Assessment of severity scoring systems for predicting mortality in critically ill patients receiving continuous renal replacement therapy

Hyunmyung Park; Jihyun Yang; Byung Chul Chun

doi:10.1371/journal.pone.0286246

. 2023 May 25;18(5):e0286246. doi: 10.1371/journal.pone.0286246

Assessment of severity scoring systems for predicting mortality in critically ill patients receiving continuous renal replacement therapy

Hyunmyung Park ¹, Jihyun Yang ², Byung Chul Chun ^3,^*

Editor: Giuseppe Remuzzi⁴

PMCID: PMC10212150 PMID: 37228073

Abstract

The incidence of acute kidney injury (AKI) is increasing every year and many patients with AKI admitted to the intensive care unit (ICU) require continuous renal replacement therapy (CRRT). This study compared and analyzed severity scoring systems to assess their suitability in predicting mortality in critically ill patients receiving CRRT. Data from 612 patients receiving CRRT in four ICUs of the Korea University Medical Center between January 2016 and November 2018 were retrospectively collected. The mean age of all patients was 67.6 ± 14.8 years, and the proportion of males was 59.6%. The endpoints were in-hospital mortality and 7-day mortality from the day of CRRT initiation to the date of death. The Program to Improve Care in Acute Renal Disease (PICARD), Demirjian’s, Acute Physiology and Chronic Health Evaluation (APACHE) II, Simplified Acute Physiology Score (SAPS) 3, Sequential Organ Failure Assessment (SOFA), Multiple Organ Dysfunction Score (MODS), and Liano’s scores were used to predict mortality. The in-hospital and 7-day mortality rates in the study population were 72.7% and 45.1%, respectively. The area under the receiver operator characteristic curve (AUROC) revealed the highest discrimination ability for Demirjian’s score (0.770), followed by Liano’s score (0.728) and APACHE II (0.710). The AUROC curves for the SAPS 3, MODS, and PICARD were 0.671, 0.665, and 0.658, respectively. The AUROC of Demirjian’s score was significantly higher than that of the other scores, except for Liano’s score. The Hosmer-Lemeshow test on Demirjian’s score showed a poor fit in our analysis; however, it was more acceptable than general severity scores. Kidney-specific severity scoring systems showed better performance in predicting mortality in critically ill patients receiving CRRT than general severity scoring systems.

Introduction

Acute kidney injury (AKI) occurs in 15%–38% of hospitalized patients, with an in-hospital mortality rate of 23.9%–60.3%, especially in critically ill patients, with a high incidence of up to 74.5% [1–6]. Furthermore, AKI causes additional complications and aggravates the underlying disease, leading to increased hospital stay duration and medical costs [1, 2].

Despite continued progress in medical technology, the incidence of AKI is increasing every year [7] and has become a major public health concern [8]. Among the patients with AKI admitted to the intensive care unit (ICU), 72.5% required renal replacement therapy and 80% received continuous renal replacement therapy (CRRT) [6], primarily because this therapy is hemodynamically more stable than intermittent hemodialysis and fluid balance can be easily controlled [9].

However, CRRT has the disadvantages of high cost and requiring skilled personnel due to the risks of arrhythmia, bleeding, and hypotension [9–11]. Therefore, it is necessary to assess patient severity for predicting prognosis and identifying meaningful information necessary for medical staff to discuss and make correct decisions about patient prognosis, and for providing future treatment directions to patients and care givers [12]. Moreover, predicting the mortality rate of patients admitted to the ICU is critical for assessing the severity of the disease and adjudicating the value of new treatments, interventions, and health care policies [13]. Estimates of mortality risk can be useful for the efficient allocation of resources and judgment of treatment adequacy in medical institutions by comparing actual and expected outcomes [14, 15].

The severity scoring system can be divided into assessing the overall health status and measuring severity by focusing on specific organs. Various scoring systems have been developed to predict disease prognosis [16–22]. Several studies have used severity scoring systems to effectively apply CRRT according to the acuity of illness parameters [23, 24]. Additionally, to improve the quality of CRRT, such as the optimal start time of CRRT [25, 26], severity scores are used for providing the objectivity and reliability of the study with population stratification and balanced randomization to ensure that disease severity does not affect the statistical outcome [27]. Although other studies have compared the predictive abilities of severity scoring systems in patients with AKI, most studies included all patients diagnosed with AKI who received renal replacement therapy, such as intermittent hemodialysis, and only few focused solely on CRRT [28–34].

This study aimed to evaluate the predictive ability of severity scoring systems for mortality in critically ill patients receiving CRRT. This study compared and analyzed the Acute Physiology and Chronic Health Evaluation (APACHE) II score, Simplified Acute Physiology Score (SAPS) 3, Sequential Organ Failure Assessment (SOFA) score, and Multiple Organ Dysfunction Score (MODS), which are general severity scoring systems predicting mortality, and Liano’s, Program to Improve Care in Acute Renal Disease (PICARD), and Demirjian’s scores, which are kidney-specific severity scoring systems. The results of this study could be used as a basis for selecting scoring systems suitable for severity assessment in patients with AKI receiving CRRT, and for developing a new severity scoring system.

Materials and methods

Study population

From January 2016 to November 2018, patients who received CRRT in four ICUs at the Korea University of Medicine were selected. The target group included patients with AKI and chronic kidney disease (CKD) who did not undergo dialysis. Patients who had previously undergone hemodialysis or peritoneal dialysis for end-stage renal disease or who had received a kidney transplant were excluded. Of the 768 patients who received CRRT during this period, 612 were included in the final analysis, excluding 119 patients who had previously undergone hemodialysis, 23 who had undergone peritoneal dialysis, and 14 who had kidney transplants. In cases where CRRT was performed several times during hospitalization, the clinical symptoms and diagnostic test results at the time of initial treatment were investigated.

This study was approved by the Institutional Review Board of the Korea University Anam Hospital (approval number: 2018AN0415). Informed consent was waived by the board as this study was conducted retrospectively and the data were de-identified prior to analysis. All study methods were performed in accordance with relevant guidelines and regulations.

Data

Clinical data of patients receiving CRRT were retrospectively collected from electronic medical records. The patients’ age, sex, in-hospital mortality, survival at 7 days from the initiation of CRRT, comorbidities, reason for CRRT, mean arterial pressure (MAP) at the initiation of treatment, use of vasopressors, mechanical ventilation, and laboratory findings, including serum hemoglobin, serum creatinine, albumin, arterial blood gas analysis, and C-reactive protein levels, were investigated.

The scores of APACHE II, SAPS 3, SOFA, MODS, Liano’s, PICARD, and Demirjian’s were calculated and used to predict mortality. Severity scoring was performed at the initiation of CRRT, and all variables for severity scores were collected within 24 h prior to the initiation of CRRT. The endpoints were in-hospital and 7-day mortality according to the duration from the day of CRRT initiation to the date of death.

Statistical analysis

The general characteristics of the participants were recorded using mean value and standard deviation, and the area under the receiver operating characteristic curve (AUROC) of each severity score was calculated to assess discrimination among the severity scoring systems.

Calibration of the severity score was assessed using the Hosmer-Lemeshow Goodness-of-Fit test. Data analysis was performed using IBM SPSS statistical software, version 23 (SPSS Inc., Chicago, IL, USA), and ROC comparisons were performed using the MedCalc statistical software (MedCalc, Ostend, Belgium). The statistical significance level was set at p-value >0.05.

Results

Baseline characteristics and mortality of the study population

A total of 612 participants were enrolled in the study. The mean age of all participants was 67.6 ± 14.8 years and 59.6% were males (Table 1). The in-hospital mortality rate was 72.7% and 7-day mortality was 45.1%. Patients with CKD accounted for 11.9% of the total study population, and there was no significant difference in the proportions of survivors and non-survivors. There were no significant differences between survivors and non-survivors in terms of mean age, sex, and frequency of AKI causes. Table 2 shows the clinical test results and acute physiology of the study population at the time of CRRT initiation. pH, serum albumin, and platelet count in non-survivors were significantly lower than those in survivors (p < 0.01 for all).

Table 1. Demographics and clinical characteristics of the study population.

Parameter	All patients n = 612 (%)	Survivors n = 167 (%)	Non-survivors n = 445 (%)	p-value
Age (years)	67.6 ± 14.8	67.4 ± 14.9	67.7 ± 14.7	0.851^*
Sex
Male	365 (59.6)	95 (56.9)	270 (60.7)	0.406⁺
Female	247 (40.4)	72 (43.1)	175 (39.3)	0.406⁺
Etiology of AKI
Sepsis	400 (65.4)	108 (64.7)	292 (65.6)	0.849⁺
Nephrotoxic	226 (36.9)	54 (32.3)	172 (38.7)	0.159⁺
Ischemic	204 (33.3)	52 (31.1)	152 (34.2)	0.502⁺
Others	62 (10.1)	18 (10.8)	44 (9.9)	0.764⁺
CKD history	73 (11.9)	24 (11.4)	49 (11.0)	0.264
Comorbidities
Diabetes mellitus	231 (37.7)	81 (48.5)	150 (33.7)	0.001⁺
Hypertension	336 (54.9)	99 (59.3)	237 (53.3)	0.202⁺
Heart Failure	91 (14.9)	27 (16.2)	64 (14.4)	0.610⁺
Coronary artery disease	94 (15.4)	24 (14.4)	70 (15.7)	0.708⁺
COPD	13 (2.1)	2 (1.2)	11 (2.5)	0.530⁺
Cancer	111 (18.1)	21 (12.6)	90 (20.2)	0.034⁺
Liver disease	100 (16.3)	18 (10.8)	82 (18.4)	0.027⁺

Open in a new tab

Continuous data are presented as mean ± SD, and categorial data as number of patients (%). AKI, acute kidney injury; COPD, chronic obstructive pulmonary disease.

