Review Papers Against Their Own Results

 Long Explanation Copied

Study Design & Key Findings

Hamed et al. analyzed 801 deceased-donor kidney-only transplants (2002–2012) at Cambridge, finding EGL in 50/801 (6.2%). Risk factors by multivariate logistic regression were DCD donor type (OR 2.88, p=0.006), ECD donor type (OR 4.22, p=0.010), donor age (OR 1.03/year, p=0.044), and recipient history of venous thrombosis (OR 4.91, p=0.001). EGL conferred 12.28× increased odds of death within 1 year (p<0.001), yet long-term survival after EGL still exceeded remaining untransplanted on the waiting list — risk equalized at 4 months (vs. non-transplanted) and 23 months (vs. all waitlisted) post-transplant Hamed et al. AJT 2015 EGL study

EGL causes were predominantly non-immunologic: PNF (20, 2.5%), thrombosis (18, 2.2%), haemorrhage (6, 0.7%), acute rejection (3, 0.4%). Consistent with the donor-quality hypothesis, ECD kidneys had worse 5-year graft survival (83.7% vs 91.4% SCD, p=0.017), while DCD vs DBD long-term graft survival was similar (86.4% vs 90.0%, p=0.257), suggesting DCD's effect is front-loaded in the first year EGL causes and long-term graft survival data

Critical Assessment

Strengths: Prospective database, dual donor classification (DBD/DCD and SCD/ECD analyzed separately due to collinearity — 60.4% of ECD donors were also DCD), explicit EGL definitions with biopsy confirmation, and a decision-relevant waiting-list comparator analysis. The thrombosis-history finding (7/50 EGL patients; OR 4.91) is a novel, actionable risk factor, consistent with thrombosis accounting for up to one-third of early losses in prior series Renal vein thrombosis case series. The waiting-list comparison aligns with the broader view that marginal kidneys often still beat waiting Deceased donor characteristics narrative review.

Limitations: Single-center, non-randomized, 2002–2012 era; the authors themselves acknowledge selection bias in the transplanted-vs-waitlisted comparison (more optimal patients get transplanted). EGL events are few (n=50), so the 12.28× mortality OR has a wide CI (4.60–32.91); only 5 of 16 post-EGL deaths were directly attributable to surgical complications, so the mortality mechanism remains partly unexplained (authors attribute the rest to return to dialysis + immunosuppression exposure). Donor age significance (OR 1.03) vanishes in the SCD/ECD model due to collinearity with ECD — a collinearity the authors handled but not eliminated. Retrospective re-transplant survival comparison (85.7% vs 90.8%, p=0.635) is underpowered.

Contextualizing Evidence

The finding that DCD confers short-term but not long-term graft risk matches UK registry data showing no 5-year graft survival difference between first-time DCD (n=739) and DBD (n=6,759) recipients (HR 1.01, 95% CI 0.83–1.19) UK DCD vs DBD registry comparison. Donor AKI meta-analysis (RR 1.76 for DGF) supports the mechanistic link between warm ischemic injury and early non-function implicated in PNF Donor AKI and DGF meta-analysis.

Bottom line: Moderately strong single-center observational evidence that suboptimal donor kidneys raise EGL risk but that even EGL is, in expectation, better than staying on the waiting list — with the caveat that the counterfactual comparison is unadjusted for selection bias. Confidence: moderate. Falsification would require a multi-center analysis showing EGL risk factors do not replicate, or that waitlisted patients fare as well as EGL patients.