Secondary Logo

Journal Logo

Review Article

Inverse Correlation Between Incidence and Mortality of Acute Kidney Injury in Critically Ill Patients: A Systematic Review

Komaru, Yohei; Inokuchi, Ryota; Iwagami, Masao; Hamasaki, Yoshifumi§; Nangaku, Masaomi∗,§; Doi, Kent

Author Information
doi: 10.1097/SHK.0000000000001511



Acute kidney injury (AKI) is frequently observed in intensive care units (ICUs) and has a considerable negative impact on the mortality of critically ill patients (1–3). Impaired renal function affects the balance of biological homeostasis, complicates fluid management, and eventually leads to poor patient outcome, especially in ICUs (4). Since the introduction of the standardized diagnostic criteria for AKI, including those of the Risk, Injury, Failure, Loss of kidney function, End-stage kidney disease (RIFLE) (5), Acute Kidney Injury Network (AKIN) (6), and Kidney Disease: Improving Global Outcomes (KDIGO) (7), there have been many clinical studies reporting the incidence and prognosis of AKI using these criteria. Susantitaphong et al. (8), in a systematic review and meta-analysis, comprehensively researched these such epidemiological studies on AKI that were published since the introduction of RIFLE criteria and confirmed the poor outcomes associated with the severity of AKI using KDIGO-equivalent criteria.

Notably, the reported incidence and mortality of AKI are remarkably different even when limited to critical care settings (9–11), which may be attributed to the differences in underlying diseases and the etiology of AKI. Conversely, increasing awareness of AKI will potentially improve the care of patients with AKI. Recently, an e-alert system has been developed for establishing an early diagnosis of AKI for better sensitivity (12). Moreover, a machine learning system for detecting AKI is currently being investigated (13). One interventional study applied a standardized AKI care bundle to the high-risk group and demonstrated significant improvement in post-cardiac surgery AKIs (14). Therefore, a high incidence of AKI may not necessarily be associated with a worse outcome if there is sufficient awareness regarding AKI and proper management of AKI without significant delay.

We hypothesize that there exists a negative correlation between the incidence and mortality of AKI in each study. For this, we systematically reviewed recent clinical studies regarding AKI incidence and mortality and evaluated the incidence of AKI along with the mortality in patients with AKI and the entire cohort.


Study design and search strategy

We systematically reviewed published manuscripts, which evaluated the incidence of AKI and the mortality of adult patients with AKI in ICU. Eligible studies were limited to those applying KDIGO-equivalent AKI criteria: RIFLE, AKIN, and KDIGO diagnostic and staging criteria. The study protocol was registered in the international prospective register of systematic reviews database (registration number, CRD 42019129322), and we followed the PRISMA guideline checklist for this systematic review (see Table S1, Supplemental Digital Content 1, (15). This study specifically aimed to identify clinical studies that included terms indicating both the stages of AKI and their diagnostic criteria (see Table S2, Supplemental Digital Content 2,, which showed full search terms of our systematic review). The search terms were almost identical to those used in an earlier report by Susantitaphong et al. (8). First, we referred to the aforementioned previous study to identify eligible research studies published from 2004 to August 2012 (8). On May 15, 2018, we conducted a search for studies published from August 2012 to May 1, 2018, using MEDLINE, EMBASE, and Cochrane Library. The following types of studies were excluded: those involving pediatric cohorts, those lacking KDIGO-equivalent AKI definition, those without a comparable non-AKI cohort, those without data on the mortality of AKI and/or non-AKI groups, those not evaluating short-term (≤30 days) outcomes, those with a small sample size (< 500 participants), case-control studies or case reports, second reports on the same cohort, and reports on non-ICU settings. We also excluded those not reported in English language. We subsequently evaluated all the included manuscripts for their quality, using the Newcastle-Ottawa quality scale for cohort studies (16). The study selection and quality assessment process were conducted by two pairs of independent physicians (YK and RI or MI), and any discrepancies between the physicians were resolved by discussion with a third ICU physician (KD) to reach a final decision.

Data extraction

A full-text review was conducted for all the included studies. The characteristics of each study (study design using AKI criteria [5–7]), participant demographics (sample size, age, and sex), and clinical outcomes (the number of patients with AKI, incidence of AKI in each cohort, proportion of patients with severe AKI, mortality of the entire cohort, and mortality of patients with AKI) were analyzed. Either ICU, hospital, or 30-day mortality was assessed as the outcome; when more than one of these data were available, the order of precedence was ICU mortality, hospital mortality, and then 30-day mortality. The category of ICU (e.g., general, surgical, and medical) was also referred. In this study, mild AKI was defined as stage 1, severe AKI as composite of stages 2 and 3, diagnosed by KDIGO-equivalent criteria (17).