*p-value by Student’s t-test

+p-value by Fisher’s exact test

Table 2. Clinical test results and acute physiology of study population at continuous renal replacement therapy (CRRT) initiation.

Parameter	All patients (n = 612) mean ± SD	Survivors (n = 167) mean ± SD	Non-survivors (n = 445) mean ± SD	p-value
Serum creatinine (mg/dL)	3.1 ± 1.9	3.8 ± 2.4	2.9 ± 1.8	<0.001^**
BUN (mg/dL)	57.8 ± 33.8	61.8 ± 37.4	56.5 ± 32.6	0.236^**
Urine volume (mL/day)	759 ± 1003	1169.9 ± 1244.9	634.0 ± 898.4	<0.001^**
Hb (g/dL)	9.9 ± 2.2	10.0 ± 2.1	9.9 ± 2.3	0.615^*
WBC count (x10 ³ /μL)	14.5 ± 9.9	15.3 ± 9.2	14.4 ± 10.3	0.092^**
Lactic acid (mmol/L)	7.3 ± 6.1	4.9 ± 4.9	8.2 ± 6.3	<0.001^**
CRP (mg/L)	125.6 ± 112.3	117.7 ± 102.8	128.9 ± 115.6	0.373^**
Arterial pH	7.2 ± 0.14	7.4 ± 0.1	7.3 ± 0.1	<0.001^**
Arterial bicarbonate (mmol/L)	16.5 ± 5.6	17.3 ± 5.9	16.2 ± 5.5	0.039^*
Total bilirubin (mg/dL)	3.4 ± 7.1	3.2 ± 8.3	3.5 ± 6.5	<0.001^**
INR	1.7 ± 1.1	1.5 ± 0.9	1.8 ± 1.1	<0.001^**
Platelet (x10 ³ /μL)	110.8 ± 91.8	141.4 ± 94.0	99.3 ± 88.4	<0.001^**
Serum albumin (g/dL)	2.7 ± 0.6	2.9 ± 0.6	2.6 ± 0.6	<0.001^*
Serum phosphate (mg/dL)	5.5 ± 2.8	5.0 ± 2.5	5.7 ± 2.9	0.010^**
MBP (mmHg)	73.5 ± 16.3	79.0 ± 16.1	71.4 ± 15.9	<0.001^*
Heart rate (bpm)	110.8 ± 29.3	102.8 ± 27.2	113.8 ± 29.5	<0.001^*
Mechanical ventilation	395 (64.5)	71 (42.5)	324 (72.8)	<0.001⁺
Vasopressor use	445 (72.7)	85 (50.9)	360 (80.9)	<0.001⁺

Open in a new tab

Continuous data are presented as mean ± SD, and categorial data as number of patients (%). BUN, blood urea nitrogen; Hb, hemoglobin; WBC, white blood cell; CRP, C-reactive protein; INR, international normalized ratio; MBP, mean blood pressure.

*p-value by Student’s t-test

**p-value by Mann-Whitney test

Severity scores of study population

The mean severity scores for the study population were as follows: APACHE II score, 35.5; SAPS 3, 84.6; SOFA score, 9.0; MODS, 10.7; Liano’s score, 0.55; PICARD score, 0.43; and Demirjian’s score, 0.60. Table 3 shows a comparison of the severity scores between survivors and non-survivors. There was a significant difference in the mean of all severity scores between survivors and non-survivors (p < 0.01 for all).

Table 3. Comparison of severity score between survivors and non-survivors.

Severity score	All patients (n = 612) mean ± SD	Survivors (n = 167) mean ± SD	Non-survivors (n = 445) mean ± SD	p-value
APACHE II score	35.5 ± 9.3	30.6 ± 8.8	37.3 ± 8.8	<0.001^*
SAPS 3	84.6 ± 19.9	73.6 ± 17.7	87.7 ± 19.8	<0.001^*
SOFA score	9.0 ± 3.3	8.0 ± 3.3	9.4 ± 3.2	<0.001^*
MODS	10.7 ± 4.3	9.8 ± 0.5	11.1 ± 0.1	0.001^*
Liano’s score	0.55 ± 0.22	0.42 ± 0.19	0.60 ± 0.21	<0.001^*
PICARD score	0.43 ± 0.16	0.37 ± 0.15	0.46 ± 0.15	<0.001^*
Demirjian’s score	0.60 ± 0.31	0.38 ± 0.29	0.68 ± 0.28	<0.001^**

Open in a new tab

Data are presented as mean ± SD. APACHE, Acute Physiology and Chronic Health Evaluation; SAPS, Simplified Acute Physiology Score; SOFA, Sequential Organ Failure Assessment; MODS, Multiple Organ Dysfunction Score; PICARD, The Program to Improve Care in Acute Renal Disease.

*p-value by Student’s t-test

**p-value by Mann-Whitney test

Discrimination of each severity scoring system

The AUROCs for in-hospital mortality are shown in Fig 1. AUROC revealed acceptable discrimination ability for Demirjian’s score, followed by Liano’s score. Table 4 shows the results of the comparison of AUROC between the scoring systems. Demirjian’s score was not significantly different from Liano’s score but was significantly higher than the rest.

Table 4. Pairwise comparison of receiver operating characteristic curves for in-hospital mortality.

Severity score system		Compared scores	95% CI	p-value
Demirjian’s score	~	Liano’s score	-0.005–0.091	0.081
	~	APACHE II score	0.009–0.111	0.020
	~	SAPS 3	0.050–0.149	<0.001
	~	MODS	0.053–0.157	<0.001
	~	PICARD score	0.065–0.159	<0.001
	~	SOFA score	0.093–0.204	<0.001
Liano’s score	~	APACHE II score	-0.029–0.063	0.459
	~	SAPS 3	0.008–0.106	0.023
	~	MODS	0.004–0.121	0.036
	~	PICARD score	0.006–0.132	0.031
	~	SOFA score	0.042–0.017	0.001
APACHE II score	~	SAPS 3	-0.002–0.082	0.064
	~	MODS	-0.005–0.095	0.076
	~	PICARD score	-0.011–0.115	0.105
	~	SOFA score	0.028–0.148	0.004
SAPS 3	~	MODS	-0.042–0.054	0.816
	~	PICARD score	-0.044–0.069	0.664
	~	SOFA score	-0.005–0.102	0.073
MODS	~	PICARD score	-0.053–0.067	0.825
	~	SOFA score	0.001–0.086	0.048
PICARD score	~	SOFA score	-0.025–0.097	0.241

Open in a new tab

APACHE, Acute Physiology and Chronic Health Evaluation; SAPS, Simplified Acute Physiology Score; SOFA, Sequential Organ Failure Assessment; MODS, Multiple Organ Dysfunction Score; PICARD, The Program to Improve Care in Acute Renal Disease.

To compare the discrepancies in severity scores according to the survival period, cases of death within 7 days after CRRT initiation were evaluated. The AUROC for 7-day mortality was as follows: APACHE II score, 0.707 (95% CI, 0.666–0.748); SAPS 3, 0.629 (95% CI, 0.585–0.673); SOFA score, 0.590 (95% CI, 0.545–0.635); MODS, 0.651 (95% CI, 0.607–0.694); Liano’s score, 0.725 (95% CI, 0.686–0.765); PICARD score, 0.569 (95% CI, 0.523–0.614); and Demirjian’s score, 0.768 (95% CI, 0.731–0.805). Similar to in-hospital mortality, Demirjian’s score showed a relatively high value in predicting the 7-day mortality.

Calibration of each severity scoring systems

Fig 2 shows the calibration of each severity score. Two severity scoring systems were excluded: the SOFA score and MODS, which do not generate the probability of death but only count points. Except for Liano’s score (χ2 = 7.555, p = 0.478) and the PICARD score (χ2 = 14.835, p = 0.062), the Hosmer-Lemeshow test for in-hospital mortality demonstrated a poor fit of the prediction models (p < 0.05). The calibration for 7-day mortality was similar to that for in-hospital mortality. Only SAPS 3 showed different results, and the result was significant for 7-day mortality (χ2 = 9.224, p = 0.324).

Discussion

This study evaluated and compared the predictive ability of the severity scores of patients who received CRRT. The primary result was that the kidney severity scores performed better than the general severity scores because of comparing the predictive ability between the severity scores. The AUROC for in-hospital mortality revealed acceptable discrimination ability of the Demirjian’s score (0.770), followed by the Liano’s (0.728) and APACHE II (0.710) scores. Demirjian’s score also showed the highest predictive value for 7-day mortality, followed by the Liano’s and APACHE II scores. The AUROC comparison showed that Demirjian’s score was significantly higher than the other scores except for Liano’s score. In addition, the Hosmer-Lemeshow test results of five scores, which provide predicted mortality, showed poor calibration of all scores except for those of Liano’s and PICARD. The calibration for 7-day mortality was similar to that for in-hospital mortality.

