Acute Kidney Injury With a Rising Potassium

Learning objectives#

• Recognize acute kidney injury by change from baseline creatinine and urine output rather than by a single high value.
• Separate true hyperkalemia from a sampling artifact without delaying emergency care when the clinical risk is high.
• Organize acute hyperkalemia treatment into cardiac protection, temporary intracellular shift, potassium removal, cause control, and repeated monitoring.
• Identify medication, perfusion, obstruction, acid-base, and tissue-injury contributors to acute kidney injury and hyperkalemia.
• Explain when nephrology, critical care, and kidney replacement therapy must enter the plan.

Initial presentation#

A 68-year-old man is admitted after four days of vomiting and watery diarrhea. At the emergency assessment he was tired but alert, with dry mucous membranes, cool hands, and low blood pressure that improved after carefully reassessed intravenous fluid. He had no chest pressure, focal weakness, fever, bloody stool, or difficulty urinating. Stool frequency has now slowed.

His history includes hypertension, type 2 diabetes, heart failure with a previously preserved ejection fraction, and knee osteoarthritis. A laboratory record from six weeks earlier shows creatinine 88 micromol/L and potassium 4.5 mmol/L. His medicines include an angiotensin-converting enzyme inhibitor, a mineralocorticoid-receptor antagonist, a sodium-glucose cotransporter 2 inhibitor, and an anti-inflammatory medicine that he began taking several times daily when knee pain worsened. He continued all of them while he was unable to keep fluids down.

On the first inpatient morning, blood pressure is 102/64 mm Hg, pulse 96 per minute, respiratory rate 20 per minute, oxygen saturation 96 percent on room air, and temperature is normal. He is oriented and breathing comfortably. The jugular venous pressure appears low. Lungs are clear, heart rhythm is regular, the abdomen is soft, and there is no edema. A bladder scan after a small void shows little retained urine.

Creatinine has risen from 88 to 194 micromol/L. Bicarbonate is 18 mmol/L, and potassium is reported as 5.8 mmol/L with a laboratory comment that mild hemolysis may have affected the sample. During the previous eight hours, measured urine output has been approximately 0.3 mL/kg/hour. A trainee suggests waiting for the next morning's laboratory panel because the patient has no palpitations and the first sample may be unreliable.

Problem representation#

This is an acutely volume-depleted inpatient with a creatinine more than twice the recent baseline, sustained oliguria, mild metabolic acidosis, and a potassium result that is both plausibly real and potentially amplified by hemolysis. Renal perfusion has been impaired by gastrointestinal loss, while several medicines can reduce glomerular filtration, potassium excretion, or physiologic reserve during acute illness. There is no strong evidence of pulmonary congestion or bladder retention at this moment.

The immediate problem is not simply acute kidney injury. It is acute kidney injury with impaired potassium clearance and a result near a dangerous range that may be rising. Confirmation, ECG assessment, and monitoring must proceed in parallel. So must medication review, volume and obstruction assessment, and a plan for emergency treatment. The hemolysis comment changes how the result is verified; it does not justify walking away from it.

Prioritized differential#

1. Hemodynamic acute kidney injury with medication amplification#

Reasoning for: Several days of gastrointestinal loss, low intake, an initially low blood pressure, low jugular venous pressure, and oliguria support reduced kidney perfusion. The anti-inflammatory medicine can impair afferent arteriolar adaptation, while renin-angiotensin-aldosterone system medicines can reduce filtration reserve and potassium secretion during acute illness. The timing fits a multifactorial hemodynamic injury.

Reasoning against or still uncertain: Improvement in blood pressure does not prove that kidney perfusion has fully recovered. A rising creatinine may lag behind the injury, and prolonged hypoperfusion can progress to tubular injury. Heart failure makes both underfilling and excessive fluid possible, so volume assessment must be repeated rather than inferred from the admission label.

2. Acute tubular injury after sustained hypoperfusion#

Reasoning for: Persistent oliguria and continued creatinine rise despite initial circulatory improvement make structural tubular injury plausible. A bland or granular urine sediment, the time course, and response over the next several days can help, but no single urine index proves the diagnosis.

Reasoning against or still uncertain: The patient may still be under-resuscitated, and the current urine findings are not yet available. Hemodynamic and tubular mechanisms often overlap rather than occur as mutually exclusive categories.

