studies, and describes our protocols to restart the transplant program.

Methods: Data were analyzed on kidney transplant, dialysis, and waitlisted patients tested positive for SARS-CoV-2 (nasopharyngeal swab polymerase chain reaction [PCR] test) between March 1, 2020, and June 30, 2020, together with a meta-analysis of 16 studies.

Results: Twenty-three of 1494 kidney transplant patients tested positive for SARS-CoV-2 compared with 123 of 1278 hemodialysis patients (1.5% vs. 9.6%, P < 0.001) and 12 of 253 waitlisted patients (1.5% vs. 4.7%, P = 0.002). Nineteen patients required hospital admission, of whom 6 died and 13 developed AKI. The overall case fatality ratio was 26.1% compared with patients on hemodialysis (27.6%, P = 0.99) and waitlisted patients (8.3%, P = 0.38). Within our entire cohort, 0.4% of transplant patients died compared with 0.4% of waitlisted patients and 2.7% of hemodialysis patients. Patients who died were older (alive [median age 71 years] vs. dead [median age 59 years], P = 0.01).In a meta-analysis of 16 studies, including ours, the pooled case fatality ratio was 24% (95% confidence interval [CI] 19%, 28%); AKI proportion in 10 studies was 50% (95% CI 45%, 56%), with some evidence against no heterogeneity between studies (P = 0.02).

Conclusions: From our cohort of transplant patients, a significantly lower proportion of patients contracted COVID-19 compared with waitlisted and dialysis patients. The case fatality ratio was comparable to that of the dialysis cohort and to a pooled case fatality ratio from a meta-analysis of 16 studies. The pooled AKI ratio in the meta-analysis was similar to our results.

Graphical abstract

Figure 1

Meta-analyses of COVID-19 in transplant patients: case fatality ratio. (a) The pooled case fatality ratio was 24% (95% CI 19%, 28%). There was moderate heterogeneity between the studies (I² = 51.5% [variation in effect size (ES) attributable to heterogeneity], heterogeneity χ² = 30.90 [df = 15], P = 0.01). The New York Montefiore 2 Study—the third most influential in this analysis—exhibited a case fatality ratio of 38% (95% CI 29%, 48%), well above the pooled estimate of 24% (95% CI 19%, 28%). (b) We then analyzed 14 studies excluding this study, and with this analysis, the I² drops to 34.3% with a P value = 0.09, consistent with the null hypothesis of not much heterogeneity between studies. The pooled case fatality ratio in this analysis was 22% (95% CI 18%, 27%).

Figure 2

Meta-analyses of COVID-19 in transplant patients: (a) AKI, all stages. The pooled proportion of AKI was 50% (95% CI 45%, 56%). There was no significant heterogeneity between the studies, χ² = 11.02 (df = 9), P = 0.27; I² (variation in effect size [ES] attributable to heterogeneity) = 18.37%. Therefore, the pooled proportion of AKI is 50% (95% CI 45%, 56%). (b, c) AKI, stage 3 / RRT requirement. The pooled proportion of severe AKI (stage 3 / requiring RRT) was 18% (95% CI 12%, 25%) (b). However, there was a significant heterogeneity: I² = 66.27%, P < 0.001. Reanalysis after removal of the Bologna study that showed a high stage 3 AKI percentage of 45% yielded results with pooled stage 3 / RRT-requiring AKI estimate of 16% (95% CI 10%, 22%) but the heterogeneity, although improved, remained significant, with I² = 56.96%, P < 0.02 (c). Therefore, it appears from these analyses that the pooled proportion of severe AKI is 16% to 18%.

Figure 2

Meta-analyses of COVID-19 in transplant patients: (a) AKI, all stages. The pooled proportion of AKI was 50% (95% CI 45%, 56%). There was no significant heterogeneity between the studies, χ² = 11.02 (df = 9), P = 0.27; I² (variation in effect size [ES] attributable to heterogeneity) = 18.37%. Therefore, the pooled proportion of AKI is 50% (95% CI 45%, 56%). (b, c) AKI, stage 3 / RRT requirement. The pooled proportion of severe AKI (stage 3 / requiring RRT) was 18% (95% CI 12%, 25%) (b). However, there was a significant heterogeneity: I² = 66.27%, P < 0.001. Reanalysis after removal of the Bologna study that showed a high stage 3 AKI percentage of 45% yielded results with pooled stage 3 / RRT-requiring AKI estimate of 16% (95% CI 10%, 22%) but the heterogeneity, although improved, remained significant, with I² = 56.96%, P < 0.02 (c). Therefore, it appears from these analyses that the pooled proportion of severe AKI is 16% to 18%.