The in-hospital mortality was 72.7% in this study, which is higher than the 23.9%–60.3% due to AKI [1–6], implying that among patients with AKI, patients undergoing CRRT have a higher mortality rate. In a previous study of 1738 patients with AKI, 76.2% received mechanical ventilation, 69.1% received vasopressors, and 47.5% had sepsis at the onset of CRRT [6]. The use of mechanical ventilation was lower in this study than in a previous study; however, the use of vasopressors and sepsis rates were higher. These differences may be related to differences in mortality rates. In the non-survivor group, the MBP was lower at the start of CRRT, and mechanical ventilation and vasopressor use were more frequent, indicating that vital signs at the beginning of CRRT were worse. However, the mortality rate of our study population, which was relatively higher than that of other studies, is a potential limitation.

The general severity scores evaluated in our study were lower than the AUROC of 0.7, except for the APACHE II score. The APACHE II score was more discriminative than the other general scores; however, the results of the fitness test showed poor calibration, and the calibration line tended to underestimate mortality.

In previous studies that evaluated mortality based on general severity scores in patients with AKI, discriminant assessments were inconsistent. Passos et al. compared the APACHE II score, SAPS 3, and SOFA score in 186 patients with sepsis who underwent CRRT, and the AUROC showed poor discrimination, with 0.57, 0.48, and 0.58, respectively [33]. A study of 1169 patients with AKI in China from 1996 to 2013 showed that the AUROC of the SOFA score was 0.78 [34]. Of the 731 patients, only 56.1% underwent RRT, and the overall mortality rate was 13.8%, indicating that the severity of the disease was low.

Liano’s score was developed by Liano et al. in Spain [20]. The discriminant ability of Liano’s score was higher than that of the other scores, except for Demirjian’s score in this study. Additionally, the Hosmer-Lemeshow test on Liano’s score showed good calibration. Liano’s scores have been evaluated for external validity in several previous studies. Uchino et al. conducted a prospective multinational multicenter study of patients with AKI involving 54 medical institutions in 23 countries from 2000 to 2001 [28]. A total of four kidney-specific severity scores (Mehta [35], Liano’s, Chertow [36], and Paganini [37]) and two general severity scores (SAPS II and SOFA) were calculated to compare the predictive ability. The AUROC of Liano’s score was 0.698, which was more discriminative than the other scores; however, all were less than 0.7. Calibration was poor for all except for Liano’s score in this study.

Maccariello et al. compared the mortality predictive ability of the APACHE II score, SAPS II, Logistic Organ Dysfunction [38], Organ Dysfunction and Infection [39], Liano’s score, and Mehta score in 467 patients with AKI who received RRT in ICUs [29]. The AUROC score was above 0.7 for the SAPS II and Mehta scores, and all the scores except the Mehta score showed good calibration. In this study, the high proportion of patients with sepsis (76%) and mechanical ventilation dependence (81%) may have influenced the results. Ohnuma et al., performed a retrospective data analysis of 343 patients with AKI who underwent CRRT in Japan [31]. The mortality predictive external validity of the Mehta, SHARF II [40], PICARD, VELLORE [41], Liano’s, and Demirjian’s scores, which are kidney-specific severity scores, were compared with the SOFA score; all were less than an AUROC of 0.7. The results of the goodness-of-fit tests were poor, except for Liano’s score.

The PICARD score was developed by Chertow et al. based on the Program to Improve Care in Acute Renal Disease (PICARD), a multicenter study of 618 patients with AKI in five U.S. medical institutions from 1999 to 2001 [21]. A formula for predicting 60-day mortality was developed by dividing the time of AKI diagnosis, consultation, and initiation of dialysis. In this study, a prediction formula was applied and analyzed based on the dialysis initiation time, where the AUROC was the highest in internal validation. The PICARD score showed good calibration but the lowest discrimination among the kidney-specific severity scores in our analysis. Discrimination is affected by the distribution of the target group, which is poor in the homogeneous group and good in the heterogeneous group [39]. The predicted mortality rate of PICARD was 43%, indicating its tendency to underestimate the mortality rate. This is thought to be attributed to a 60-day mortality criterion and a low mortality rate of 37% among the populations that developed these scores.

Demirjian’s score exhibited the highest discriminative ability. Demirjian’s score was developed from the Veterans Affairs/National Institutes of Health Acute Renal Failure trial network study in the United States [22] to predict the 60-day mortality by selecting 21 variables affecting mortality among patients with AKI who received CRRT. The Hosmer-Lemeshow test on Demirjian’s score showed a poor fit in our analysis; however, it was more acceptable than general severity scores.

Although this study is limited in that it analyzed retrospectively collected data in a single-institution ICU, it has the strength of assessing the mortality predictability of kidney-specific severity scores only in patients who received CRRT. In several previous studies, general severity scores were used for population stratification and balanced randomization to improve the quality of CRRT. For example, Zarbock et al. compared the effect of early and delayed RRT initiation on mortality in critically ill patients with AKI, in which randomization was stratified according to SOFA cardiovascular scores [25]. In the study by Barbar et al. on the timing of RRT in patients with AKI, randomization was performed based on a minimization technique with stratification according to center, age, SOFA score, and type of infection [26]. This study supports the fact that the kidney-specific severity scores have higher discriminative ability than systemic scores in predicting mortality in patients receiving CRRT, and highlights the need to develop more predictable tools for patients with AKI receiving CRRT.

Patients with CKD were included in the study population, except those who received renal replacement therapy, such as intermittent hemodialysis, peritoneal dialysis, and kidney transplantation. Since the focus was on patients receiving CRRT, the study results are unlikely to change due to the characteristics of this cohort; however, the lack of information, such as baseline creatinine or eGFR, is a limitation of this study.

Conclusions

In summary, compared with general severity scores, kidney-specific severity scores demonstrated better calibration and discrimination in predicting mortality in patients with AKI receiving CRRT. However, none of the parameters evaluated in this study exhibited both excellent differentiation and calibration. In conclusion, all severity scoring systems included in this study had a limited ability to predict mortality in critically ill patients requiring CRRT. Therefore, we emphasize the need to develop novel severity scores with good calibration and high discrimination abilities.

Data Availability

Data cannot be shared publicly because of confidential information. Data are available from the Korea University Anam Hosptial Institutional Review board (e-mali: selfmaster@korea.ac.kr) for researchers who meet the criteria for access to confidential data.

Funding Statement

The authors received no specific funding for this work.