Statistical analyses

For summarizing baseline characteristics, continuous parameters were expressed as medians with interquartile range (IQR) and categorical variables as frequencies with percentages. Correlations between the incidence of AKI in each cohort and the mortality of the entire cohort or patients with AKI were evaluated. If the regression line was statistically significant, squared correlation coefficient (R2), and regression coefficient (β) were calculated using the least-square approach weighted with the number of patients in each cohort. Subsequently, a subgroup analysis was conducted using ICU categories. Moreover, the association between the incidence and severity of AKI and that between the incidence of mild or severe AKI and the mortality were analyzed. All P values reported were two-sided, and values of P < 0.05 were considered to be statistically significant. We used the JMP Pro Version 13 software (SAS Institute, Cary, NC) for conducting all statistical analyses.


Systematic review

Evaluating the previous systematic review (8) yielded 22 cohorts that satisfied our eligibility criteria (Fig. 1). We subsequently conducted a literature search using MEDLINE, EMBASE, and Cochrane Library and identified 2,051; 1,786; and 552 potentially relevant reports, respectively. A total of 4,340 reports were excluded from the study because they met at least one of the exclusion criteria (Fig. 1). Two reports identified using MEDLINE included more than one independent eligible cohort in each literature; thus, a total of 49 reports identified produced 54 eligible ICU cohorts for analysis in our systematic review process. As a result, 76 cohorts published during 2007 to 2018 comprised 564,455 adult patients in ICU who were included in further analyses in the present study. The quality of the included manuscripts varied from point of 5 to 9 in the Newcastle-Ottawa scale for cohorts studies (16); none was classified as poor quality (points 0–3), 31 (43.7%) were of fair quality (4–6), and 40 (56.3%) were of good quality (7–9). A complete reference list of the cohorts is available in Table S3 (Supplemental Digital Content 3,

Fig. 1:
Search process and literature selection.

Baseline characteristics

The baseline characteristics of the cohorts included in this study are shown in Table 1. The median age of participants was 60.5 years, and 59.5% of the participants were men. The median incidence of AKI in each cohort was 35.6% (IQR, 25.9–50.8). The mortality of the entire cohort and patients with AKI in each cohort was 14.9% (IQR, 9.6–19.8) and 28.0% (IQR, 17.0–36.3), respectively. Cohorts from Europe, North America, and Asia accounted for 78.9% (60/76) of all the included cohorts. The proportion of patients in general ICU, which included both medical and surgical critically ill patients, was the highest (51/76, 67.1%) among the study cohorts.

Table 1:
Characteristics of included cohorts.

Incidence of AKI and mortality

The association between the incidence of AKI and the mortality in each cohort is represented in Fig. 2. The mortality of all patients in each cohort showed no correlation with the proportion of AKI (P = 0.73) (Fig. 2A). Conversely, the mortality of patients with AKI decreased as the incidence of AKI increased (R2 = 0.18, β = −0.25, P < 0.001) (Fig. 2B). Moreover, 51 cohorts in general ICU, which was the most frequent clinical setting, showed a similar association between the incidence of AKI and the mortality of AKI patients (R2 = 0.44, β = −0.38, P < 0.001) (Fig. 2C).

Fig. 2:
Correlations between the incidence of AKI and mortality.

We further investigated the association between the incidence of AKI in each cohort and the proportion of severe AKI in all patients with AKI. Data regarding mortality based on patients with mild or severe AKI were available in 51 cohorts. As shown in Fig. 3, the proportion of severe AKI was significantly higher in cohorts with a higher incidence of AKI than in those with a lower incidence of AKI (R2 = 0.11, β = 0.30, P = 0.006). Correlations between the incidence of AKI and the mortality of patients with mild or severe AKI are shown in Figure 4A and B. Both results reproduced the negative correlation between the incidence of AKI and the mortality (R2 = 0.27, β = −0.31, P < 0.001 for patients with mild AKI; R2 = 0.23, β = −0.28, P < 0.001 for patients with severe AKI).

Fig. 3:
Correlations between the incidence and severity of AKI.
Fig. 4:
Mortality of patients with mild or severe AKIs.