3. Medication-associated and acidosis-associated hyperkalemia#

Reduced filtration and distal sodium delivery limit potassium excretion. The mineralocorticoid-receptor antagonist and angiotensin-converting enzyme inhibitor can further impair potassium elimination, while acid-base disturbance can contribute to redistribution. Continued potassium-containing fluids, supplements, salt substitutes, or enteral products would add to the load and must be checked.

4. Pseudohyperkalemia from collection or processing#

The hemolysis comment makes a falsely increased result possible. Difficult venipuncture, prolonged tourniquet time, and fist clenching can create misleading values. So can delayed processing, marked thrombocytosis, and marked leukocytosis. Yet this patient also has genuine physiologic reasons for hyperkalemia. A carefully collected repeat specimen and rapid point-of-care measurement are appropriate, but high-risk treatment should not be delayed if repeat results or ECG findings confirm danger.

5. Postrenal obstruction#

An older man with low urine output requires obstruction reasoning even without pain. The low post-void bladder volume argues against simple lower-tract retention, but it does not exclude upper-tract obstruction. Ultrasound becomes urgent if no cause is established, obstruction risk rises, or pyonephrosis is suspected. Routine imaging is less useful when the cause is clear and recovery is prompt.

6. Intrinsic renal disease, tissue breakdown, or another potassium source#

Glomerulonephritis, interstitial nephritis, and vascular disease can produce acute kidney injury or hyperkalemia. So can rhabdomyolysis, tumor lysis, and hemolysis inside the body. So can adrenal insufficiency, uncontrolled diabetes, and severe infection. Acute illness plus a sodium-glucose cotransporter 2 inhibitor also raises euglycemic ketoacidosis as an acid-base danger even when glucose is not high. Targeted tests must respond to these clues.

Focused history and examination#

Every item should answer whether the patient is unstable, why filtration fell, why potassium rose, or what treatment could harm him.

• Current instability: Ask about weakness, paralysis, palpitations, chest pressure, fainting, new breathlessness, and confusion. Repeat blood pressure, perfusion, respiratory effort, oxygenation, and mental status. Hyperkalemia may be asymptomatic before an arrhythmia.
• Fluid balance: Quantify vomiting, diarrhea, intake, urine, weight change, thirst, and fluid already given. Reassess jugular venous pressure, lungs, edema, capillary refill, weight, and response to small, clinically supervised fluid decisions. Heart failure narrows the margin for error.
• Urinary tract: Ask about hesitancy, weak stream, retention, stones, pelvic or prostate disease, prior obstruction, flank pain, and a solitary kidney. Confirm catheter patency if present rather than assuming the recorded urine output is biologic.
• Medicine reconciliation: Verify actual recent use of anti-inflammatory medicines, renin-angiotensin-aldosterone system drugs, diuretics, glucose-lowering drugs, trimethoprim-containing antibiotics, potassium supplements, salt substitutes, herbal products, and intravenous or enteral potassium. Record the last dose and indication.
• Alternative mechanisms: Ask about severe muscle pain, prolonged immobility, seizures, burns, transfusion, cancer therapy, infection, rash, joint inflammation, bloody urine, and recent contrast imaging.
• Baseline and goals: Retrieve prior creatinine, potassium, urine findings, cardiac function, kidney history, and the patient's preferences regarding intensive support or temporary dialysis if deterioration occurs.

The examination also looks for pulmonary edema, pericardial rub, and focal muscle tenderness. It looks for rash, embolic signs, bladder distention, and features of infection. None is sufficiently sensitive to replace laboratory trends, ECG, urine output, and bedside reassessment.

Diagnostic strategy#

Verify and risk-stratify potassium now#

The team obtains an urgent 12-lead ECG, places the patient on continuous cardiac monitoring because a rapid rise is plausible, and repeats potassium using a carefully collected sample. A point-of-care blood gas potassium can provide a rapid result while the central laboratory measurement is pending, but the team understands that some point-of-care systems cannot identify hemolysis.

The repeat strategy is not a delay tactic. If there were severe hyperkalemia, convincing ECG toxicity, or clinical deterioration, treatment would begin while confirmation proceeds. Conversely, treating a spurious result with insulin or other agents can cause dangerous hypokalemia or hypoglycemia. The clinical setting, rate of change, ECG, collection quality, and repeated measurement belong together.

Define acute kidney injury and its cause#

NICE and KDIGO-based criteria recognize acute kidney injury through a recent creatinine rise or reduced urine output. This patient's creatinine more than doubled from a documented baseline, and urine output has remained below 0.5 mL/kg/hour for more than six hours. Both signals matter, even though creatinine is not yet at a steady state.