References

1.Lewington AJ, Cerdá J, Mehta RL. Raising awareness of acute kidney injury: a global perspective of a silent killer. Kidney Int. 2013;84: 457–467. doi: 10.1038/ki.2013.153 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Doyle JF, Forni LG. Acute kidney injury: short-term and long-term effects. Crit Care. 2016;20: 188. doi: 10.1186/s13054-016-1353-y [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Meersch M, Küllmar M, Schmidt C, Gerss J, Weinhage T, Margraf A, et al. Long-term clinical outcomes after early initiation of RRT in critically ill patients with AKI. J Am Soc Nephrol. 2018;29: 1011–1019. doi: 10.1681/ASN.2017060694 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Fuhrman DY, Kane-Gill S, Goldstein SL, Priyanka P, Kellum JA. Acute kidney injury epidemiology, risk factors, and outcomes in critically ill patients 16–25 years of age treated in an adult intensive care unit. Ann Intensive Care. 2018;8: 26. doi: 10.1186/s13613-018-0373-y [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Bellomo R, Ronco C, Mehta RL, Asfar P, Boisramé-Helms J, Darmon M, et al. Acute kidney injury in the ICU: from injury to recovery: reports from the 5th Paris International Conference. Ann Intensive Care. 2017;7: 49. doi: 10.1186/s13613-017-0260-y [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, et al. Acute renal failure in critically ill patients: a multinational, multicenter study. JAMA. 2005;294: 813–818. doi: 10.1001/jama.294.7.813 [DOI] [PubMed] [Google Scholar]
7.Sawhney S, Fraser SD. Epidemiology of AKI: utilizing large databases to determine the burden of AKI. Adv Chronic Kidney Dis. 2017;24: 194–204. doi: 10.1053/j.ackd.2017.05.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Li PK, Burdmann EA, Mehta RL, World Kidney Day Steering Committee 2013. Acute kidney injury: global health alert. Arab J Nephrol Transplant. 2013;6: 75–81. [PubMed] [Google Scholar]
9.Iqbal S, Faruq MO. Renal replacement therapy in critically ill: current trend and new direction. Bangladesh Crit Care J. 2015;3: 17–21. doi: 10.3329/bccj.v3i1.24096 [DOI] [Google Scholar]
10.Vanholder R, Van Biesen WV, Lameire N. What is the renal replacement method of first choice for intensive care patients? J Am Soc Nephrol. 2001;12 Supplement 17: S40–S43. doi: 10.1681/ASN.V12suppl_1s40 [DOI] [PubMed] [Google Scholar]
11.Uchino S, Bellomo R, Morimatsu H, Morgera S, Schetz M, Tan I, et al. Continuous renal replacement therapy: a worldwide practice survey. The beginning and ending supportive therapy for the kidney (B.E.S.T. kidney) investigators. Intensive Care Med. 2007;33: 1563–1570. doi: 10.1007/s00134-007-0754-4 [DOI] [PubMed] [Google Scholar]
12.Lemeshow S, Klar J, Teres D. Outcome prediction for individual intensive care patients: useful, misused, or abused? Intensive Care Med. 1995;21: 770–776. doi: 10.1007/BF01704747 [DOI] [PubMed] [Google Scholar]
13.Pirracchio R, Petersen ML, Carone M, Rigon MR, Chevret S, van der Laan MJ. Mortality prediction in intensive care units with the Super ICU Learner Algorithm (SICULA): a population-based study. Lancet Respir Med. 2015;3: 42–52. doi: 10.1016/S2213-2600(14)70239-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Kee YK, Kim EJ, Park KS, Han SG, Han IM, Yoon CY, et al. The effect of specialized continuous renal replacement therapy team in acute kidney injury patients treatment. Yonsei Med J. 2015;56: 658–665. doi: 10.3349/ymj.2015.56.3.658 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Fieghen HE, Friedrich JO, Burns KE, Nisenbaum R, Adhikari NK, Hladunewich MA, et al. The hemodynamic tolerability and feasibility of sustained low efficiency dialysis in the management of critically ill patients with acute kidney injury. BMC Nephrol. 2010;11: 32. doi: 10.1186/1471-2369-11-32 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Knaus WA, Draper EA, Wagner DP, Zimmerman JE. Apache II: a severity of disease classification system. Crit Care Med. 1985;13: 818–829. doi: 10.1097/00003246-198510000-00009 [DOI] [PubMed] [Google Scholar]
17.Moreno RP, Metnitz PG, Almeida E, Jordan B, Bauer P, Campos RA, et al. SAPS 3—From evaluation of the patient to evaluation of the intensive care unit. Part 2: Development of a prognostic model for hospital mortality at ICU admission. Intensive Care Med. 2005;31: 1345–1355. doi: 10.1007/s00134-005-2763-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Vincent JL, Moreno R, Takala J, Willatts S, De Mendonça AD, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22: 707–710. doi: 10.1007/BF01709751 [DOI] [PubMed] [Google Scholar]
19.Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score:–a reliable descriptor of a complex clinical outcome. Crit Care Med. 1995;23: 1638–1652. doi: 10.1097/00003246-199510000-00007 [DOI] [PubMed] [Google Scholar]
20.Liaño F, Gallego A, Pascual J, García-Martín F, Teruel JL, Marcén R, et al. Prognosis of acute tubular necrosis: an extended prospectively contrasted study. Nephron. 1993;63: 21–31. doi: 10.1159/000187139 [DOI] [PubMed] [Google Scholar]
21.Chertow GM, Soroko SH, Paganini EP, Cho KC, Himmelfarb J, Ikizler TA, et al. Mortality after acute renal failure: models for prognostic stratification and risk adjustment. Kidney Int. 2006;70: 1120–1126. doi: 10.1038/sj.ki.5001579 [DOI] [PubMed] [Google Scholar]
22.Demirjian S, Chertow GM, Zhang JH, O’Connor TZ, Vitale J, Paganini EP, et al. Model to predict mortality in critically ill adults with acute kidney injury. Clin J Am Soc Nephrol. 2011;6: 2114–2120. doi: 10.2215/CJN.02900311 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Yoon BR, Leem AY, Park MS, Kim YS, Chung KS. Optimal timing of initiating continuous renal replacement therapy in septic shock patients with acute kidney injury. Sci Rep. 2019;9: 11981. doi: 10.1038/s41598-019-48418-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Ruiz EF, Ortiz-Soriano VM, Talbott M, Klein BA, Thompson Bastin ML, Mayer KP, et al. Development, implementation and outcomes of a quality assurance system for the provision of continuous renal replacement therapy in the intensive care unit. Sci Rep. 2020;10: 20616. doi: 10.1038/s41598-020-76785-w [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Zarbock A, Kellum JA, Schmidt C, Van Aken H, Wempe C, Pavenstädt H, et al. Effect of early vs delayed initiation of renal replacement therapy on mortality in critically ill patients with acute kidney injury: The ELAIN Randomized Clinical Trial. JAMA. 2016;315: 2190–2199. doi: 10.1001/jama.2016.5828 [DOI] [PubMed] [Google Scholar]
26.Barbar SD, Clere-Jehl R, Bourredjem A, Hernu R, Montini F, Bruyère R, et al. Timing of renal-replacement therapy in patients with acute kidney injury and sepsis. N Engl J Med. 2018;379: 1431–1442. doi: 10.1056/NEJMoa1803213 [DOI] [PubMed] [Google Scholar]
27.Suresh K. An overview of randomization techniques: an unbiased assessment of outcome in clinical research. J Hum Reprod Sci. 2011;4: 8–11. doi: 10.4103/0974-1208.82352 [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
28.Uchino S, Bellomo R, Morimatsu H, Morgera S, Schetz M, Tan I, et al. External validation of severity scoring systems for acute renal failure using a multinational database. Crit Care Med. 2005;33: 1961–1967. doi: 10.1097/01.ccm.0000172279.66229.07 [DOI] [PubMed] [Google Scholar]
29.Maccariello ER, Valente C, Nogueira L, Ismael M, Valença RVR, Machado JES, et al. Performance of six prognostic scores in critically ILL patients receiving renal replacement therapy. Rev Bras Ter Intensiva. 2008;20: 115–123. [PubMed] [Google Scholar]
30.Varricatt VP, Rau NR, Attur RP, Baig WW. Validation of Liano score in acute renal failure: a prospective study in Indian patients. Clin Exp Nephrol. 2009;13: 33–37. doi: 10.1007/s10157-008-0073-2 [DOI] [PubMed] [Google Scholar]
31.Ohnuma T, Uchino S, Toki N, Takeda K, Namba Y, Katayama S, et al. External validation for acute kidney injury severity scores: A multicenter retrospective study in 14 Japanese ICUs. Am J Nephrol. 2015;42: 57–64. doi: 10.1159/000439118 [DOI] [PubMed] [Google Scholar]
32.Skarupskienė I, Adukauskienė D, Kuzminskienė J, Rimkutė L, Balčiuvienė V, Žiginskienė E, et al. Mortality prediction in patients with acute kidney injury requiring renal replacement therapy after cardiac surgery. Medicina (Kaunas). 2017;53: 217–223. doi: 10.1016/j.medici.2017.06.003 [DOI] [PubMed] [Google Scholar]
33.da Hora Passos R, Ramos JG, Mendonça EJ, Miranda EA, Dutra FRD, Coelho MF, et al. A clinical score to predict mortality in septic acute kidney injury patients requiring continuous renal replacement therapy: the HELENICC score. BMC Anesthesiol. 2017;17: 21. doi: 10.1186/s12871-017-0312-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Luo M, Yang Y, Xu J, Cheng W, Li XW, Tang MM, et al. A new scoring model for the prediction of mortality in patients with acute kidney injury. Sci Rep. 2017;7: 7862. doi: 10.1038/s41598-017-08440-w [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Mehta RL, Pascual MT, Gruta CG, Zhuang S, Chertow GM. Refining predictive models in critically ill patients with acute renal failure. J Am Soc Nephrol. 2002;13: 1350–1357. doi: 10.1097/01.asn.0000014692.19351.52 [DOI] [PubMed] [Google Scholar]
36.Chertow GM, Lazarus JM, Paganini EP, Allgren RL, Lafayette RA, Sayegh MH. Predictors of mortality and the provision of dialysis in patients with acute tubular necrosis. The auriculin anaritide acute renal failure study group. J Am Soc Nephrol. 1998;9: 692–698. doi: 10.1681/ASN.V94692 [DOI] [PubMed] [Google Scholar]
37.Paganini EP, Halstenberg WK, Goormastic M. Risk modeling in acute renal failure requiring dialysis: the introduction of a new model. Clin Nephrol. 1996;46: 206–211. [PubMed] [Google Scholar]
38.Le Gall JR, Klar J, Lemeshow S, Saulnier F, Alberti C, Artigas A, et al. The logistic organ dysfunction system. A new way to assess organ dysfunction in the Intensive Care Unit. ICU Scoring Group. JAMA. 1996;276: 802–810. doi: 10.1001/jama.276.10.802 [DOI] [PubMed] [Google Scholar]
39.Fagon JY, Chastre J, Novara A, Medioni P, Gibert C. Characterization of intensive care unit patients using a model based on the presence or absence of organ dysfunctions and/or infection: the Odin model. Intensive Care Med. 1993;19: 137–144. doi: 10.1007/BF01720528 [DOI] [PubMed] [Google Scholar]
40.Lins RL, Elseviers MM, Daelemans R, Arnouts P, Billiouw JM, Couttenye M, et al. Re-evaluation and modification of the Stuivenberg hospital acute renal failure (SHARF) scoring system for the prognosis of acute renal failure: an independent multicentre, prospective study. Nephrol Dial Transplant. 2004;19: 2282–2288. doi: 10.1093/ndt/gfh364 [DOI] [PubMed] [Google Scholar]
41.Dharan KS, John GT, Antonisamy B, Kirubakaran MG, Jacob CK. Prediction of mortality in acute renal failure in the tropics. Ren Fail. 2005;27: 289–296. doi: 10.1081/JDI-56612 [DOI] [PubMed] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0286246.r001