This systematic review included 564,455 adult patients in ICU from 76 cohorts and revealed that the mortality of patients with AKI decreased in ICU cohorts with an increase in the incidence of AKI. To the best of our knowledge, this is the first report in the relevant literature to report a negative correlation between the incidence and the mortality of AKI among patients in ICUs using a systematic review approach.

In our systematic review, the proportion of AKI in studies on critically ill patients showed considerable variation (5.9%–82.5%). This variety may be due to differences in the baseline characteristics of patients, the severity of underlying diseases, the etiology of AKI, and the process of management in each ICU (1). In the present study, the mortality of the entire cohort also varied and was not associated with the incidence of AKI in each cohort (Fig. 2A).

Nevertheless, the mortality of patients diagnosed with AKI was unexpectedly lower in cohorts with a higher incidence of AKI than in cohorts with lower incidence of AKI (Fig. 2B). Did the severity of AKI have an impact on these unexpected results in this study? The proportion of severe AKI over total AKI was relatively lower in cohorts with a low incidence of AKI (Fig. 3). The subgroup analysis on the mortality of the mild or severe AKI group also showed an inverse correlation between the incidence and mortality of patients with AKI (Fig. 4, A and B); therefore, AKI severity itself could not explain the inverse correlation. Alternatively, mild AKI showed a relatively higher mortality rate in the ICU cohort with a low incidence of AKI. This association may be because of the sufficient awareness of AKI that plays a role in reducing the mortality of patients with AKI. An ICU setting that provides adequate AKI awareness may have several advantages such as standardized AKI care bundles, especially for postoperative patients, and will provide favorable outcomes (14, 18); easy access of intensivists and physicians to nephrologists; and better management of AKI by the available critical care nephrologists (19). Although the effect of hospital- and case-volume on mortality of patients with AKI has not been investigated till date, several clinical studies have reported that patients with acute respiratory failure who require extracorporeal membrane oxygenation in larger hospitals with sufficient experience showed low mortality (20).

The prevention of AKI in critically ill patients is important in improving clinical outcomes. However, data obtained in this study suggested that using the incidence of AKI as a quality indicator for managing AKI is inappropriate because a high incidence of AKI was not associated with the mortality of patients with AKI. Recently, several trials that aimed to validate the effect of an e-alert system or a machine learning program were successful in realizing an early diagnosis of AKI and increasing AKI awareness by physicians (12, 13, 21). The present study suggests that AKI awareness contributes to the outcomes that can be measured by the incidence of AKI and the mortality of patients with AKI rather than the mortality of all patients in ICU.

There are several limitations to the present study. First, all the studies included in this review were observational studies and more than 50% were retrospective cohort studies. These study backgrounds do not exclude heterogeneity and potential confounders. Second, studies with a small number (< 500) of participants were excluded, resulting in the exclusion of randomized studies and most studies from developing countries. Finally, the assessment of a causal relationship between clinical factors and observed outcomes was not possible because all the analyzed studies were observational studies.

In conclusion, the incidence and mortality of patients with AKI in ICUs were remarkably different even with recently available standardized AKI criteria. The inverse association of the incidence and mortality of patients with AKI in ICU may indicate the advantages of high awareness, good management, and case-volume effect of ICUs with more experience in treating patients with AKI.


The authors thank Enago ( for the English-language review.


