Fig 1.
Flow diagram for patient enrollment.
CRRT, continuous renal replacement therapy; ICU, intensive care unit; PCT, procalcitonin; SIAKI, sepsis-induced acute kidney injury.
Table 1.
Baseline characteristics stratified by survival at 28 days after CRRT initiation (n = 649).
Table 2.
Baseline characteristics stratified by recovery from dialysis at 28 days after CRRT initiation among the survivors (n = 332).
Table 3.
Univariable and multivariable Cox regression analyses for survival within 28 days after CRRT initiation (n = 649).
Fig 2.
HR plots of % PCT decrease for survival (A) and recovery from dialysis (B) in predefined subgroups.
In a multivariable Cox regression analysis, the % PCT decrease was an independent predictor of the 28-day survival in predefined subgroups, including age > 65 or ≤ 65 years, male or female sex, diabetes or no diabetes, oliguria or no oliguria, high SOFA group or low SOFA group, and early CRRT group or late CRRT group. In the survivors, the % PCT decrease was an independent predictor of the recovery from dialysis within 28 days after CRRT initiation across predefined subgroups. CI, confidence interval; CRRT, continuous renal replacement therapy; HR, hazard ratio; PCT, procalcitonin; SOFA, Sequential Organ Failure Assessment.
Table 4.
Univariable and multivariable Cox regression analyses for recovery from dialysis within 24 h after CRRT initiation among the survivors (n = 332).
Fig 3.
Receiver-operating characteristic curves of % PCT decrease for predicting survival (A) and recovery from dialysis (B) within 28 days after CRRT initiation in patients with SIAKI receiving CRRT.
For all the participants (n = 649), the best cut-off value of % PCT decrease for predicting survival was > 31%, with an associated sensitivity of 64.8% and specificity of 83.6% (AUC: 0.802, 95% CI: 0.769–0.832, P < 0.001, Youden index: 0.48). In the survivors (n = 332), the best cut-off of % PCT decrease for predicting the recovery from dialysis was > 69%, with an associated sensitivity of 73.5% and specificity of 80.5% (AUC: 0.825, 95% CI: 0.780–0.864, P < 0.001, Youden index: 0.54). AKI, acute kidney injury; AUC, area under the curve; CI, confidence interval; CRRT, continuous renal replacement therapy; PCT, procalcitonin; SIAKI, sepsis-induced acute kidney injury.
Fig 4.
Probability of survival (A) and recovery from dialysis (B) according to the best cut-off value of % PCT decrease in patients with SIAKI receiving CRRT.
In all the study subjects (n = 649), patients with a % PCT decrease > 31% showed a significant increase in the survival compared with those with % PCT decrease ≤ 31% (28-day survival: 80.5 vs. 30.6%, P < 0.001). In survivors (n = 332), patients with a % PCT decrease > 69% showed a significant higher rate of recovery from dialysis than those with a % PCT decrease ≤ 69% (28-day recovery rate from dialysis: 58.5% vs. 25.7%, P < 0.001). CRRT, continuous renal replacement therapy; PCT, procalcitonin; SIAKI, sepsis-induced acute kidney injury.