Risk Factors for Mortality in Patients with AKI Receiving CKRT

Critically ill patients commonly develop acute kidney injury (AKI) and may require continuous kidney replacement therapy (CKRT). Developing AKI is associated with adverse patient outcomes and increases mortality rate to approximately 60% to 80%. Identification of risk factors for outcomes in patients with AKI requiring CKRT is crucial.

Results of previous studies have suggested associations between lower ratios of serum creatinine to cystatin C with both lower muscle mass and adverse clinical outcomes in multiple disease conditions. Chan-Young Jung, MD, and colleagues conducted a retrospective cohort study to examine the association of creatinine-cystatin C ratio in patients with AKI undergoing CKRT. Results were reported in the American Journal of Kidney Diseases [2021;77(4):509-516].

The primary outcome  of interest was age- and sex-adjusted 90-day mortality after initiation of CKRT. Secondary outcomes included age- and sex-adjusted 30-day mortality, CKRT maintenance duration, KRT dependency at discharge, and length of intensive care unit (ICU) and hospital stays. The researchers evaluated 10 years of data in critically ill patients with AKI requiring CKRT who were treated at a single-center ICU setting. Cox proportional models were used to estimate the association between creatinine-cystatin C ratio and the study outcomes.

Patients at the 99-bed ICU of the Yonsei University Health System in Seoul, South Korea, who underwent CKRT between August 2009 and June 2019 were screened. Exclusion criteria were age <18 years and kidney failure requiring maintenance kidney replacement therapy prior to initiation of CKRT. The final analysis included 1588 patients.

Electronic medical records were used to retrieve laboratory and demographic data; baseline was defined as the time of initiation of CKRT. Decisions to initiate CKRT were made by attending nephrologists. Indications included volume overload, metabolic acidosis, hyperkalemia, and oliguria.

Mean age of the analysis cohort was 64.7 years, 40% (n=635) were female, 55.8% (n=886) had hypertension, and 37.6% (n=597) were being treated for diabetes mellitus. Median creatinine and cystatin C levels at baseline were 2.2 mg/L and 2.5 mg/L, respectively, with a creatinine-cystatin C range of 0.08 to 10.48.

Following stratification into quartiles based on baseline creatinine-cystatin C ratio, the proportion of women in each group was progressively lower across progressively greater creatinine-cystatin C ratios (P<.001). In groups with higher creatinine-cystatin C ratios, systolic blood pressure, diastolic blood pressure, and mean arterial pressure were significantly higher (P<.001). Higher creatinine-cystatin C ratios also tended to be associated with higher Acute Physiology and Chronic Health Evaluation II scores (P<.001). Charlson Comorbidity Index scores and 2-hour urine output prior to initiation of CKRT were comparable across quartiles.

Within 30 days of CKRT initiation, 954 patients died; the number of deaths increased to 1055 (66.4%) at 90 days post-CKRT initiation. Patients received CKRT on average for 6.3 days. Mean length of stay in the ICU was 15.7 days and mean hospital length of stay was 50.3 days. At discharge, the rate of KRT dependence was 12.3%.  There was a statistically significant trend toward lower 30- and 90-day mortality at higher creatinine-cystatin C ratios (P<.001). The length of ICU and hospital stays were significantly shorter for the higher ratio groups (P<.001 for both). Duration of CKRT and KRT dependence at discharge were comparable across creatinine-cystatin C ratio quartiles.

Kaplan-Meier curves revealed that cumulative 10- and 90-day survival probabilities were significantly lower for patients in the lowest baseline creatinine-cystatin C ratio quartile (quartile 1), compared with the other three quartiles (P<.001). Cumulative 30- and 90-day survival probability for each quartile sequentially improved with greater creatinine-cystatin C ratio.

In multivariate Cox proportional hazards regression analyses, mortality hazard ratios (HRs) for the quartiles were successively lower with greater creatinine-cystatin C ratios. For quartile 4 (highest creatinine-cystatin C ratios), HR for 30-day mortality was 0.63 (95% confidence interval [CI], 0.52-0.75); HR for 90-day mortality was 0.59 (95% CI, 0.50-0.71); both mortalities were significantly lower than in quartile 1 (P<.001). Following adjustment for confounding factors, the observed relationship between creatinine-cystatin C ratio and quartile mortality was maintained. There were no interactions between creatinine-cystatin C ratio and sex (P=.3); no sex-specific evaluations were made.

Cubic spline analyses were used to further examine the association of creatinine-cystatin C ratio with 30-day and 90-day mortality risk. There was significant decline in HRs for 30- and 90-day morality at creatinine-cystatin C ratios >0.35. The risks for 30- and 90-day mortality became progressively lower with progressively greater creatinine-cystatin C ratios. Following adjustments for confounding factors, the relationships remained.

The researchers cited some limitations to the study, including the need to use caution in interpreting the independent relationship between creatinine-cystatin C ratio and mortality due to the retrospective design of the study, the lack of data on nutritional status, and creatinine and cystatin C likely not in a steady state in the setting of AKI.

The researchers said, “In conclusion, creatinine-cystatin C ratio is associated with survival in ICU patients with AKI undergoing CKRT. This ratio may be used as a practical prognosis risk factor for ICU patients with AKI. However, further evaluations are needed for its general application.”

Takeaway Points

  1. Researchers conducted a retrospective cohort study to examine the association between creatinine-cystatin C ratio and outcomes in patients with acute kidney injury (AKI) undergoing continuous kidney replacement therapy (CKRT).
  2. The 30- and 90-day mortality rates were significantly lower for patients with higher creatinine-cystatin C ratios.
  3. Higher serum creatinine-cystatin C ratios were associated with better survival in patients with AKI in intensive care units receiving CKRT.