Urea and creatinine changes in critical illness and recovery: the natural history of muscle and kidney injury

Abstract

Acute kidney injury affects over half of patients admitted to intensive care. However, the estimation of kidney function during and after critical illness, reliant on changes in urea and creatinine, is poorly understood. Development of chronic kidney disease contributes to cardiovascular disease, mortality, and poorer quality of life. I hypothesised the effects of critical illness, specifically muscle wasting, impacted our interpretation of commonly used kidney biomarkers which has implications for the diagnosis of varying kidney function and catabolism. Several analysis methods were used to address the complexities of longitudinal data in the heterogenous intensive care population. Focusing on major trauma patients reduced the impact of underlying heterogeneity on selected analyses. Chapter 3 examined the burden of acute kidney injury in major trauma requiring ICU admission using a systematic review of observational studies. From 17 included articles of 24,267 trauma patients the incidence of AKI was 20.4% (95% confidence interval, 16.5-24.9). Chapter 4 explored longitudinal trends of urea, creatinine, and muscle mass in a retrospective analysis of two large datasets of major trauma patients in intensive care. Falls in creatinine and increases in urea resulting in an elevated urea- to-creatinine ratio were associated with catabolism. Chapter 5 expanded the analysis to the general critical care population. Through the application of joint models, longitudinal urea-to-creatinine ratio changes (a surrogate of catabolism) were associated with increased mortality (hazard ratio of 2.15, 95% credible interval 1.66– 2.82, for a two-fold greater urea-to-creatinine ratio). Finally, Chapter 6 presents a prospective observational study that examined the impact of catabolism on diagnosis of kidney function. By ICU discharge the data suggested a difference between creatinine and cystatin C (a kidney function biomarker not affected by catabolic changes of critical illness) estimated glomerular filtration rate was 33 ml/min/1.73m2 (95% credible interval 24–41). Prevalent catabolism resulted in creatinine systematically over-estimating kidney function, potentially resulting in significant under-diagnosis of persistent kidney disease.