1. Hoste EAJ, Kellum JA, Selby NM, Zarbock A, Palevsky PM, Bagshaw SM, Goldstein SL, Cerda J, Chawla LS. Global epidemiology and outcomes of acute kidney injury. Nat Rev Nephrol 14 (10):607–625, 2018.
2. Bellomo R, Vaara ST, Kellum JA. How to improve the care of patients with acute kidney injury. Intensive Care Med 43:727–729, 2017.
3. Mehta RL, Cerda J, Burdmann EA, Tonelli M, Garcia-Garcia G, Jha V, Susantitaphong P, Rocco M, Vanholder R, Sever MS, et al. International Society of Nephrology's 0by25 initiative for acute kidney injury (zero preventable deaths by 2025): a human rights case for nephrology. Lancet 385:2616–2643, 2015.
4. Kellum JA, Prowle JR. Paradigms of acute kidney injury in the intensive care setting. Nat Rev Nephrol 14:217–230, 2018.
5. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. Acute renal failure-definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 8:R204–212, 2004.
6. Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, Levin A. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 11:R31, 2007.
7. Kidney Disease: Improving Global Outcomes (KDIGO). Clinical Practice Guideline for Acute Kidney Injury. Kidney Int Suppl 2:1–138, 2012.
8. Susantitaphong P, Cruz DN, Cerda J, Abulfaraj M, Alqahtani F, Koulouridis I, Jaber BL. World incidence of AKI: a meta-analysis. Clin J Am Soc Nephrol 8:1482–1493, 2013.
9. Nisula S, Kaukonen KM, Vaara ST, Korhonen AM, Poukkanen M, Karlsson S, Haapio M, Inkinen O, Parviainen I, Suojaranta-Ylinen R, et al. Incidence, risk factors and 90-day mortality of patients with acute kidney injury in Finnish intensive care units: the FINNAKI study. Intensive Care Med 39:420–428, 2013.
10. Hoste EA, Bagshaw SM, Bellomo R, Cely CM, Colman R, Cruz DN, Edipidis K, Forni LG, Gomersall CD, Govil D, et al. Epidemiology of acute kidney injury in critically ill patients: the multinational AKI-EPI study. Intensive Care Med 41:1411–1423, 2015.
11. Fujii T, Uchino S, Doi K, Sato T, Kawamura T. Diagnosis, management, and prognosis of patients with acute kidney injury in Japanese intensive care units: the JAKID study. J Crit Care 47:185–191, 2018.
12. Wilson FP, Shashaty M, Testani J, Aqeel I, Borovskiy Y, Ellenberg SS, Feldman HI, Fernandez H, Gitelman Y, Lin J, et al. Automated, electronic alerts for acute kidney injury: a single-blind, parallel-group, randomised controlled trial. Lancet 385:1966–1974, 2015.
13. Flechet M, Guiza F, Schetz M, Wouters P, Vanhorebeek I, Derese I, Gunst J, Spriet I, Casaer M, Van den Berghe G, et al. AKIpredictor, an online prognostic calculator for acute kidney injury in adult critically ill patients: development, validation and comparison to serum neutrophil gelatinase-associated lipocalin. Intensive Care Med 43:764–773, 2017.
14. Meersch M, Schmidt C, Hoffmeier A, Van Aken H, Wempe C, Gerss J, Zarbock A. Prevention of cardiac surgery-associated AKI by implementing the KDIGO guidelines in high risk patients identified by biomarkers: the PrevAKI randomized controlled trial. Intensive Care Med 43:1551–1561, 2017.
15. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 6:e1000097, 2009.
16. Wells GA, Shea B, O’Connell D, Peterson J, Welch V, Losos M, Tugwell P. The Newcastle-Ottawa Scale (NOS) for assessing the quality if nonrandomized studies in meta-analyses. Available at: Accessed October 30, 2019.
17. Yang X, Chen C, Teng S, Fu X, Zha Y, Liu H, Wang L, Tian J, Zhang X, Liu Y, et al. Urinary matrix metalloproteinase-7 predicts severe AKI and poor outcomes after cardiac surgery. J Am Soc Nephrol 28:3373–3382, 2017.
18. Kolhe NV, Reilly T, Leung J, Fluck RJ, Swinscoe KE, Selby NM, Taal MW. A simple care bundle for use in acute kidney injury: a propensity score-matched cohort study. Nephrol Dial Transplant 31:1846–1854, 2016.
19. Askenazi DJ, Heung M, Connor MJ Jr, Basu RK, Cerda J, Doi K, Koyner JL, Bihorac A, Golestaneh L, Vijayan A, et al. Optimal role of the nephrologist in the intensive care unit. Blood Purif 43:68–77, 2017.
20. Barbaro RP, Odetola FO, Kidwell KM, Paden ML, Bartlett RH, Davis MM, Annich GM. Association of hospital-level volume of extracorporeal membrane oxygenation cases and mortality. Analysis of the extracorporeal life support organization registry. Am J Respir Crit Care Med 191:894–901, 2015.
21. Holmes J, Rainer T, Geen J, Roberts G, May K, Wilson N, Williams JD, Phillips AO. Acute kidney injury in the era of the AKI E-Alert. Clin J Am Soc Nephrol 11:2123–2131, 2016.

Acute kidney injury; acute renal failure; case-volume effect; intensive care unit; systematic review; AKI; acute kidney injury; AKIN; acute kidney injury network; ICU; intensive care unit; IQR; interquartile range; KDIGO; Kidney Disease, Improving Global Outcomes; RIFLE; risk, injury, failure, loss of kidney function, End-stage kidney disease

Supplemental Digital Content

Copyright © 2020 by the Shock Society