Meta-analysis of AKI to CKD transition in perioperative patients.

Background: Recent research shows AKI increases the risk of incident CKD. We hypothesized that perioperative AKI may confer increased risk of subsequent CKD compared to nonperioperative AKI.

Methods: A MEDLINE search was performed for "AKI, CKD, chronic renal insufficiency, surgery, and perioperative" and related terms yielded 5209 articles. One thousand sixty-five relevant studies were reviewed. One thousand six were excluded because they were review, animal, or pediatric studies. Fifty-nine studies underwent full manuscript review by two independent evaluators. Seventeen met all inclusion criteria and underwent analysis. Two-by-two tables were constructed from AKI +/- and CKD +/- data. The R package metafor was employed to determine odds ratio (OR), and a random-effects model was used to calculate weighted ORs. Leave-1-out, funnel analysis, and structured analysis were used to estimate effects of study heterogeneity and bias.

Results: Nonperioperative studies included studies of oncology, percutaneous coronary intervention, and myocardial infarction patients. Perioperative studies comprised patients from cardiac surgery, vascular surgery, and burns. There was significant heterogeneity, but risk of bias was overall assessed as low. The OR for AKI versus non-AKI patients developing CKD in all studies was 4.31 (95% CI 3.01-6.17; p < 0.01). Nonperioperative subjects demonstrated OR 3.32 for developing CKD compared to non-AKI patients (95% CI 2.06-5.34; p < 0.01) while perioperative patients demonstrated OR 5.20 (95% CI 3.12-8.66; p < 0.01) for the same event.

Conclusions: We conclude that studies conducted in perioperative and nonperioperative patient populations suggest similar risk of development of CKD after AKI.