Diagnostic systems to warn of AKI

Acute kidney injury (AKI) is a key challenge for healthcare systems in developed countries with ageing populations and current systems for its earlier detection are discussed

Tarek Abdelaziz MRCP

Jyoti Bahrani FRCP

Mark Thomas FRCP

Department of Renal Medicine,

Heart of England Foundation Trust,

Birmingham, UK

Acute kidney injury (AKI) is the Cinderella of modern healthcare. Its ‘Cinderella’ nature is demonstrated by the number of excess deaths due to AKI in England alone – more than 40,000 per annum.1 This compares, for example, to about 35,000 deaths yearly in the entire UK from lung cancer, yet many health policy makers or members of the public have not even heard of AKI. Patients die due to AKI and not just from their comorbid conditions.2

The condition is known to cause increased mortality, length of stay and inpatient costs. Between 15% and 20% of all hospitalised patients develop AKI.3 Across all three stages, about 25% of patients with AKI will die.2 AKI-related inpatient care alone (ignoring outpatient costs) is estimated at 1% of the total NHS budget.1 AKI occurs in different medical and surgical settings, being caused by common factors such as hypovolaemia, dehydration, sepsis and drugs.4 The term AKI was introduced to replace the old term acute renal failure (ARF), emphasising that moderate derangement of kidney function often occurs without frank renal failure and the need for dialysis.

Until very recently, care of patients with AKI has often been very poor.5 In particular, in the UK the National Confidential Enquiry into Patient Outcome and Death (NCEPOD)5 in 2009 showed poor care for AKI, leading to a sea change in attitudes to AKI. This report also led to the development of NICE guidelines on AKI4 and a national programme aimed at improving  AKI care in the UK. 

The definition of AKI

There is a clear need to have a single pragmatic definition of AKI to enable clinicians and researchers to unify the management across different healthcare systems. In 2011, the International Kidney Disease: Improving Global Outcome (KDIGO) group merged the previous sister definitions RIFLE and AKIN, to produce the current definition, as shown in Table 1.6 Creatinine is a by-product of muscle metabolism, filtered from the blood by the kidneys and excreted in the urine. Its blood concentration is used to estimate the level of kidney filtration (and therefore function).

This is called the glomerular filtration rate (GFR). At the onset of AKI, assuming a simple single insult to the kidneys such as a period of ischaemia, the GFR will drop rapidly to a very low level. Serum creatinine then rises over hours to days. Some AKI patients sustain multiple renal insults over a prolonged period, complicating the assessment of renal function. The creatinine rise used to detect AKI is measured as a percentage or an absolute rise. The complexities of using changes in creatinine levels to diagnose AKI are discussed in detail elsewhere.2 Therefore, looking at the KDIGO definition of AKI (Table 1), it can be seen that a sudden rise in creatinine, either in absolute terms (≥26µmol/l within 48 hours) or in percentage terms (≥50% within seven days) is used to detect AKI.2 Note that AKI can be ‘detected’ as soon as the creatinine rise reaches the specified level. The maximum extent of AKI or stage can only be identified retrospectively when the creatinine value peaks.

Systems to detect the sick patient with or at risk of AKI

Early warning scores (EWS)

There has been considerable work in critical care to develop and implement physiological track and trigger systems for hospital inpatients, using patient observations or vital signs.7 In the care of the deteriorating patient with abnormal signs these trigger the ‘afferent limb’ or warning mechanism. The need to have an effective response or ‘efferent limb’ led to the development of medical emergency teams or critical care outreach teams. Systematic reviews have shown that such rapid response teams reduce cardiac arrests. Although they have been widely adopted, there is in fact no clear evidence that they reduce hospital mortality in adults.7–9

EWS and AKI

EWS might have two roles in the field – as part of a system to detect patients at risk of AKI, or to prevent deterioration in patients with AKI. A combination of age, certain vital signs (respiratory rate and disturbed consciousness) and the presence of certain clinical conditions (chronic kidney disease, diabetes mellitus, congestive cardiac failure and liver disease) can identify patients at high risk of AKI.10 Electronic vital signs monitoring may in future combine with the electronic patient record and lab testing to reduce the occurrence and severity of AKI. However, at present there is little or no evidence of the cost-effectiveness of either electronic vital signs monitoring or rapid response systems.7

