The Effects of Intensive Versus Routine Treatment in Patients with Acute Kidney Injury

An explorative randomized controlled study

Background: In patients with acute kidney injury (AKI), specialized treatment—initiated in response to an early-warning system—may be beneficial compared with routine treatment.

Method: To explore effect estimators in a pilot trial (DRKS00010530), patients with AKI on regular wards of a university hospital were treated either in the usual way (control group) or more intensively (intervention group). The subjects were allotted randomly to the two treatment groups. The more intensive treatment consisted of an early warning system for a rise in the serum creatinine concentration, immediate consultation of a specialist, and the issuance of a patient kidney passport. The primary endpoint was recovery of renal function after AKI during the index hospitalization. Renal complications and process indicators were the secondary endpoints.

Results: The proportion of patients whose renal function returned to baseline after AKI was 50% in the intervention group (N = 26) and 42% in the control group (N = 26) (odds ratio 1.4, 95% confidence interval [0.5; 4.0], p = 0.58). The calculated glomerular filtration rate went down, from hospital admission to discharge, by 3 mL/min/1.73 m² (1st–3rd quartile: [6; –20]) in the intervention group and by 13 mL/min/1.73 m² in the control group (1st–3rd quartile: [0; –25]; p = 0.09). Complications of AKI such as hyperkalemia, pulmonary edema, and renal acidosis were rarer in the intervention group (15% versus 39%; p = 0.03). In the intervention group, compared with the control group, the cause of AKI was identified more frequently (27% versus 4%; p = 0.05); drugs with relevance to the kidney were discontinued more frequently (65% versus 31%; p = 0.01); and the diagnosis of AKI was more frequently documented in the patient’s chart (58% versus 37%; p = 0.03).

Conclusion: Specialized consultations supported by an early warning system for AKI seem to be beneficial for patients. The findings of this pilot trial should be verified in larger-scale randomized controlled trials.

With an incidence of of 4–21%, acute kidney injury (AKI; acute kidney failure N 17.9 in the ICD classification) is one of the most frequently occurring and cost-intensive acute diseases in inpatients (1, 2, 3, 4, 5, 6). Patients with AKI more frequently develop chronic renal failure (hazard ratio: 2.7 [7]) or terminal renal failure (hazard ratio: 4.8 [7]). The prevalence of chronic renal failure in Germany is estimated at around 2 million persons [8]. Cardiorenal complications, including death, chronic dialysis, decline in renal function, heart failure, and stroke, occur more frequently following AKI than after myocardial infarction (hazard ratio 1.4 [9]). In particular, undetected AKI, or delayed treatment of AKI, is an independent risk factor for increased hospital mortality (1). Extrapolating the data from cost surveys in England and Canada (10, 11), approximately € 1 billion is spent annually on the treatment of AKI and AKI-related complications in Germany.

The specialist medical societies are calling for urgent multimodal treatment, as part of which triggering factors as well as complications are promptly detected and resolved (12, 13). A specialist can be consulted for expert advice on treatment. Outpatient follow-up should be initiated. Nevertheless, there are deficits in terms of diagnosis and acute treatment. In approximately 75% of cases the diagnosis is not made, and treatment is often delayed by a matter of days. Fewer than 10% of patients receive specialist co-treatment (1, 14, 15, 16). In a survey of physicians involved in the treatment of patients with AKI at a total of 423 hospitals, nephrologists were involved in the most severe disease manifestation, i.e., AKI requiring dialysis, in only 22% of cases (17). The diagnosis of AKI can be challenging, since the diagnostic relevance of an increase in creatinine needs to be evaluated not only on the basis of two different cut-off values of creatinine concentration, but also taking into account two different time windows. This is further complicated by the often initially asymptomatic course. AKI early warning systems, in combination with specialist co-treatment if indicated, represent a possible means to support physicians in the diagnosis and treatment of AKI. A short summary of the most important effects of AKI early warning systems in randomized controlled studies can be found in eTable 1.

eTable 1

Results of randomized controlled studies on inpatient early warning systems for acute kidney injury

This explorative randomized controlled study evaluated the variability of effect estimates and feasibility indicators of intensified treatment compared with routine treatment of patients with AKI.

