Acute Kidney Injury Recognition
Definition and clinical scope
Acute kidney injury (AKI) describes an abrupt reduction in kidney function, recognised by a change in serum creatinine, urine output or both. It is a clinical syndrome with many causes, not a diagnosis of a single renal disease. The clinical task is to recognise the change early, evaluate its severity and identify reversible causes while managing complications. A patient can have substantial physiological deterioration before creatinine has increased appreciably. A normal creatinine result should therefore not override concern about urine output, perfusion, medications and clinical trajectory. Distinguish AKI from long-standing chronic kidney disease and remember that the conditions can coexist.
Underlying pathophysiology
Reduced renal perfusion, direct tubular or interstitial injury, glomerular inflammation and obstruction of urinary drainage are different mechanisms that can all lead to AKI. Hypovolaemia, haemorrhage, heart failure and systemic vasodilatation may reduce effective kidney perfusion. Sepsis may cause complex microcirculatory and inflammatory effects, sometimes without obvious systemic hypotension. Drug exposures can alter intrarenal haemodynamics or cause direct renal injury. Urinary obstruction may occur at the level of the ureter, bladder outlet or urethra. These mechanisms influence assessment and management but frequently overlap in a single patient.
Who requires assessment
Evaluate patients with acute illness and pre-existing CKD, sepsis, hypovolaemia, diabetes, liver disease, heart failure, recent exposure to potentially nephrotoxic drugs or iodine-based contrast, urinary symptoms or deteriorating early warning scores. Older adults and people reliant on others for fluid access may have increased vulnerability. After surgery, consider haemodynamic changes, bleeding, infection and obstruction. A normal serum creatinine on admission does not establish absence of evolving injury; compare with previous measurements and repeat according to clinical need. Urine output should be assessed in context rather than inferred from a single brief period without a recorded void.
Recognition and diagnostic criteria
NICE NG148 recommends recognising AKI using an increase in creatinine of at least 26 micromol/L within 48 hours, an increase of at least 50% from baseline within seven days, or urine output under 0.5 mL/kg/hour for more than six hours in adults. These thresholds are recognition criteria, not a substitute for judging clinical severity. Check units and the timing of samples. Calculate changes using the most defensible baseline available, while acknowledging uncertainty if pre-illness values are unknown. A patient may meet urine output criteria without yet meeting the creatinine criterion.
Clinical presentation
AKI is frequently asymptomatic early in its course and may be identified only through laboratory testing or urine output monitoring. Clinical features may instead reflect the precipitating illness: fever, diarrhoea, bleeding, shock, new heart failure, urinary retention or drug exposure. Features of complications can include fluid overload, pulmonary oedema, confusion, cardiac rhythm disturbance and symptoms of uraemia. Severe hyperkalaemia may be life-threatening even before kidney-related symptoms develop. Oliguria requires assessment, but urine output may be maintained in non-oliguric AKI.
Initial assessment
Assess airway, breathing, circulation, haemodynamic stability and need for urgent escalation. Review trends in blood pressure, pulse, temperature, oxygenation, weight and fluid balance. Identify fluid losses, recent medication changes, contrast exposure, infection, urinary symptoms and prior renal function. Examination should assess hydration and perfusion alongside possible fluid overload; neither a dry mouth nor oedema alone defines intravascular volume status. Examine for evidence of urinary retention or systemic disease. Consider bedside bladder assessment and renal ultrasound where obstruction is suspected or the cause remains unclear.
Investigations and monitoring
Repeat serum creatinine and electrolytes at a frequency proportionate to risk; daily measurement is common for hospital inpatients with AKI, while unstable patients may need more frequent testing. Assess potassium, bicarbonate and acid–base status where relevant. Record measured urine output accurately, including body weight where using weight-based criteria. Urine dipstick can help identify blood or protein and support consideration of intrinsic renal pathology. A medication review is essential. Imaging is guided by suspicion of obstruction and other findings, not automatically necessary in every case. Interpret any estimated GFR cautiously in a rapidly changing creatinine state.
Differential and contributing causes
Separate likely prerenal or perfusion-related factors, intrinsic renal disorders and postrenal obstruction as a practical initial framework, while recognising these are not mutually exclusive. Sepsis, acute cardiac dysfunction, bleeding and dehydration can coexist. Consider glomerulonephritis when renal dysfunction accompanies haematuria or proteinuria, particularly with systemic features. Consider rhabdomyolysis, drug toxicity and acute interstitial nephritis when the history fits. A clear precipitant does not eliminate the possibility of another simultaneous cause. A review of earlier test results and the patient’s longitudinal clinical history can materially change the differential.