Decision Letter 0

Giuseppe Remuzzi

9 Feb 2023

PONE-D-22-26854Assessment of severity scoring systems for predicting mortality in critically ill patients receiving continuous renal replacement therapyPLOS ONE

Dear Dr. Chun,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The manuscript focuses as a topic of potential interest. The study, however, has major shortcomings that preclude sound conclusions, and should be addressed. To mention some of them, i) concern about the fact that many of the scores being compared have used in their developmental cohort patients only with AKI needing CRRT unlike author study here where CKD patients needed dialysis is also included; ii) need to mention what percent of their patients had CKD among survivors vs non-survivors; iii) there is no mention of timing of scoring with respect to CRRT start; iv) concern about the fact that the authors overstate how good any of the models are; v) need to further explain that none of these tests are particularly good, and therefore they should more strongly emphasize that the primary conclusion of this study is that additional research is needed to develop better scoring systems; vi) unclear why the mortality of the cohort was so high; vii) need to include this high mortality as a potential limitation of the study, because these results may not be as relevant to centers with different patient selection and/or better CRRT outcomes, viii) need to report the results of the Hosmer-Lemeshow tests by simply providing the terms “significant” or “not significant”; ix) unclear if this is a secondary analysis of a cohort which was generated for a prior study; x) concern about Table 3, which is somewhat confusing; xi) need to review some statements in the Introduction that are somewhat misleading; xii) need to provide the distribution of baseline creatinine or eGFR and the proportion of patients with advanced CKD (stage 4 or 5); xiii) need to provide the relative number of patients with de novo AKI, AKI on CKD, and progressive CKD.

Please submit your revised manuscript by Mar 26 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Giuseppe Remuzzi

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please ensure that you have specified (1) whether consent was informed and (2) what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

If your study used retrospective hospital data- in ethics statement in the manuscript please ensure that you have discussed whether all data/samples were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data/samples from their medical records used in research, please include this information.

3. Please include your full ethics statement in the ‘Methods’ section of your manuscript file. In your statement, please include the full name of the IRB or ethics committee who approved or waived your study, as well as whether or not you obtained informed written or verbal consent. If consent was waived for your study, please include this information in your statement as well.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The study by Chun et al is an original research to attempt using well validated scoring systems to predict 7 day and in hospital mortality in patients on CRRT. The study does follow the necessary statistical requirement to compare scoring systems and uses the standard metrics to judge prognostic scoring systems including discrimination and calibration. Unfortunately the conclusions are not valid since there are methodological flaws. Firstly many of the scores being compared have used in their developmental cohort patients only with AKI needing CRRT unlike authors study where CKD patients needing dialysis is also included. Additionally, they do not mention what percentage of their patients had CKD among survivors vs non-survivors. Besides this there is no mention of timing of scoring with respect to CRRT start.

Minor issues: Sentences need to structured for more clarity: for instance

Line 71 : which study is mentioned here

Line 72: Another several studies is grammatically incorrect. I think authors want to say that other studies have compared predictive power of disease severity scoring systems

Line 74 : needs clarification

Other than that the research meets all applicable standards for the ethics of experimentation and research integrity and seems to adhere to appropriate reporting guidelines and community standards for data availability.

Reviewer #2: Chun and colleagues present a single-center retrospective study from Korea of the ability of a variety of general and kidney-specific disease severity scores to predict mortality in patients with acute kidney injury (AKI) requiring continuous renal replacement therapy (CRRT). In short, though they somewhat overstate the utility of the kidney-specific scores, they found no single score performs that well (i.e., none had excellent discrimination nor calibration), though overall the kidney-specific scores performed somewhat better. Though the study has a lot of limitations, it is a relevant topic and reasonably thought provoking. The methods used appear appropriate and straightforward, and the data are presented clearly. However, the manuscript needs significant editing to improve the quality and clarify of the English language. In addition, I have the following recommendations, questions, and concerns:

1. The authors overstate how good any of these models are. For example, in the results the authors state that, "The AUROC revealed high discrimination ability of Demirjian’s score followed by Liano's score." In the discussion, they state that, "The AUROC for in-hospital mortality revealed high discrimination ability of Demirjian’s score (0.770) followed by Liano's score (0.728) and APACHE II score (0.710)." I would not agree with these statements. Though it's somewhat arbitrary, most would consider AUROC of 0.7-0.8 to have "moderate" or "acceptable" predictive ability. For example, a frequently cited interpretation of AUROC is as follows:

0.5 = No discrimination

0.5-0.7 = Poor discrimination

0.7-0.8 = Acceptable discrimination

0.8-0.9= Excellent discrimination

>0.9 = Outstanding discrimination

[From: Hosmer, D.W., Jr., Lemeshow, S. and Sturdivant, R.X. (2013). Assessing the Fit of the Model. In Applied Logistic Regression (eds D.W. Hosmer, S. Lemeshow and R.X. Sturdivant). https://doi.org/10.1002/9781118548387.ch5]

Overall, the results of this study are disappointing -- only 1 test was had borderline good AUC (Demirjian’s), but even that test had poor fit. I think the authors needs to do a better job of explaining that none of these tests are particularly good, and therefore they should more strongly emphasize that the primary conclusion of this study is that additional research is needed to develop better scoring systems.

2. Why was the mortality of the cohort so high? The authors should try to better address this. The authors suggest that AKI treated specifically with CRRT has higher mortality than AKI patients overall, which is true, but even for AKI requiring CRRT this is high. In most studies, the mortality of AKI requiring CRRT is closer to 50%. Comparing it to the BEST study (reference 6) -- which reported data from patients treated >20 years -- doesn't seem to adequate address the extremely high mortality seen here. To the degree that these authors are looking to evaluate mortality prediction tools, analyzing a cohort that has a much higher mortality than other centers could make the findings less generalizable. To help address this, they should report the mean or median and distribution of the disease severity scores of this cohort, particularly those that are a reflection of overall disease severity (i.e., SOFA, SAPS 3, APACHE 2, MODS). Regardless of whether they are able to somehow justify such a high mortality, this high mortality should be included as a potential limitation of the study, because these results may not be as relevant to centers with different patient selection and/or better CRRT outcomes.

3. This is extremely poorly worded: "Except for Liano's score (χ2=7.555, 166 p=0.478) and PICARD score (χ2=14.835, p=0.062), the Hosmer-Lemeshow test for in-hospital mortality demonstrated that calibration ability of all scores was not significant." Essentially, the authors are misusing the word significant here. The term "significant" when describing a statistical test indicates you reject the null hypothesis because the observed findings are unlikely to be due to change (i.e., p is <0.05). Technically, for Liano and PICARD in this study, the results of the Hosmer-Lemeshow test are *not* statistically significant (i.e., the null hypothesis that the predicted and observed outcomes are the same is not rejected). For Hosmer-Lemeshow test, a significant test indicates that the model is *not* a good fit, and a non-significant test indicates a good fit. I would suggest the authors report the results of the Hosmer-Lemeshow tests by simply providing the appropriate interpretation and avoiding the terms "significant" or "not significant", which are confusing in general for this type of test. For example, the authors could state, "Except for Liano's score (χ2=7.555, 166 p=0.478) and PICARD score (χ2=14.835, p=0.062), the Hosmer-Lemeshow test for in-hospital mortality demonstrated poor fit of the prediction models (p <0.05)." Similarly, this statement, "The APACHE II score was more discriminative compared to other general scores, but the results of the fitness test were not significant" could be changed to "The APACHE II score was more discriminative compared to other general scores, but the results of the fitness test showed poor calibration." Similarly, the abstract should state, "The Hosmer-Lemeshow test demonstrated good fit of Liano's score and PICARD scores." Likely, in the discussion I suggest, "The Hosmer-Lemeshow test on Demirjian’s score showed poor fit in our analysis, but it was more acceptable compared to the general severity scores."

4. Why are the authors publishing retrospective data from 2016-2018 in 2023? That seems odd. It certainly takes time to do this research, but a 4-year interval seems much. Is this a secondary analysis of a cohort which was generated for a prior study? If so, the authors should cite the prior work. Otherwise, at least a brief explanation as to why these patients from a cohort that is >4y old at publication seems warranted.

5. Table 3, as written, is somewhat confusing. It is hard to follow the directionality of comparisons. To help make it easier to follow, I suggest putting the score with the highest AUC first (Demirjian), followed by the second highest (Liano), etc. That should make it a little easier to read.

6. In the introduction, I found this statement to be somewhat misleading: "This is because the therapy is hemodynamically more stable than the intermittent hemodialysis therapy, and it is easy to control fluid balance and to correct metabolic acidosis or electrolyte imbalance and to correct nutritional deficiency [9]." In general, intermittent HD corrects acidosis and electrolytes just as well as CRRT. I also don't understand what is meant by "correct nutritional deficiency". I would simply end the sentence after "...control fluid balance." If they want to claim that CRRT is better for acidosis or electrolytes, a much more complicated discussion about instantaneous clearance vs. today daily dose of RRT (i.e., equilibrated Kt/V) would be needed, but it’s just best to avoid suggesting that CRRT is better than IHD for acidosis or electrolytes.