Patients and methods

Search algorithm to detect acute kidney injury

The preliminary study and the explorative randomized study used an AKI search algorithm programmed in a hospital data system. The algorithm was based on the assessment of at least two serum creatinine levels and identified an increase according to the recommendations of the Kidney Disease Improving Global Outcomes (KDIGO) initiative (12). The algorithm also detected an increase of >50% of the previous level within a maximum of 7 days or of >0.3 mg/dL (>26.4 μmol/L) within a maximum of 2 days or of >0.5 mg/dL (>44 μmol/L) to >4.0 mg/dL (>354 μmol/L). Diuresis was not taken into account for AKI detection.

Study design

Preliminary study

In a preparatory step, the following parameters were recorded over an 8-month study period at Magdeburg University Hospital, Magdeburg, Germany, which has 1130 inpatient beds and approximately 45 000 inpatient cases per year:

• Incidence
• Onset and resolution (date/time) of AKI
• The distribution of affected patients across specialist departments and regular units
• Status in terms of specialist consultation

Explorative randomized study

The study design and approach used in the randomized study are described in the eBox. The measures undertaken in the context of specialist consultations followed a standardized protocol based on the KDIGO recommendations (12). The corresponding measures are listed in Box 1.

Box 1

Specialist consultation for acute kidney injury

eBox

Detailed description of the explorative randomized study

Ethical evaluation and registration

The relevant ethics committee was informed in advance about the data collection to be carried out in the preliminary study; approval from the ethics committee and written patient consent were obtained for participation in the explorative randomized study (EK 171/13, EK 61/14). The randomized study was registered in the German Register of Clinical Trials (DRKS00010530).

Study population of the explorative randomized trial

Male and female patients aged 18 years or more who were treated on regular units for AKI and provided written consent were included in the study. The exclusion criteria were chronic dialysis dependence, kidney transplantation, treatment on a nephrology or intensive care unit, infection with human immunodeficiency virus (HIV) or hepatitis viruses, pregnancy, and participation in a drug or medical device trial.

Endpoints

The primary endpoint of the explorative randomized study was restoration of renal function following AKI, defined as the proportion of patients that regained baseline renal function, or as the change in estimated glomerular filtration rate (eGFR) between admission and discharge during the index hospital stay. Attainment of baseline renal function was defined as an increase in eGFR to at least 90% of baseline by the time of hospital discharge (6). The secondary endpoints were kidney-related complications and process indicators of clinical care and the time requirements of the study (Box 2).

Box 2

Secondary endpoints of the explorative randomized study

Statistical analysis

The normal distribution assumption was checked using histograms. In the case of normal distribution, the mean and standard deviation were reported; otherwise, the median and the distance between the 1^st and 3^rd quartiles were given. Student’s t-test was used for parametric two-group comparisons and the Mann–Whitney U test for non-parametric two-group comparisons. The chi-squared test was used for dichotomous variables and Fisher’s exact test for two-group comparisons. Since this was an explorative study, p-values were used as descriptive p-values and not adjusted for multiplicity.

Results

Preliminary study

Over the course of the preliminary study, 30 367 patients received inpatient treatment. The incidence of AKI was 4.2% (1267/30 367 patients). Around half of the AKI patients (49.6%) were treated on regular units (eTable 2). The highest number of these were treated in specialist surgical departments (42.0%), followed by internal medicine (36.7%) and specialist neurology departments (9.5%) (eFigure 2a). In over 85% of patients, the diagnostically indicative rise in serum creatinine occurred between 8 am and 3 pm (eFigure 2b). On regular units, specialist consultation was requested in 8.4% of cases (53/629). Requests for specialist consultation were made on average 2.8 days after the diagnostically indicative rise in serum creatinine concentration.

Figure 2

Primary endpoint: renal function recovery

eFigure 1

Patient medical ID card for acute kidney injury (kidney passport)

eFigure 2

Findings of the preliminary study preceding the randomized pilot study

eTable 2

Preliminary study

Explorative randomized study: patient flow and characterization

The AKI early warning system detected 96 patients with a diagnostic increase in serum creatinine (Figure 1). After excluding 44 patients, 52 patients with AKI were included. The study groups were comparable in terms of demography, comorbidities, admitting specialist department, status on admission, renal function at time of hospital admission, AKI stage, and range of triggering factors (eTable 3). The average time required for screening, randomization, and patient information was 28.7 ± 3.7 min. Data documentation was completed in 147.9 ± 32.9 min per patient.