The initial evaluation includes:

• repeat electrolytes, urea, creatinine, glucose, magnesium, calcium, and a blood gas for acid-base context, with anion gap and ketones when ketoacidosis is plausible;
• urinalysis for blood, protein, leukocytes, nitrites, and glucose, followed by microscopy when available;
• complete blood count and targeted infection studies if the clinical picture changes;
• creatine kinase if muscle injury becomes plausible;
• bladder assessment and urinary-tract ultrasound when the cause remains unclear or obstruction risk is meaningful;
• review of all inputs, outputs, fluids, nutrition, and medicines;
• frequent creatinine and urine-output monitoring, with frequency driven by acuity rather than a routine daily schedule alone.

Fractional excretion calculations are not treated as verdicts. Diuretics, chronic kidney disease, sepsis, mixed injury, and timing can undermine their interpretation. A large undirected autoimmune panel is also unnecessary without hematuria, proteinuria, systemic findings, or another signal of inflammatory kidney disease.

Progressive results and interpretation#

The repeat nonhemolyzed plasma potassium returns at 6.4 mmol/L. A near-simultaneous blood gas result is similar. Creatinine is now 239 micromol/L, bicarbonate is 16 mmol/L, and glucose is within the laboratory reference interval. The ECG shows new tall, narrow T waves compared with an old tracing, but the QRS complex is not widened. The patient still denies symptoms.

These findings establish true, rapidly worsening hyperkalemia with ECG effects in acute kidney injury. Lack of symptoms is not reassuring. The old tracing strengthens the interpretation, but waiting for a textbook sequence of ECG changes would be unsafe because ECG manifestations are variable and may not progress predictably.

Urinalysis shows trace protein without blood, leukocytes, or nitrites. Microscopy shows several granular casts but no cellular casts. Creatine kinase and ketones are not elevated, and the acid-base pattern does not support ketoacidosis. The bladder remains decompressed, and ultrasound shows no hydronephrosis. These results make glomerulonephritis, major muscle breakdown, ketoacidosis, and obstruction less likely. The best current explanation is severe hemodynamic injury with evolving tubular injury, amplified by acute-illness medicine effects.

Emergency treatment under the hospital's verified hyperkalemia protocol produces an early fall in potassium to 5.6 mmol/L. Four hours later, however, it is 6.1 mmol/L. Urine output remains very low, and the acid-base disturbance has worsened slightly. This is rebound after temporary redistribution in a patient who still cannot eliminate potassium. It is not a treatment failure that can be solved by repeating shifts forever without a removal strategy.

Management plan#

Protect, shift, remove, and reassess#

Because ECG changes are present, intravenous calcium is given through the hospital emergency protocol to reduce cardiac membrane instability. The team documents that calcium does not lower total-body potassium and confirms the follow-up ECG response. Product choice, route, and repeat use depend on resuscitation status and vascular access. They also depend on ECG evolution and the local protocol, so exact administration instructions are intentionally omitted.

Insulin with glucose is used to shift potassium into cells, with an inhaled beta-2 agonist as an adjunct when appropriate. Neither removes potassium from the body, and the beta-2 agonist is not relied on alone. Blood glucose is checked before treatment and repeatedly afterward because delayed hypoglycemia is a major treatment hazard, including in people without diabetes. Cardiac disease, tachyarrhythmia risk, and variable beta-agonist response are reviewed.

Potassium sources are stopped, and the team evaluates an appropriate removal method. Options may include a modern potassium binder, kaliuresis in a patient who is producing urine and can safely receive a diuretic, or kidney replacement therapy. Binders differ in onset, evidence, sodium burden, gastrointestinal constraints, and formulary status. Older resins with unpredictable onset and gastrointestinal harms are not assumed to be suitable emergency rescue.

Sodium bicarbonate is not a routine standalone treatment for acute hyperkalemia. Severe clinically relevant metabolic acidosis may alter the overall plan, but sodium and volume load, uncertain potassium effect, and the underlying cause must be considered. Loop diuretics are not used merely to treat acute kidney injury; they may have a role for fluid overload or potassium elimination only when urine production and volume status make that strategy plausible.