Interpreting severity and trajectory
A single creatinine result is less informative than a sequence with collection times, baseline context and fluid balance. Creatinine can lag behind an acute change in filtration, and dilution during fluid administration can alter measured concentration. Reduced muscle mass may obscure the significance of seemingly modest elevations. Urine output trends provide complementary information but can be affected by diuretics and measurement limitations. Review whether the patient is improving, stable or deteriorating after correcting suspected causes. Reassess risks and potential complications each time the clinical state changes.
Management principles
Management follows identification and treatment of the underlying cause alongside prevention of further injury. Address haemodynamic compromise, infection or obstruction as appropriate to the patient and care setting. Review medication doses and medicines that may exacerbate kidney dysfunction, balancing the benefits and risks of withholding essential treatments. Avoid unnecessary nephrotoxic exposures and adjust renally eliminated therapies using current clinical judgement and local guidance. Fluid decisions should reflect measured status and response: excess fluid can be as harmful as inadequate perfusion in susceptible patients. Refer promptly for specialist input when intrinsic renal disease or severe complications are suspected.
Safety and escalation
Urgent escalation is warranted for serious electrolyte abnormalities, severe acidosis, pulmonary oedema, rapidly progressive renal deterioration, suspected obstruction requiring intervention or features suggesting severe intrinsic renal disease. The need for kidney replacement therapy is based on the overall clinical situation and refractory complications, rather than a creatinine number alone. Hyperkalaemia or respiratory compromise must be managed immediately through established emergency pathways. A clinical scoring tool or an isolated lab value cannot substitute for examination and timely senior review.
Worked clinical example
An adult weighing 70 kg has a previous creatinine of 80 micromol/L and a current result of 112 micromol/L collected 36 hours later. The rise is 32 micromol/L, meeting the short-interval creatinine criterion for AKI. If accurately measured urine output has also averaged 25 mL/hour for seven hours, that corresponds to about 0.36 mL/kg/hour and independently meets the adult urine-output criterion. This establishes recognition, not cause or treatment. The next assessment must address blood pressure, fluid status, infection, medication exposures, renal tract symptoms, potassium and other complications. All values in this case are illustrative.
Common errors and pitfalls
Common pitfalls include treating eGFR as reliable during a rapid creatinine change, using an incorrectly dated baseline, overlooking AKI in frail patients with low muscle mass and assuming adequate renal perfusion from a normal blood pressure alone. Another error is labelling all oliguria as dehydration and administering fluids without evaluating overload or obstruction. A urinary catheter may improve measurement accuracy in selected patients, but catheterisation itself has risks and should follow clinical indication. Do not let numerical stage classifications distract from hyperkalaemia, pulmonary oedema or an evolving systemic illness.
Aftercare and recovery
Kidney recovery may lag behind improvement in the precipitating illness. At discharge, communicate the episode, peak and most recent kidney function, outstanding investigations, medicine changes and follow-up plan. Review kidney function according to clinical stability, severity and residual impairment. Patients may need advice on hydration, temporary medicine adjustments during future illnesses and when to seek help, using locally approved patient information. Reassessment is particularly important after severe AKI, incomplete recovery or pre-existing CKD. A past AKI episode is relevant to future prescribing and risk assessment.
Relationship to other clinical tools
AKI recognition sits alongside rather than replacing assessment of fluid balance, organ dysfunction, sepsis and renal dosing. When comparing related measures, maintain the distinction between a dynamic change in kidney function and a steady-state estimate. A creatinine-based estimated GFR can help with chronic kidney assessment but is less reliable during acute change. Clinical deterioration scores identify broad physiological instability but do not diagnose AKI. These resources are complementary and should be interpreted alongside symptoms, serial findings and appropriate investigation.
Evidence and limitations
NICE NG148 provides a clear UK reference for detection, prevention, monitoring and referral. KDIGO consensus criteria form a widely used international framework. Clinical implementation requires attention to local pathways, measurement quality, populations covered by the guidance and updates to evidence. Numerical detection criteria identify possible injury but do not specify a unique cause; an algorithm may therefore support consistent recognition while still requiring qualified interpretation. All clinical examples should be checked against current local practice before the article is approved.