Figure 1

Flow diagram for patients in the randomized pilot study

eTable 3

Patient characteristics and intervention

Explorative randomized study: course of renal function

In all, 50% of patients in the intervention group and 42% of patients in the control group regained their baseline renal function (odds ratio with control group as reference: 1.4; 95% confidence interval [0.5; 4.0], p = 0.58) (Figure 2a). From hospital admission to discharge, eGFR fell by –3 mL/min/1.73 m² (1^st–3^rd quartile: [6; −20]) in the intervention group and by –13 mL/min/1.73 m² (1^st–3^rd quartile: [0; −25]; p = 0.09) in the control group (Figure 2b). AKI-related complications such as hyperkalemia, pulmonary edema, and renal acidosis occurred less frequently in the intervention group than in the control group (p = 0.03) (Figure 3), with this group difference being due mainly to the higher incidence of hyperkalemia in the control group (Table). Patients with AKI-related complications had stage-1 AKI in half of the cases (Table). Three of 11 patients (27.3%) in the intervention group and three of eight patients (37.5%) in the control group developed newly diagnosed chronic renal failure or worsening of previously known chronic renal failure within 1 year of AKI (p >0.99). Seven patients in each of the two study groups were admitted as inpatients within 30 days of discharge. Here, recurrence of AKI was seen in no patients of the intervention group and in four patients of the control group (Table).

Figure 3

Secondary endpoints

Table

Secondary endpoints: measures, course, and documentation

Explorative randomized study: documentation and measures

Patients in the intervention group received prompt specialist consultation more frequently than those in the control group (65% versus 4%, p <0.001) (eTable 3). The median time of consultation was the day of AKI onset in the intervention group and 2 days following AKI onset in the control group (p = 0.003) (eTable 3). The diagnosis of AKI was noted more frequently in the records of patients receiving intensified treatment than for those receiving routine treatment (Figure 2). Triggering factors for AKI remained undocumented and untreated less often in the intervention group than in the control group (4% versus 27%, p = 0.05).

In the intervention group, non-steroidal anti-inflammatory drugs (NSAID) were discontinued more frequently and renin–angiotensin–aldosterone system inhibitors suspended more often than in the control group (Figure 2). The recommendation for specialist outpatient follow-up was made in 13.6% of patients receiving intensified treatment, but in none of the patients receiving routine treatment (Table). There were no differences between the two groups in terms of documentation of diuresis and body weight, AKI diagnosis in the medical discharge report, AKI coding (ICD N17.9x), or the recommendation for primary care monitoring of renal function.

Discussion

Following a retrospective preliminary study, patients with AKI treated on regular units were randomly assigned to receive either intensified treatment or routine treatment. The preliminary study found relevant patient numbers and endpoints, as well as a balanced distribution of patients across regular units in relation to the total number of patients treated. AKI onset predominantly occurred during core working hours and a specialist consultation was rarely requested.

The randomized study groups appeared to be patient cohorts of similar structure. Specialist consultations took place in the intervention group significantly more often, and 2 days earlier, than in the control group.

The proportion of patients that regained their baseline renal function was 50% and 42% for intensified and routine treatment, respectively. From hospital admission to discharge, eGFR fell in the intervention group by –3 mL/min/1.73 m² and in the control group by −13 mL/min/1.73 m². Drugs were adjusted more frequently and AKI-related complications seen less often in patients in the intervention group than in the control group: in addition, causes of AKI were less likely to remain undocumented with intensified treatment. The diagnosis of acute renal injury was documented more frequently in patient records in the intervention group than in the control group. Cross-sectoral communication about AKI was suboptimal in both study groups.

A non-randomized study conducted in the USA without AKI early warning systems showed that AKI was less likely to progress to moderate or severe stages if specialist support was provided (3.3%) than with routine treatment (12.9%) (18). A systematic review found a higher number of nephroprotective measures in non-randomized original studies in which AKI early warning systems were linked to clinical action (2). A randomized study without treatment recommendations found no effects for an AKI early warning system, but demonstrated pronounced treatment deficits, including continuation of NSAID administration in patients with AKI (19). A Korean non-randomized quality assurance study reported a higher rate of renal function recovery in patients whose management included an AKI early warning system and thus the associated specialist co-treatment (20). A stepped-wedge cluster-randomized study conducted in the UK found hospital stays to be reduced by approximately 1 day for AKI patients when the AKI early warning system was combined with a reference to a bundle of clinical measures (21).