7. For the calibration tests, why did the author present results for only 5 of the 7 scoring systems? They should present them all or explain why they excluded SOFA and MODS.

8. In the discussion, for all the other prediction scores used (e.g., Mehta, Chertow, Paganini, SHARF II, and VELLORE), the authors should cite the original publications describing these scoring systems.

9. Do the authors have the distribution of baseline creatinine or eGFR? Or do they have the proportion of patients with advanced (e.g., stage 4 or 5) CKD? The authors suggest some patients with advanced CKD were included in the cohort. These patients (which many of which may be better classified as new ESKD rather than AKI) could be vastly different than patients with AKI. To better understand this cohort, more information about the relative number of patients with de novo AKI, AKI on CKD, and progressive CKD would be good.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Anirban Ganguli

Reviewer #2: Yes: J. Pedro Teixeira

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2023 May 25;18(5):e0286246. doi: 10.1371/journal.pone.0286246.r002

Author response to Decision Letter 0

27 Mar 2023

Response to reviewer(s)’ comments

Dear Reviewer(s),

We are grateful for the valuable comments. After thorough discussions, we have revised our paper to reflect your helpful recommendations. Our point-by-point responses to the comments appear in blue in the following, and the revised manuscript can be found enclosed with the submission.

Reviewer 1

The study by Chun et al is an original research to attempt using well validated scoring systems to predict 7 day and in hospital mortality in patients on CRRT. The study does follow the necessary statistical requirement to compare scoring systems and uses the standard metrics to judge prognostic scoring systems including discrimination and calibration. Unfortunately the conclusions are not valid since there are methodological flaws. Firstly many of the scores being compared have used in their developmental cohort patients only with AKI needing CRRT unlike authors study where CKD patients needing dialysis is also included. Additionally, they do not mention what percentage of their patients had CKD among survivors vs non-survivors.

� Thank you for your thorough comments. Many studies included all patients diagnosed with AKI who received renal replacement therapy such intermittent hemodialysis. Since our study focused on CRRT, we included CKD patients who received CRRT, except for those who received renal replacement therapy such as intermittent hemodialysis, peritoneal dialysis, or kidney transplants. As you mentioned, we added proportion of CKD patients as following:

(Line 123-125) CKD patients accounted for 11.9% of the total study population, and there was no significant difference in the proportion of survivors and non-survivors.

(Table 1: Line 130-135)

Table 1. Demographics and Clinical Characteristics of Study Population

Besides this there is no mention of timing of scoring with respect to CRRT start.

� Following your recommendation, we added the timing of scoring as following:

(Line 105-107) Severity scoring was performed when it was decided to start CRRT, and all variables for severity scores were collected within 24 hours prior to the initiation of CRRT.

Minor issues: Sentences need to structured for more clarity: for instance

Line 71 : which study is mentioned here

Line 72: Another several studies is grammatically incorrect. I think authors want to say that other studies have compared predictive power of disease severity scoring systems

Line 74 : needs clarification

� We appreciate for your help. We rephrased the sentences as the following:

(Line 68-70) In addition, in studies to improve the quality of CRRT, such as the optimal start time of CRRT [25, 26],

(Line 71-72) Other studies have compared predictive power of severity scoring systems in patients with AKI.

(Line 72-74) However, many studies included all patients diagnosed with AKI who received renal replacement therapy such intermittent hemodialysis, and few studies focused solely on CRRT [28-34].

Reviewer 2

0.5 = No discrimination

0.5-0.7 = Poor discrimination

0.7-0.8 = Acceptable discrimination

0.8-0.9= Excellent discrimination

>0.9 = Outstanding discrimination

� Thank you for your careful review of our work and your suggestions regarding the manuscript. Following your recommendation, we rephrased the sentences and added the opinion in the conclusion as the following:

(Line 159-160) The AUROC revealed acceptable discrimination ability of Demirjian’s score followed by Liano's score.

(Line 179-180) Same as in-hospital mortality, Demirjian’s score showed relatively high value to predict the 7-day mortality.

(Line 197-198) The AUROC for in-hospital mortality revealed acceptable discrimination ability of Demirjian’s score

(Line 281-285) However, none of those evaluated in this study showed both excellent differentiation and suitability. In conclusion, all severity scoring systems included in this study were inappropriate for predicting mortality of critically ill patients requiring CRRT. Therefore, we emphasize the need to develop a novel severity scores with good calibration and high discrimination for the patients.

� Thank you for your help. We added the mean severity scores of the study population. Also we added that high mortality is a potential limitation of our study as the following:

(Line 145-148) The mean severity scores of the study population were APACHE II score 35.5, SAPS 3 84.6, SOFA score 9.0, MODS 10.7, Liano's score 0.55, PICARD score 0.43, and Demirjian's score 0.60. Table 3 shows the comparison of severity score between survivors and non-survivors. There was significant difference in the mean of all severity scores between survivors and non-survivors (all p <0.01).

(Table 3: Line 150-155)

Table 3. Comparison of severity score between survivors and non-survivors

� We appreciate for your help. We rephrased the sentences as the following:

(Line 185-186) Except for Liano's score (χ2=7.555, p=0.478) and PICARD score (χ2=14.835, p=0.062), the Hosmer-Lemeshow test for in-hospital mortality demonstrated poor fit of the prediction models (p <0.05).

(Line 202-204) In addition, the Hosmer-Lemeshow test results of five scores, which provide predicted mortality, showed poor calibration of all scores except Liano's and PICARD scores.

(Line 216-217) The APACHE II score was more discriminative compared to other general scores, but the results of the fitness test showed poor calibration,

(Line 40-42) The Hosmer-Lemeshow test on Demirjian’s score showed poor fit in our analysis, but it was more acceptable compared to the general severity scores.

� We agree with your opinion. The reason we used the data from 2016 to 2018 is that the data at that time was accessible. Since the author belonged to the department at the time, it was easy to access the data at the time. After data collection, an actual analysis was conducted from 2019 to 2020. Above all, the authors believe that the characteristics of the cohort in this study, which evaluates the adequacy of scoring systems for a particular group, will not change the relevant risk factors of the study results.

� Thank you for your thorough comments. As you mentioned, we rephrased the table as the following:

(Table 4: Line 167-172)

Table 4. Pairwise Comparison of Receiver Operating Characteristic Curves for In-hospital Mortality

� We appreciate your suggestion. We rephrased the sentence as the following:

(Line 54-55) This is because the therapy is hemodynamically more stable than the intermittent hemodialysis therapy, and it is easy to control fluid balance [9].

7. For the calibration tests, why did the author present results for only 5 of the 7 scoring systems? They should present them all or explain why they excluded SOFA and MODS.

� We added the reasons for exclusion as following:

(Line 183-184) Two severity scoring systems were excluded, SOFA score and MODS, which do not generate probability of death, only counting points.

� Additional references is added as following:

(Line 231-232) Mehta score [35], Liano's score, Chertow score [36]

(Line 236) Logistic Organ Dysfunction [38], Organ Dysfunction and Infection [39]

(Line 242) SHARF II score [40], PICARD score, VELLORE score [41],

(Line 372-398)

35. Mehta RL, Pascual MT, Gruta CG, Zhuang S, Chertow GM. Refining predictive models in critically ill patients with acute renal failure. J Am Soc Nephrol. 2002; 13:1350–7.

36. Chertow GM, Lazarus JM, Paganini EP, Allgren RL, Lafayette RA, Sayegh MH. Predictors of mortality and the provision of dialysis in patients with acute tubular necrosis. The Auriculin Anaritide Acute Renal Failure Study Group. J Am Soc Nephrol. 1998;9:692–8.

37. Paganini EP, Halstenberg WK, Goormastic M. Risk modeling in acute renal failure requiring dialysis: The introduction of a new model. Clin Nephrol. 1996; 46:206–11.

38. Le Gall JR, Klar J, Lemesho S, Saulnier F, Alberti C, Artigas A, et al. The Logistic Organ Dysfunction System. A New Way to Assess Organ Dysfunction in the Intensive Care Unit. JAMA. 1996;276(10):802-10.

39. Fagon JY, Chastre J, Novara A, Medioni P, Gibert C. Characterization of intensive care unit patients using a model based on the presence or absence of organ dysfunctions and/or infection: The ODIN model. Intensive Care Medicine. 1993;19:137-44.

40. Lins RL, Elseviers MM, Daelemans R, Arnouts P, Billiouw JM, Couttenye M, et al. Re-evaluation and modification of the stuivenberg hospital acute renal failure (SHARF) scoring system for the prognosis of acute renal failure: an independent multicentre, prospective study. Nephrol Dial Transplant. 2004;19:2282–8.

41. Dharan KS, John GT, Antonisamy B, Kirubakaran MG, Jacob CK. Prediction of mortality in acute renal failure in the tropics. Ren Fail. 2005;27:289–96.