Oliguria in AKI

In AKI there is a fall in urine output (oliguria). Oliguria is not easy to detect in a patient on a busy hospital ward, particularly if they do not have a urinary catheter. Oliguria was traditionally defined as a urine output of less than 400ml/day, equivalent to 0.24ml/kg/hr in a 70kg patient. The recent KDIGO definition of oliguria is a urine output of <0.5ml/kg/hr for more than six hours.6 It is important to monitor urine output in certain AKI patients, as reversal of any hypovolaemia guided by the urine output monitoring will prevent further worsening of AKI. Various electronic urine output-monitoring systems are available for critical care use. They have advantages over the manual measurement. A crucial issue is the time such systems take to detect a drop in urine output. Obviously hourly ‘manual’ measurement in critical care will take at least one hour to detect a change. Electronic measurement in critical care may allow detection in less than 15 minutes.11 This may improve the timeliness of response to hypovolaemia. 

Electronic alerts in AKI 

The concept of electronic alerts utilises information technology to automatically detect the rises in creatinine, based on the KDIGO criteria.6 Alert systems rapidly detect AKI, unlike doctors who often may be slow to register  changes in creatinine.12 An alert system will automatically compare the present result to the archived results of the patient, as far back as the previous year.13 Alert use has been successful in changing clinician behaviour and outcome, with reductions in nephrotoxic medication use in AKI.14,15 Moreover, the use of alerts in critical care has successfully increased interventions in AKI.16 Options for warnings include: 

(a) A passive alert, such as a warning statement with AKI stage as an addendum to the creatinine result, which might contain a hyperlink to the local AKI clinical guideline (Figure 1).

(b) An interruptive alert requiring the clinician to acknowledge the alert.

(c) An active alert where a direct response is triggered, such as an outreach team visit. 

Some alert studies have failed to change clinician behaviour,17 and furthermore, clinicians may defer a response to alerts.18 In a study of nephrotoxic medication use in AKI patients, interruptive alert warnings were deferred a median of four times prior to a definitive response. Alert fatigue remains a concern for clinicians assailed by a barrage of alerts for various conditions within IT systems. A standardised algorithm for automated KDIGO-like staging is currently undergoing mandatory implementation in the UK. Modern pathology laboratories in the developed world should also be implementing this change.

New ways of organising care of AKI

The use of alerts reduces the diagnostic time lag in AKI. As with early warning scores that trigger a warning (above), the key issue is what happens in response to an alert. Do members of the healthcare team act on such alerts? Research is underway to determine the best way to optimise the response. Strategies include education, promotion of guidelines or early specialist input (outreach). We19 as well as others20 have piloted the use of outreach teams in AKI. Further research into their use is ongoing. An example of successful treatment of AKI is shown in Figure 2.

Novel biomarkers of AKI

There is much expectation that biomarkers will be incorporated into definitions of AKI as well as routine clinical practice. However, some degree of caution is warranted.21 It has been suggested that AKI should be designated a ‘kidney attack’, analogous to heart attack,22 and that biomarkers may be used to detect AKI earlier in its course, similar to the use of cardiac troponin in the detection of myocardial infarction. 

Unfortunately, the analogy between the conditions disappears when interventions are considered. Myocardial infarction has specific treatments. Even if a biomarker reliably shows early AKI, there are currently no specific therapies for ischaemic or septic tubular injury. Demonstrating the cost effectiveness of biomarker use in addition to creatinine and urine output monitoring presents a further challenge. New biomarkers will need to demonstrate ‘added value’, over and above optimal patient care and use of traditional AKI markers (creatinine and oliguria).

Conclusions

AKI has a major impact in healthcare: it has the dual features of being common and causing an increase in mortality. It clearly increases healthcare costs. Early detection and intervention has the potential to reduce the damage caused by AKI. The early detection may be mediated by electronic alert systems using creatinine measurements, other information in the electronic patient record, electronic urine output monitoring or biomarkers. The warning from any of these systems needs to trigger an appropriate response. Developments in the coming years should refine both the AKI warning and the subsequent response systems, so as to improve outcomes from this important condition.