Recent publications have provided epidemiological data, described and quantified the treatment deficit, and proposed solutions (2, 16, 17). Information on the feasibility and potential effects of the intensified co-treatment of patients with AKI—triggered by an AKI early warning system—is not available for the German-speaking countries. Recent clinical studies showing proof of effect for a specialist intervention were not randomized (18, 20), did not use an AKI early warning system (18), or did not carry out any specialist co-treatment (19, 21, 22). No randomized study has reported the occurrence of complications related to AKI or the long-term development of renal function.

To our knowledge, the present explorative study, conducted using a randomized design, is the first to investigate the effects of an AKI early warning system and prompt specialist co-treatment of patients on a regular unit. We focused on identifying feasibility indicators and on obtaining data on the variability as well as the frequency and distribution of patient-relevant effect estimates. Due to structural differences in healthcare systems, it is important to have data on treatment outcomes for AKI in the German-speaking countries.

In agreement with the lower limit of around 4% reported in the literature on AKI incidence (4), AKI was diagnosed at a rate of approximately 15–20 patients per week on regular units during the preliminary study and the randomized study. In more than 85% of cases, AKI onset occurred during core working hours, and specialist co-treatment took place as part of routine care in less than 10% of cases. Thus, intensified treatment appears to be possible at comparable hospitals. At hospitals without a specialist nephrology department, intensified co-treatment can be guaranteed at least to a certain extent by means of a care bundle stored in the hospital’s information system and triggered by an AKI early warning system, as well as by continuous training (23, 24).

Taking into consideration the complementary improvement of renal endpoints, intensified treatment for AKI appears to confer a patient-relevant benefit. Given the limitations of these explorative data, no alternative explanation of the findings presents itself— with the exception of a possible α error. At a somewhat lower disease burden overall, routine treatment showed no benefit with regard to the endpoint.

This present study did not permit any conclusions at the level of statistical significance. However, the observations that baseline renal function was regained approximately 20% more often in the intensified treatment group, and that the eGFR loss was around 10 mL/min/1.73 m² lower, support clinical implications and point to the prospect of demonstrating the efficacy of the described intervention in a follow-up study with sufficient power. No comparable effects have yet been achieved in drug trials on acute or progressive renal insufficiency (25, 26, 27). On the basis of the available data, with an α error of 5% for a double-sided test and a power of 90% per group, 840 patients with AKI would be needed to demonstrate a higher proportion of patients regaining baseline renal function under intensified treatment (50%) than with routine treatment (42%). AKI-related complications are often observed as early as stage 1. This supports the call to implement measures for the prompt initiation of treatment at all stages of AKI.

The costs in terms of time, effort, and expense of specialist care in patients with AKI appear to be justified by the frequency and effects of the syndrome, and by the better treatment outcomes anticipated. Furthermore, we assume that the clinical benefit of specialist co-treatment is likely to more than compensate for the additional costs it causes. Patient kidney passports, information brochures, and measures aimed at effective cross-sectoral communication supported by the medical discharge report can serve to optimize inter-sectoral communication.

The internal validity of the results presented here is supported by the retrospective preliminary study with an epidemiological survey of AKI at the university hospital, together with the quantification of treatment gaps (16) and a review article on AKI early warning systems as a possible aid to improving diagnosis (2). On the basis of this, in the present randomized study we correctly and promptly identified and included the—even in the routine situation—heterogeneous patient collective. The randomization in this study resulted in a relatively even distribution of demographic variables and comorbidities. Data analysis ensued in blinded fashion; unblinding using the sealed randomization list occurred only after study completion. A significant intervention gradient emerged between the two study groups, since many more patients in the intervention group received prompt specialist co-treatment than in the control group. Nevertheless, these measures could not always be implemented in the intervention group, although it was not possible to determine the reasons for this in detail. It cannot be ruled out that the nephrologist on duty prioritized a different patient-related task instead of immediately carrying out an AKI early warning system-triggered consultation (intervention group). Since AKI was documented in the patient records in only 58% of cases in the intervention group, one can assume an additional intervention effect if all cases were to be documented and consultations carried out. This randomized study had a small sample, which was oriented towards the objective of describing feasibility indicators and the variability of possible effect estimates for the purposes of conducting a randomized controlled follow-up study with sufficient power.

The therapeutic effects demonstrated in randomized studies (eTable 1) could be further improved upon by using a continuous AKI-prediction algorithm in real time in fully networked electronic health records, as recently presented by Tomašev et al. (28), on the basis of the primary preventive approach.

Conclusions

Within the constraints of the study limitations, the results of this explorative randomized investigation describe the feasibility and effects of specialist-supported AKI early warning systems on regular units. Studies with sufficient power are needed to demonstrate proof of efficacy for intensified treatment.