Treat the cause without creating a second injury#

The anti-inflammatory medicine, angiotensin-converting enzyme inhibitor, mineralocorticoid-receptor antagonist, and other medicines requiring acute kidney review are held or adjusted by the inpatient prescribers. This is not a permanent judgment about medicines that may provide long-term cardiovascular benefit. Each restart decision will require recovery of kidney function and potassium, a current indication, and a monitoring plan.

Fluid is given in small, reassessed steps because the patient appears depleted but has heart failure. A fixed large-volume order would be unsafe if pulmonary congestion develops. Daily weight, respiratory findings, and blood pressure guide the next decision. So do perfusion, urine output, and laboratory trajectory. Nutrition and intravenous fluids are reviewed for hidden potassium.

Escalate to definitive removal#

Nephrology and critical care are involved early, before the temporary measures wear off. Rebound hyperkalemia, persistent oliguria, worsening acidosis, and inability to eliminate potassium make urgent kidney replacement therapy a live possibility. The decision is based on the whole clinical picture and response to medical management, not on a single creatinine or potassium number.

In this case, the potassium rises again despite repeat temporizing care and remains accompanied by severe oliguria and worsening acidosis. After a goals-and-risk discussion, the patient receives temporary kidney replacement therapy. Potassium, acid-base status, volume, hemodynamics, and treatment complications are followed closely. Kidney function begins to recover over the next several days, and ongoing dialysis is not required, but that outcome was not knowable when the emergency decision was made.

Escalation, referral, and safety net#

Emergency escalation is immediate for a rising or severe potassium level, ECG changes, new bradycardia or tachyarrhythmia, widening QRS complex, syncope, muscle paralysis, hypotension, shock, pulmonary edema, severe acidosis, or cardiac arrest. The patient belongs in a monitored setting with staff able to deliver the local hyperkalemia and resuscitation pathways.

Nephrology or critical care discussion should not wait until the last available treatment has failed. Refractory or recurrent hyperkalemia, worsening acidosis, pulmonary edema, uremic complications, stage 3 acute kidney injury, unclear cause, transplant status, advanced chronic kidney disease, or inadequate response to initial care all strengthen the need for specialist input. Obstructed infected kidneys, bilateral upper-tract obstruction, or obstruction of a solitary kidney require urgent urologic action.

After stabilization, the safety net follows the patient across transfer and discharge. A written handoff names who will review potassium and creatinine, which medicines remain held, what conditions must be met before restart, and when laboratory testing will occur. Reduced urine, recurrent vomiting or diarrhea, weakness, palpitations, fainting, breathlessness, swelling, or confusion requires urgent reassessment rather than self-directed medicine changes.

Communication, shared decisions, and equity#

The clinician explains the sequence in plain language: "The kidney injury has reduced your ability to clear potassium. One treatment protects the heart, other treatments move potassium temporarily, and we still need a way to remove it and treat what caused the kidney injury. That is why the blood tests and heart monitoring continue even after the first number improves."

The medicine discussion avoids blame. Continuing usual medicines during vomiting is a common systems problem, especially when labels, portal messages, and clinician instructions do not explain what to do during acute illness. Before discharge, the team uses teach-back and provides a medication list that clearly distinguishes stopped, temporarily held, restarted, and unchanged medicines. Generic "sick day" instructions are not safe unless they name the medicine, the trigger, the person to contact, and the restart plan.

Language, vision, and reading level all affect safety. So do medication cost, transportation to laboratory testing, access to a scale or blood-pressure cuff, and ability to reach a clinician. Dietary counseling does not reduce the case to a list of forbidden foods. It considers cultural staples, nutrition risk after gastrointestinal illness, salt substitutes, food access, and whether dietary potassium was actually a major contributor.

If kidney replacement therapy is being considered, the team discusses expected benefit, uncertainty about renal recovery, vascular access, hemodynamic risks, and alternatives in a way that matches urgency and the patient's goals. Urgency can shorten a conversation, but it does not erase the patient's role.

Follow-up and contingencies#

During the acute phase, potassium is reassessed soon enough to capture the peak effect and later waning of temporary therapies. Glucose monitoring continues for the protocol-defined period after insulin because hypoglycemia can be delayed. ECG and vital-sign reassessment respond to severity and clinical change. Urine output is measured accurately, not estimated from memory.

Before discharge, the team documents the baseline creatinine, peak creatinine and potassium, and presumed causes. It documents the dialysis course if any, discharge kidney function, current volume status, and unresolved tests. Follow-up frequency is based on recovery and stability. It includes creatinine, potassium, and acid-base status. It includes blood pressure, volume findings, urine abnormalities when relevant, and medicine reconciliation.