References 

1. Kerr M et al. The economic impact of acute kidney injury in England. Nephrol Dial Transplant 2014;29(7):1362–8.
2. Thomas ME et al. The definition of acute kidney injury and its use in clinical practice. Kidney Int 2015;87:62–73; doi:10.1038/ki.2014.328; published online 15 October 2014.
3. Bedford M et al. What is the real impact of acute kidney injury? BMC Nephrol 2014;15:95.
4. Acute kidney injury Prevention, detection and management of acute kidney injury up to the point of renal replacement therapy. Nice guidelines (CG169). Available at: http://www.nice.org.uk/guidance/CG169 (accessed 9 November 2014).
5. Adding insult to the injury, a review of the care of the patients who died in hospital with a primary diagnosis of  acute kidney Injury (acute renal failure), a national confidential enquiry into patient outcome and death (NCEPOD) report, 2009. Available at: http://www.ncepod.org.uk/2009aki.htm (accessed 10 November 2014).
6. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO Clinical Practice Guideline for Acute Kideny Injury. Kidney Int Suppl 2012;2:1–138.
7. NICE CG 50. Acutely ill patients in hospital: Recognition of and response to acute illness in adults in hospital. Available at: http://www.nice.org.uk/guidance/cg50 (accessed 5 November 2014).
8. Chan PS et al. Rapid response teams: A systematic review and meta-analysis. Archives of Internal Medicine 2010;170(1):18–26.
9. Winters BD et al. Rapid-Response Systems as a Patient Safety Strategy. A Systematic Review. Annals of Internal Medicine 2013;158(5 Part 2):417–25.
10. Forni LG et al. Identifying the Patient at Risk of Acute Kidney Injury: A Predictive Scoring System for the Development of Acute Kidney Injury in Acute Medical Patients. Nephron Clin Pract 2013;123(3-4):143–50.
11. Shamir MY et al. Urine flow is a novel hemodynamic monitoring tool for the detection of hypovolemia. Anesth Analg 2011;112(3):593–6.
12. Thomas M, Sitch A, Dowswell G. The initial development and assessment of an automatic alert warning of acute kidney injury. Nephrology Dialysis Transplantation 2011;26(7):2161–8.
13. NHS England patient safety alert: Algorithm for detecting Acute Kidney Injury(AKI) based on changes in serum creatinine with time. Available at: http://www.england.nhs.uk/ourwork/patientsafety/akiprogramme/aki-algorithm/ (accessed 8 November, 2014).
14. McCoy AB et al. A computerized provider order entry intervention for medication safety during acute kidney injury: a quality improvement report. Am J Kidney Dis 2010;56(5):832–41.
15. Rind DM et al. Effect of computer-based alerts on the treatment and outcomes of hospitalized patients. Arch Intern Med 1994;154(13):1511–7.
16. Colpaert K et al. Impact of real-time electronic alerting of acute kidney injury on therapeutic intervention and progression of RIFLE class. Crit Care Med 2012;40(4):1164–70.
17. Hooper MH et al. Randomized trial of automated, electronic monitoring to facilitate early detection of sepsis in the intensive care unit. Crit Care Med 2012;40(7):2096–101.
18. van der Sijs H et al. Overriding of drug safety alerts in computerized physician order entry. J Am Med Inform Assoc 2006;13(2):138–47.
19. Thomas ME et al. Earlier intervention for acute kidney injury: evaluation of an outreach service and a long-term follow-up. Nephrol Dial Transplant 2015;30(2):239–44.
20. Balasubramanian G et al. Early nephrologist involvement in hospital-acquired acute kidney injury: a pilot study. Am J Kidney Dis 2011;57(2):228–34.
21. Bagshaw SM, Zappitelli M, Chawla LS. Novel biomarkers of AKI: the challenges of progress ‘Amid the noise and the haste’. Nephrol Dial Transplant 2013;28(2):235–8.
22. Ronco C, McCullough PA, Chawla LS. Kidney attack versus heart attack: evolution of classification and diagnostic criteria. Lancet 2013;382(9896):939–40.