Funding

This investigation was funded by the B. Braun Foundation and the Dr. Werner Jackstädt Foundation.

Acknowledgments

The authors would like to thank Prof. Dr. J. Wippermann (Department of Cardiac and Thoracic Surgery), Prof. Dr. F. Walcher (Department of Trauma Surgery), Prof. Dr. R. Braun-Dullaeus (Department of Cardiology), and Dr. E. Lücke (Department of Respiratory Medicine), all Otto-von-Guericke Universität Magdeburg, Prof. Dr. C. Bruns (Department of General, Visceral, and Oncological Surgery, Cologne University Hospital), and Prof. Dr. I. Kutschka (Department of Cardiac, Thoracic, and Vascular Surgery, Göttingen University Hospital) for their suggestions during the planning of the study and their willingness to facilitate patient recruitment, as well as Mrs. J. Czupajllo (Charité – University Medical Center Berlin), Mrs. R. Iwers (Department of Cardiology, Brandenburg Heart Center, Immanuel Hospital, Bernau and Brandenburg Medical School Theodor Fontane), and Dr. A. Albert (Department of Nephrology and Endocrinology, Ernst von Bergmann Hospital, Potsdam and Diaverum Renal Care Center Am Neuen Garten, Potsdam) for their support in terms of data acquisition.

Conflict of interest statement
PD Dr. Haase-Fielitz is editor of a book on the subject and has received honoraria from Springer-Verlag. She has received research support (third party funding) from the B. Braun Foundation Melsungen and the Dr. Werner Jackstädt Foundation Wuppertal.

Prof. Robra has received research support (third party funding) from the B. Braun Foundation Melsungen.

Prof. Haase is editor of a book on the subject and has received honoraria from Springer-Verlag. He has received lecture honoraria from Abbott Diagnostics.

The remaining authors declare that no conflict of interest exists.

Manuscript received on 17 October 2019, revised version accepted on
10 February 2020

Translated from the original German by Christine Rye

Corresponding author
Prof. Dr. med. Michael Haase
Medizinische Fakultät der Otto-von-Guericke Universität Magdeburg
Leipziger Str. 44, 39120 Magdeburg, Germany

Cite this as:
Haase-Fielitz A, Elitok S, Schostak M, Ernst M, Isermann B, Albert C, Robra BP, Kribben A, Haase M: The effects of intensive versus routine treatment in patients with acute kidney injury—an explorative randomized controlled study.
Dtsch Arztebl Int 2020; 117: 289–96. DOI: 10.3238/arztebl.2020.0289

►Supplementary material

For eReferences please refer to:
www.aerzteblatt-international.de/ref1720

eFigures, eTables, eBoxes:
www.aerzteblatt-international.de/20m0289

References

1. Khadzhynov D, Schmidt D, Hardt J, et al.: The incidence of acute kidney injury and associated hospital mortality. Dtsch Arztebl Int 2019; 116: 397–404

2. Haase M, Kribben A, Zidek W, et al.: Electronic alerts for acute kidney injury. Dtsch Arztebl Int 2017; 114: 1–8

3. Bienholz A, Kribben A: Acute kidney injury: a clinical syndrome. Internist (Berl) 2016; 57: 983–93

4. Meier P, Bonfils RM, Vogt B, Burnand B, Burnier M: Referral patterns and outcomes in noncritically ill patients with hospital-acquired acute kidney injury. Clin J Am Soc Nephrol 2011; 6: 2215–25

5. Chawla LS, Amdur RL, Amodeo S, Kimmel PL, Palant CE: The severity of acute kidney injury predicts progression to chronic kidney disease. Kidney Int 2011; 79: 1361–9 MEDLINE

6. Bucaloiu ID, Kirchner HL, Norfolk ER, Hartle JE 2nd, Perkins RM: Increased risk of death and de novo chronic kidney disease following reversible acute kidney injury. Kidney Int 2012; 81: 477–85

7. See EJ, Jayasinghe K, Glassford N, et al.: Long-term risk of adverse outcomes after acute kidney injury: a systematic review and meta-analysis of cohort studies using consensus definitions of exposure. Kidney Int 2019; 95: 160–72

8. Girndt M, Trocchi P, Scheidt-Nave C, Markau S, Stang A: The prevalence of renal failure. Results from the German Health Interview and Examination Survey for Adults, 2008–2011 (DEGS1). Dtsch Arztebl Int 2016; 113: 85–91