Baseline creatinine selection and uncertain baseline
Baseline selection should be explicit in an AKI record because the same current creatinine can produce different apparent changes depending on the reference. Prefer a recent stable pre-illness measurement where available, checking whether it truly represents usual kidney function. Admission creatinine may already be elevated after several days of symptoms. Older values may be misleading after changes in weight, muscle mass, medication, pregnancy or chronic disease. An estimated or assumed baseline introduces uncertainty and should be labelled as such rather than silently treated as observed. If a baseline is unavailable, consider current clinical findings, measured urine output, prior reports of CKD and serial laboratory values. Do not delay urgent treatment of hyperkalaemia, pulmonary oedema or sepsis while attempting to establish an ideal baseline. Document dates and units for each result, the suspected timing of the precipitating event and whether the pattern suggests acute, chronic or acute-on-chronic dysfunction. Misclassification can affect communication, risk stratification and subsequent follow-up. An apparently modest creatinine rise can be important for a small or frail person, and a large absolute value may reflect known advanced CKD without establishing a new acute event. Interpretation requires trajectory, not merely crossing an isolated threshold.
Urine output measurement quality
Urine output is both clinically useful and vulnerable to measurement error. For a patient who has not voided, ask whether urine is being produced but retained rather than assuming the kidneys have stopped producing it. Collection methods, incontinence, unmeasured toilet use, bag emptying times and errors recording body weight can all create misleading hourly rates. Review whether urine is being measured continuously, intermittently or estimated from incomplete nursing notes. When calculating a rate, use a defensible interval and recorded weight: total measured urine divided by weight in kilograms and the elapsed number of hours. Do not infer a six-hour criterion from a single low hour followed by adequate output without checking the actual sustained pattern. Diuretics may increase urine volume without resolving impaired filtration or renal injury. Conversely, the absence of oliguria does not exclude significant AKI. Record the context of fluid administration, haemodynamic changes and possible urinary obstruction. Catheterisation may be clinically useful in selected critically ill patients requiring accurate monitoring, but carries infection and trauma risks and is not appropriate merely to make a dataset complete. The clinical decision for measurement technique should follow the patient needs and local protocols.
Medication review and kidney function
A structured medication review should identify recently started, intensified or interacting therapies as well as long-standing medicines. Potential nephrotoxicity, effects on renal perfusion, volume depletion, potassium balance and accumulation due to reduced clearance are different concerns. Non-steroidal anti-inflammatory drugs can impair renal prostaglandin-mediated perfusion responses in susceptible circumstances. Some renin–angiotensin system medicines and diuretics may need reassessment during acute volume depletion, but stopping or restarting them should account for the original clinical indication and overall condition. Certain antimicrobials require therapeutic drug monitoring or renal dose adjustment. A drug-associated change in creatinine does not by itself prove structural kidney damage. Review non-prescription medicines, supplements and recent imaging contrast, and involve a pharmacist when dosing is complex. Do not replace a personalised clinical decision with an automatic blanket instruction to stop all long-term medication. The reason for each medication change, the monitoring plan and the criteria for resumption should be documented. In rapidly changing kidney function, a dose recommendation based solely on a steady-state eGFR can be misleading; consult current drug-specific guidance and local pharmacy support.
Volume assessment and fluid decisions
The relationship between circulating volume and total body water is complex. Patients with cardiac failure, cirrhosis or critical illness may have oedema yet reduced effective arterial filling, and aggressive fluid administration can worsen pulmonary congestion. Evaluate the history of intake and loss, perfusion findings, blood pressure trend, heart rate, lung examination, oedema, weight trajectory and response to initial treatment. Where appropriate, bedside ultrasound and dynamic measures may supplement examination, but each method has limitations. A urine output increase following fluid treatment can help assess response without proving that all renal injury is reversible. In septic shock and other emergencies follow the relevant resuscitation pathway, revisiting the balance of benefits and harms as the situation evolves. Document why fluid is given, what physiological endpoint is being assessed and when reassessment will occur. Avoid presuming oliguria automatically means hypovolaemia or administering repeated boluses in a patient with overt fluid overload. Clinical escalation should be prompt when perfusion remains poor, oxygenation worsens or renal function deteriorates despite initial measures. Fluid type, amount and rate require local guidance, comorbidity assessment and monitoring; no universal dose is appropriate for every person with AKI.
Obstruction and intrinsic renal disease
Obstruction may be suggested by urinary retention, anuria, hydronephrosis, pelvic malignancy, urinary tract instrumentation or characteristic symptoms, but the presentation can be subtle. Examine for bladder distension where appropriate and use imaging guided by NICE recommendations and the degree of suspicion. A normal early ultrasound does not settle every diagnostic uncertainty. Patients with suspected infected obstruction require urgent specialist management because delay can be dangerous. Intrinsic renal causes deserve consideration if urine dipstick reveals blood and protein in a compatible clinical picture, if systemic inflammatory or vasculitic features are present, or if kidney function continues to worsen without a satisfactory explanation. Urinalysis findings are clues rather than definitive diagnoses. Relevant investigation can include blood count, inflammatory markers, immunology and imaging based on the history and examination, usually in consultation with renal specialists. The presence of a plausible prerenal cause does not exclude concurrent drug toxicity or underlying glomerular disease. The record should communicate suspected mechanism, alternative explanations, investigations already done and triggers for revisiting the diagnosis. Escalate unexpectedly rapid deterioration or findings suggesting a treatable systemic renal disease.