Four contingency branches remain explicit:

• Kidney function and potassium normalize: Review whether and how long-term cardiovascular medicines should be restarted, one decision at a time, with laboratory monitoring.
• Creatinine improves but does not return to baseline: Reassess chronic kidney disease risk, albuminuria, blood pressure, and nephrology follow-up.
• Potassium rises after discharge: Confirm the result, review medicines and intake, assess illness and urine output, and escalate according to severity rather than repeatedly advising dietary restriction.
• The original cause no longer fits: Reopen the differential for obstruction, inflammatory kidney disease, vascular disease, infection, or ongoing tissue injury.

An episode of acute kidney injury increases future kidney and cardiovascular risk even after apparent recovery. The discharge summary must reach primary care and relevant specialists, and the patient needs a practical plan. It should cover future vomiting, diarrhea, and infection, along with imaging and nonprescription pain treatment.

Reasoning traps and alternative pathways#

• Dismissing a hemolyzed result: A collection artifact is possible, but physiologic risk determines the urgency of confirmation and monitoring.
• Waiting for symptoms: Hyperkalemia can be dangerous before the patient feels palpitations or weakness.
• Using the ECG as a rule-out test: ECG toxicity raises urgency, but a normal tracing does not reliably exclude arrhythmia risk.
• Confusing redistribution with removal: Insulin and beta-2 agonists buy time; they do not solve ongoing potassium accumulation.
• Repeating temporary therapy without monitoring: Rebound potassium and treatment-associated hypoglycemia can appear after the first improvement.
• Treating creatinine instead of the patient: Volume depletion, heart failure, obstruction, infection, medicines, and goals determine the safe intervention.
• Giving fluid by reflex: Both persistent underfilling and fluid overload can worsen outcome. Each step needs bedside reassessment.
• Stopping beneficial medicines forever: Acute holds and long-term cardiovascular decisions are different questions.
• Dialyzing by one number: Refractory electrolyte, acid-base, volume, and uremic complications plus the full clinical context guide kidney replacement therapy.

An alternative pathway would emerge if the carefully collected repeat potassium were normal, kidney function stable, and the ECG unchanged; that would support pseudohyperkalemia and prevent harmful treatment. Hematuria and proteinuria with systemic findings would elevate glomerulonephritis. A distended bladder or hydronephrosis would move obstruction toward the center. Pulmonary edema after fluid would require a different balance of hemodynamic support, diuresis, and kidney replacement therapy.

Evidence limits and what could change#

Hyperkalemia thresholds, laboratory methods, ECG interpretation, binder availability, and emergency protocols vary across systems. The UK Kidney Association guideline offers a detailed hospital algorithm, while the KDIGO acute hyperkalemia report emphasizes the same physiologic separation between membrane stabilization, intracellular shift, and removal. The evidence behind several of these acute treatments is thinner than their widespread use might imply, and direct comparative outcome data remain limited.

The 2012 KDIGO acute kidney injury guideline remains the final KDIGO guideline while a 2026 update has been through public review. A public-review draft is not treated here as final guidance. NICE NG148 was updated in 2024 and provides current detection, cause-finding, referral, and kidney replacement principles. Local protocols, current product information, specialist review, and the patient's comorbidities still control actual care.

The patient in this analysis had gastrointestinal loss and medicine amplification, then recovered after temporary kidney support. Real patients may have chronic kidney disease, severe heart failure, or poisoning. They may have tumor lysis, rhabdomyolysis, or adrenal disease. They may have transplant status, pregnancy, or goals that produce a different sequence. The uncertainty about cause and recovery must remain visible while emergency threats are treated.

Key points#

• Acute kidney injury is recognized by change from baseline creatinine and urine output, then explained through perfusion, intrinsic, obstructive, and medication reasoning.
• A questionable potassium result should be repeated correctly, but high clinical risk requires ECG assessment and monitoring during confirmation.
• Cardiac protection, temporary potassium shift, potassium removal, cause control, and repeated potassium and glucose checks are distinct parts of one plan.
• An early potassium fall does not guarantee safety because temporary therapies can wear off while kidney clearance remains poor.
• Refractory hyperkalemia, acidosis, fluid overload, uremic complications, or clinical deterioration require early nephrology and critical-care coordination, with kidney replacement therapy based on the whole patient.