9. Chawla LS, Amdur RL, Shaw AD, Faselis C, Palant CE, Kimmel PL: Association between AKI and long-term renal and cardiovascular outcomes in United States veterans. Clin J Am Soc Nephrol 2014; 9: 448–56

10. Collister D, Pannu N, Ye F, et al.: Health care costs associated with AKI. Clin J Am Soc Nephrol 2017; 12: 1733–43

11. Kerr M, Bedford M, Matthews B, O‘Donoghue D: The economic impact of acute kidney injury in England. Nephrol Dial Transplant 2014; 29: 1362–8

12. KDIGO: Clinical practice guideline for acute kidney injury. Kidney Int Suppl 2012. Doi: 10.1038/kisup.2012. www.kdigo.org/wp-content/uploads/2016/10/KDIGO-2012-AKI-Guideline-English.pdf (last accessed on 26 February 2020).

13. Alscher MD, Erley C, Kuhlmann MK: Acute renal failure of nosocomial origin. Dtsch Arztebl Int 2019; 116: 149–58

14. Yang L, Xing G, Wang L, et al.: Acute kidney injury in China: a cross-sectional survey. Lancet 2015; 386: 1465–71

15. Stewart J, Findlay G, Smith N, Kelly K, Mason M: Adding insult to injury: a review of the care of patients who died in hospital with a primary diagnosis of acute kidney injury (acute renal failure). A report of the National Confidential Enquiry into Patient Outcome and Death (2009). London: National Confidential Enquiry into Patient Outcome and Death 2009. www.ncepod.org.uk/2009report1/Downloads/AKI_report.pdf (last accessed on 26. February 2020).

16. Haase-Fielitz A, Ernst M, Lehmanski F, et al.: Treatment, clinical course, and cross-sectoral information transmission in patients with acute-on-chronic kidney injury. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2019; 62: 773–81

17. Schindler R, Hutagalung R, Jörres A, et al.: Treatment of acute renal failure in Germany: a structural analysis. Dtsch Med Wochenschr 2014; 139: 1701–6

18. Balasubramanian G, Al-Aly Z, Moiz A, et al.: Early nephrologist involvement in hospital-acquired acute kidney injury: a pilot study. Am J Kidney Dis 2011; 57: 228–34

19. Wilson FP, Shashaty M, Testani J, et al.: Automated, electronic alerts for acute kidney injury: a single-blind, parallel-group, randomised controlled trial. Lancet 2015; 385: 1966–74

20. Park S, Baek SH, Ahn S, et al.: Impact of electronic acute kidney injury (AKI) alerts with automated nephrologist consultation on detection and severity of AKI: a quality improvement study. Am J Kidney Dis 2018; 71: 9–19

21. Selby NM, Casula A, Lamming L, et al.: An organizational-level program of intervention for AKI: a pragmatic stepped wedge cluster randomized trial. J Am Soc Nephrol 2019; 30: 505–15

22. Wu Y, Chen Y, Li S, et al.: Value of electronic alerts for acute kidney injury in high-risk wards: a pilot randomized controlled trial. Int Urol Nephrol 2018; 50: 1483–8

23. Kolhe NV, Staples D, Reilly T, et al.: Impact of compliance with a care bundle on acute kidney injury outcomes: a prospective observational study. PLoS One 2015; 10: e0132279

24. Chandrasekar T, Sharma A, Tennent L, Wong C, Chamberlain P, Abraham KA: A whole system approach to improving mortality associated with acute kidney injury. QJM. 2017; 110: 657–66

25. de Zeeuw D, Akizawa T, Audhya P, et al.: Bardoxolone methyl in type 2 diabetes and stage 4 chronic kidney disease. N Engl J Med 2013; 369: 2492–503

26. Perkovic V, Jardine MJ, Neal B, et al.: Canagliflozin and renal outcomes in type 2 diabetes and nephropathy. N Engl J Med 2019; 380: 2295–306

27. Jo YY, Kim JY, Lee JY, Choi CH, Chang YJ, Kwak HJ: The effect of intraoperative dexmedetomidine on acute kidney injury after pediatric congenital heart surgery: A prospective randomized trial. Medicine (Baltimore) 2017; 96: e7480

28. Tomašev N, Glorot X, Rae JW, et al.: A clinically applicable approach to continuous prediction of future acute kidney injury. Nature 2019; 572: 116–9