Documenting a useful AKI handover
A safe handover goes beyond writing AKI in the problem list. State the observed baseline and current creatinine with dates, whether urine output criteria were met, the suspected cause and the level of certainty. Describe the clinical state: blood pressure trends, oxygenation, fluid balance, evidence of sepsis or obstruction and any important laboratory complications. Include actions already taken, response to those actions, remaining investigations and the next scheduled review. If potassium has been abnormal, communicate the result, treatment and timing of repeat testing. Record relevant medication changes and whether specialist advice has been obtained. The handover should specify warning signs that demand urgent reassessment, not imply that a stable single creatinine result means the problem is resolved. At discharge or transfer, note recovery status and arrangements for repeat kidney function testing, medication reconciliation and communication to community clinicians. Where relevant, explain the episode to the patient in language appropriate to their understanding and include advice from locally approved materials. A consistent handover reduces the risk that an apparently improving patient is discharged without an appropriate follow-up plan.
Questions for multidisciplinary review
Does the patient meet a documented creatinine or urine-output recognition criterion, and how reliable are the measurements? Is there a defensible baseline, or is its absence clearly acknowledged? What is the most plausible mechanism today, and which alternative causes would materially change management? Is the patient currently hypovolaemic, overloaded or physiologically unstable? Are potassium, acid–base status and respiratory findings safe? Have medicines and renal dosing been reviewed? Is urinary obstruction plausible, and is imaging or bladder assessment required? Are there features of intrinsic renal inflammation or systemic disease that warrant specialist input? Has the patient improved after appropriate intervention, or have serial data contradicted the original explanation? Are the next measurements, escalation criteria and follow-up responsibilities clear? These questions are prompts to improve clinical reasoning, not a substitute for an established AKI pathway or the judgement of the treating team.
Limitations of calculators and classifications
The purpose of an AKI recognition calculator is to standardise data interpretation, not to diagnose the mechanism of injury or select a treatment. It may correctly identify a threshold using dates, creatinine and weight-based urine output while missing important clinical features that are not part of the inputs. Its accuracy depends on the values entered, particularly baseline creatinine, time interval, weight and urine volumes. Where dates are wrong or observations are incomplete, a numerical output can create false certainty. Stage classification can aid communication and population-level measurement, but the urgency of care depends on complications, trajectory and context. No numeric category can safely rule out hyperkalaemia, pulmonary oedema, rapidly developing sepsis or urinary obstruction. Clinicians should retain the ability to override an apparently reassuring calculation when the clinical condition is concerning. These limitations should be displayed alongside calculator outputs and explained within training materials.
Scope of this educational article
This article is designed to help healthcare professionals recognise a pattern that warrants assessment and investigation. It does not provide patient-specific treatment orders, resuscitation prescriptions, specialist replacement-therapy criteria or a substitute for emergency clinical pathways. The examples are fictional and intentionally simplified to show how recording and temporal comparison affect recognition. The clinical significance of each result depends on the full observation set and the patient’s baseline characteristics. Readers should consult current UK and international guidance and check cited references for relevance to individual clinical decisions. If the local organisation uses a formal AKI alert or electronic reporting tool, its outputs should be reconciled with clinical assessment rather than treated as automatically decisive. Discussion of risk factors, tests and referral indications should be kept current as standards evolve.
Interpreting results across care settings
Recognition and response depend partly on the setting in which laboratory tests are obtained. In an emergency department a rapidly evolving physiological crisis may require action before a full renal history is available. On a hospital ward, trends in urine output, fluid balance and blood results may provide earlier warning of deterioration than a single review of the latest creatinine. In primary care, a newly abnormal result may require tracing earlier results, contacting the patient and assessing whether urgent same-day clinical review is needed. In an outpatient specialist service, an unexpected change may require clarification of recent illnesses, medicine changes and laboratory timing. These are different operational challenges, but the underlying questions remain the same: is there credible evidence of an acute change, is the patient clinically stable, and what could be reversible or immediately dangerous? Safe processes also depend on assigning responsibility for results that return after a patient has left the testing location. A clinical pathway should make clear who reviews repeat results, communicates concerning changes and arranges escalation. This is especially important when a result is generated by electronic rules but not yet assessed by a clinician. A technically correct alert can still be unsafe if no response process exists. Clinical teams should assess not only whether recognition criteria are met but also whether the patient’s immediate clinical needs and follow-up arrangements are properly addressed.