� Thank you for your careful comments. We provided the proportion of CKD patients regarding your comments. We also added this characteristic of the cohort as a limitation of our study as following:

(Line 123-125) CKD patients accounted for 11.9% of the total study population, and there was no significant difference in the proportion of survivors and non-survivors.

(Table 1: Line 130-135)

Table 1. Demographics and Clinical Characteristics of Study Population

(Line 273-277) In this study, CKD patients were included in the study population, except for patients who received renal replacement therapy such as intermittent hemodialysis, peritoneal dialysis, and kidney transplantation. Since the focus is on patients receiving CRRT, the study results are unlikely to change due to the characteristics of this cohort, but the lack of information such as baseline creatinine or eGFR is a limitation of this study.

Attachment

Submitted filename: (20230308) Response to reviewers.docx

Click here for additional data file.^{(48.2KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0286246.r003

Decision Letter 1

Giuseppe Remuzzi

14 Apr 2023

PONE-D-22-26854R1Assessment of severity scoring systems for predicting mortality in critically ill patients receiving continuous renal replacement therapyPLOS ONE

Dear Dr. Chun,

The revised manuscript is significantly improved. Most of the reviewer’s comments have been addressed. However, the authors need to further consider the few remaining minor recommendations by Reviewer 2 dealing with the conclusions. Moreover, additional English language editing of the manuscript is required.

Please submit your revised manuscript by May 29 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Giuseppe Remuzzi

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Reviewer #1: (No Response)

Reviewer #2: I think the manuscript is significantly improved. I have few additional minor recommendations:

1. In the discussion [lines 263-265 in the clean version] I once again suggest , "The Hosmer-Lemeshow test on Demirjian’s score showed poor fit in our analysis, but it was more acceptable compared to the general severity scores."

2. The term suitability is vague. In the conclusions, I recommend changing, "However, none of those evaluated in this study showed both excellent differentiation and suitability" to a more precise statement, "However, none of those evaluated in this study showed both excellent differentiation and calibration."

3. In the conclusion I also suggest being a bit less extreme by stating, "In conclusion, all severity scoring systems included in this study had limited ability to predict mortality of critically ill patients requiring CRRT."

4. As I commented in my first review, I think the manuscript could significantly benefit from additional English language editing.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Anirban Ganguli

Reviewer #2: Yes: J. Pedro Teixeira

**********

PLoS One. 2023 May 25;18(5):e0286246. doi: 10.1371/journal.pone.0286246.r004

Author response to Decision Letter 1

24 Apr 2023

Dear Reviewers (s):

We are grateful for your insightful comments. After thorough discussions, we have revised our paper to reflect your valuable recommendations. Our point-by-point responses to the comments are provided in blue in the following text, and the revised manuscript is enclosed with the submission.

Reviewer 2

1. In the discussion [lines 263-265 in the clean version] I once again suggest, "The Hosmer-Lemeshow test on Demirjian’s score showed poor fit in our analysis, but it was more acceptable compared to the general severity scores."

Response: Thank you for your careful review of our work and for your suggestions regarding this manuscript. Following your recommendation, we have rephrased the sentence as follows:

(Line 270-271) The Hosmer-Lemeshow test on Demirjian’s score showed a poor fit in our analysis; however, it was more acceptable than general severity scores.

Response: Thank you for your insightful comment. We have rephrased the sentence as follows:

(Line 292-293) However, none of the parameters evaluated in this study exhibited both excellent differentiation and calibration.

Response: Thank you for your valuable comments. As per your suggestion, we have rephrased the sentence as follows:

(Line 294-295) In conclusion, all severity scoring systems included in this study had a limited ability to predict mortality in critically ill patients requiring CRRT.

4. As I commented in my first review, I think the manuscript could significantly benefit from additional English language editing.

Response: Thank you for the valuable suggestion. We would like to thank Editage (www.editage.co.kr) for reviewing and editing the revised manuscript for English language.

Attachment

Submitted filename: (20230422)_Response to reviewers.docx

Click here for additional data file.^{(22.8KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0286246.r005

Decision Letter 2

Giuseppe Remuzzi

12 May 2023

Assessment of severity scoring systems for predicting mortality in critically ill patients receiving continuous renal replacement therapy

PONE-D-22-26854R2

Dear Dr. Chun,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

The re-revised version of the manuscript is definitely improved. The authors have properly addressed all the remaining issues raised by the Reviewer 2, and previously all comments from Reviewer 1. Thus, the manuscript is now suitable for publication.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Giuseppe Remuzzi

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #2: Yes

**********

6. Review Comments to the Author

Reviewer #2: The manuscript is significantly improved with the changes made. I have no additional comments to make at this point.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: J. Pedro Teixeira

**********

PLoS One. doi: 10.1371/journal.pone.0286246.r006

Acceptance letter

Giuseppe Remuzzi

17 May 2023

PONE-D-22-26854R2

Assessment of severity scoring systems for predicting mortality in critically ill patients receiving continuous renal replacement therapy

Dear Dr. Chun:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Giuseppe Remuzzi

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Attachment

Submitted filename: (20230308) Response to reviewers.docx

Click here for additional data file.^{(48.2KB, docx)}

Attachment

Submitted filename: (20230422)_Response to reviewers.docx

Click here for additional data file.^{(22.8KB, docx)}

Data Availability Statement

[pone.0286246.ref001] 1.Lewington AJ, Cerdá J, Mehta RL. Raising awareness of acute kidney injury: a global perspective of a silent killer. Kidney Int. 2013;84: 457–467. doi: 10.1038/ki.2013.153 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref002] 2.Doyle JF, Forni LG. Acute kidney injury: short-term and long-term effects. Crit Care. 2016;20: 188. doi: 10.1186/s13054-016-1353-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref003] 3.Meersch M, Küllmar M, Schmidt C, Gerss J, Weinhage T, Margraf A, et al. Long-term clinical outcomes after early initiation of RRT in critically ill patients with AKI. J Am Soc Nephrol. 2018;29: 1011–1019. doi: 10.1681/ASN.2017060694 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref004] 4.Fuhrman DY, Kane-Gill S, Goldstein SL, Priyanka P, Kellum JA. Acute kidney injury epidemiology, risk factors, and outcomes in critically ill patients 16–25 years of age treated in an adult intensive care unit. Ann Intensive Care. 2018;8: 26. doi: 10.1186/s13613-018-0373-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref005] 5.Bellomo R, Ronco C, Mehta RL, Asfar P, Boisramé-Helms J, Darmon M, et al. Acute kidney injury in the ICU: from injury to recovery: reports from the 5th Paris International Conference. Ann Intensive Care. 2017;7: 49. doi: 10.1186/s13613-017-0260-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref006] 6.Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, et al. Acute renal failure in critically ill patients: a multinational, multicenter study. JAMA. 2005;294: 813–818. doi: 10.1001/jama.294.7.813 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref007] 7.Sawhney S, Fraser SD. Epidemiology of AKI: utilizing large databases to determine the burden of AKI. Adv Chronic Kidney Dis. 2017;24: 194–204. doi: 10.1053/j.ackd.2017.05.001 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref008] 8.Li PK, Burdmann EA, Mehta RL, World Kidney Day Steering Committee 2013. Acute kidney injury: global health alert. Arab J Nephrol Transplant. 2013;6: 75–81. [PubMed] [Google Scholar]

[pone.0286246.ref009] 9.Iqbal S, Faruq MO. Renal replacement therapy in critically ill: current trend and new direction. Bangladesh Crit Care J. 2015;3: 17–21. doi: 10.3329/bccj.v3i1.24096 [DOI] [Google Scholar]

[pone.0286246.ref010] 10.Vanholder R, Van Biesen WV, Lameire N. What is the renal replacement method of first choice for intensive care patients? J Am Soc Nephrol. 2001;12 Supplement 17: S40–S43. doi: 10.1681/ASN.V12suppl_1s40 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref011] 11.Uchino S, Bellomo R, Morimatsu H, Morgera S, Schetz M, Tan I, et al. Continuous renal replacement therapy: a worldwide practice survey. The beginning and ending supportive therapy for the kidney (B.E.S.T. kidney) investigators. Intensive Care Med. 2007;33: 1563–1570. doi: 10.1007/s00134-007-0754-4 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref012] 12.Lemeshow S, Klar J, Teres D. Outcome prediction for individual intensive care patients: useful, misused, or abused? Intensive Care Med. 1995;21: 770–776. doi: 10.1007/BF01704747 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref013] 13.Pirracchio R, Petersen ML, Carone M, Rigon MR, Chevret S, van der Laan MJ. Mortality prediction in intensive care units with the Super ICU Learner Algorithm (SICULA): a population-based study. Lancet Respir Med. 2015;3: 42–52. doi: 10.1016/S2213-2600(14)70239-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref014] 14.Kee YK, Kim EJ, Park KS, Han SG, Han IM, Yoon CY, et al. The effect of specialized continuous renal replacement therapy team in acute kidney injury patients treatment. Yonsei Med J. 2015;56: 658–665. doi: 10.3349/ymj.2015.56.3.658 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref015] 15.Fieghen HE, Friedrich JO, Burns KE, Nisenbaum R, Adhikari NK, Hladunewich MA, et al. The hemodynamic tolerability and feasibility of sustained low efficiency dialysis in the management of critically ill patients with acute kidney injury. BMC Nephrol. 2010;11: 32. doi: 10.1186/1471-2369-11-32 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref016] 16.Knaus WA, Draper EA, Wagner DP, Zimmerman JE. Apache II: a severity of disease classification system. Crit Care Med. 1985;13: 818–829. doi: 10.1097/00003246-198510000-00009 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref017] 17.Moreno RP, Metnitz PG, Almeida E, Jordan B, Bauer P, Campos RA, et al. SAPS 3—From evaluation of the patient to evaluation of the intensive care unit. Part 2: Development of a prognostic model for hospital mortality at ICU admission. Intensive Care Med. 2005;31: 1345–1355. doi: 10.1007/s00134-005-2763-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref018] 18.Vincent JL, Moreno R, Takala J, Willatts S, De Mendonça AD, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22: 707–710. doi: 10.1007/BF01709751 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref019] 19.Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score:–a reliable descriptor of a complex clinical outcome. Crit Care Med. 1995;23: 1638–1652. doi: 10.1097/00003246-199510000-00007 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref020] 20.Liaño F, Gallego A, Pascual J, García-Martín F, Teruel JL, Marcén R, et al. Prognosis of acute tubular necrosis: an extended prospectively contrasted study. Nephron. 1993;63: 21–31. doi: 10.1159/000187139 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref021] 21.Chertow GM, Soroko SH, Paganini EP, Cho KC, Himmelfarb J, Ikizler TA, et al. Mortality after acute renal failure: models for prognostic stratification and risk adjustment. Kidney Int. 2006;70: 1120–1126. doi: 10.1038/sj.ki.5001579 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref022] 22.Demirjian S, Chertow GM, Zhang JH, O’Connor TZ, Vitale J, Paganini EP, et al. Model to predict mortality in critically ill adults with acute kidney injury. Clin J Am Soc Nephrol. 2011;6: 2114–2120. doi: 10.2215/CJN.02900311 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref023] 23.Yoon BR, Leem AY, Park MS, Kim YS, Chung KS. Optimal timing of initiating continuous renal replacement therapy in septic shock patients with acute kidney injury. Sci Rep. 2019;9: 11981. doi: 10.1038/s41598-019-48418-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref024] 24.Ruiz EF, Ortiz-Soriano VM, Talbott M, Klein BA, Thompson Bastin ML, Mayer KP, et al. Development, implementation and outcomes of a quality assurance system for the provision of continuous renal replacement therapy in the intensive care unit. Sci Rep. 2020;10: 20616. doi: 10.1038/s41598-020-76785-w [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref025] 25.Zarbock A, Kellum JA, Schmidt C, Van Aken H, Wempe C, Pavenstädt H, et al. Effect of early vs delayed initiation of renal replacement therapy on mortality in critically ill patients with acute kidney injury: The ELAIN Randomized Clinical Trial. JAMA. 2016;315: 2190–2199. doi: 10.1001/jama.2016.5828 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref026] 26.Barbar SD, Clere-Jehl R, Bourredjem A, Hernu R, Montini F, Bruyère R, et al. Timing of renal-replacement therapy in patients with acute kidney injury and sepsis. N Engl J Med. 2018;379: 1431–1442. doi: 10.1056/NEJMoa1803213 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref027] 27.Suresh K. An overview of randomization techniques: an unbiased assessment of outcome in clinical research. J Hum Reprod Sci. 2011;4: 8–11. doi: 10.4103/0974-1208.82352 [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]

[pone.0286246.ref028] 28.Uchino S, Bellomo R, Morimatsu H, Morgera S, Schetz M, Tan I, et al. External validation of severity scoring systems for acute renal failure using a multinational database. Crit Care Med. 2005;33: 1961–1967. doi: 10.1097/01.ccm.0000172279.66229.07 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref029] 29.Maccariello ER, Valente C, Nogueira L, Ismael M, Valença RVR, Machado JES, et al. Performance of six prognostic scores in critically ILL patients receiving renal replacement therapy. Rev Bras Ter Intensiva. 2008;20: 115–123. [PubMed] [Google Scholar]

[pone.0286246.ref030] 30.Varricatt VP, Rau NR, Attur RP, Baig WW. Validation of Liano score in acute renal failure: a prospective study in Indian patients. Clin Exp Nephrol. 2009;13: 33–37. doi: 10.1007/s10157-008-0073-2 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref031] 31.Ohnuma T, Uchino S, Toki N, Takeda K, Namba Y, Katayama S, et al. External validation for acute kidney injury severity scores: A multicenter retrospective study in 14 Japanese ICUs. Am J Nephrol. 2015;42: 57–64. doi: 10.1159/000439118 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref032] 32.Skarupskienė I, Adukauskienė D, Kuzminskienė J, Rimkutė L, Balčiuvienė V, Žiginskienė E, et al. Mortality prediction in patients with acute kidney injury requiring renal replacement therapy after cardiac surgery. Medicina (Kaunas). 2017;53: 217–223. doi: 10.1016/j.medici.2017.06.003 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref033] 33.da Hora Passos R, Ramos JG, Mendonça EJ, Miranda EA, Dutra FRD, Coelho MF, et al. A clinical score to predict mortality in septic acute kidney injury patients requiring continuous renal replacement therapy: the HELENICC score. BMC Anesthesiol. 2017;17: 21. doi: 10.1186/s12871-017-0312-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref034] 34.Luo M, Yang Y, Xu J, Cheng W, Li XW, Tang MM, et al. A new scoring model for the prediction of mortality in patients with acute kidney injury. Sci Rep. 2017;7: 7862. doi: 10.1038/s41598-017-08440-w [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0286246.ref035] 35.Mehta RL, Pascual MT, Gruta CG, Zhuang S, Chertow GM. Refining predictive models in critically ill patients with acute renal failure. J Am Soc Nephrol. 2002;13: 1350–1357. doi: 10.1097/01.asn.0000014692.19351.52 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref036] 36.Chertow GM, Lazarus JM, Paganini EP, Allgren RL, Lafayette RA, Sayegh MH. Predictors of mortality and the provision of dialysis in patients with acute tubular necrosis. The auriculin anaritide acute renal failure study group. J Am Soc Nephrol. 1998;9: 692–698. doi: 10.1681/ASN.V94692 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref037] 37.Paganini EP, Halstenberg WK, Goormastic M. Risk modeling in acute renal failure requiring dialysis: the introduction of a new model. Clin Nephrol. 1996;46: 206–211. [PubMed] [Google Scholar]

[pone.0286246.ref038] 38.Le Gall JR, Klar J, Lemeshow S, Saulnier F, Alberti C, Artigas A, et al. The logistic organ dysfunction system. A new way to assess organ dysfunction in the Intensive Care Unit. ICU Scoring Group. JAMA. 1996;276: 802–810. doi: 10.1001/jama.276.10.802 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref039] 39.Fagon JY, Chastre J, Novara A, Medioni P, Gibert C. Characterization of intensive care unit patients using a model based on the presence or absence of organ dysfunctions and/or infection: the Odin model. Intensive Care Med. 1993;19: 137–144. doi: 10.1007/BF01720528 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref040] 40.Lins RL, Elseviers MM, Daelemans R, Arnouts P, Billiouw JM, Couttenye M, et al. Re-evaluation and modification of the Stuivenberg hospital acute renal failure (SHARF) scoring system for the prognosis of acute renal failure: an independent multicentre, prospective study. Nephrol Dial Transplant. 2004;19: 2282–2288. doi: 10.1093/ndt/gfh364 [DOI] [PubMed] [Google Scholar]

[pone.0286246.ref041] 41.Dharan KS, John GT, Antonisamy B, Kirubakaran MG, Jacob CK. Prediction of mortality in acute renal failure in the tropics. Ren Fail. 2005;27: 289–296. doi: 10.1081/JDI-56612 [DOI] [PubMed] [Google Scholar]

PERMALINK

Assessment of severity scoring systems for predicting mortality in critically ill patients receiving continuous renal replacement therapy

Hyunmyung Park

Jihyun Yang

Byung Chul Chun

Roles

Abstract

Introduction

Materials and methods

Study population

Data

Statistical analysis

Results

Baseline characteristics and mortality of the study population

Table 1. Demographics and clinical characteristics of the study population.

Table 2. Clinical test results and acute physiology of study population at continuous renal replacement therapy (CRRT) initiation.

Severity scores of study population

Table 3. Comparison of severity score between survivors and non-survivors.

Discrimination of each severity scoring system

Fig 1. Area under receiver operating characteristic curves (AUROCs) of the seven severity scores for in-hospital mortality.

Table 4. Pairwise comparison of receiver operating characteristic curves for in-hospital mortality.

Calibration of each severity scoring systems

Fig 2.

Discussion

Conclusions

Data Availability

Funding Statement

References

Decision Letter 0

Giuseppe Remuzzi

Roles

Author response to Decision Letter 0

Decision Letter 1

Giuseppe Remuzzi

Roles

Author response to Decision Letter 1

Decision Letter 2

Giuseppe Remuzzi

Roles

Acceptance letter

Giuseppe Remuzzi

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases