Severe acute kidney injury often leads clinicians to stop medications that affect kidney filtration, blood pressure, or fluid balance. That caution makes sense during active illness. However, a new study raises a different question for the recovery period: once the patient becomes stable, can waiting too long also carry risk?
A JAMA Network Open study published on August 27, 2026, examined adults with type 2 diabetes who survived acute kidney injury severe enough to require dialysis. All participants subsequently stopped dialysis. Starting an SGLT2 inhibitor within 30 days after discharge, rather than waiting until days 31–90, was associated with fewer serious kidney events.
The study does not establish a universal 30-day deadline. Instead, it suggests a more thoughtful principle: recent kidney injury should prompt careful reassessment, but it should not automatically produce months of therapeutic delay.
Why Recovery After Acute Kidney Injury Matters
Acute kidney injury, or AKI, represents a sudden decline in kidney function. It may arise from severe infection, dehydration, surgery, heart failure, low blood pressure, medication effects, or several interacting problems.
When AKI requires dialysis, the injury has reached a particularly severe level. Stopping dialysis marks important progress, but it does not mean that the kidneys have completely recovered. Patients may still face an increased risk of recurrent AKI, advanced chronic kidney disease, renewed dialysis, cardiovascular complications, and death.
The period after hospitalization therefore functions as a bridge between acute illness and long-term care. Clinicians must protect fragile recovery without allowing appropriate preventive treatment to disappear into a prolonged therapeutic vacuum.
What the New SGLT2 Study Found
Investigators used electronic health records to identify 4,406 adults with type 2 diabetes who had required dialysis for AKI and later discontinued it. None had received an SGLT2 inhibitor during the preceding year.
The researchers compared patients who started treatment within 30 days after discharge with those who started between days 31 and 90. After propensity matching, each group included 1,912 patients with broadly similar measured characteristics.
| Outcome | Started Within 30 Days | Started During Days 31–90 | Main Finding |
|---|---|---|---|
| Adverse kidney events | 3.9% | 6.2% | Lower with earlier treatment |
| Adjusted hazard ratio | — | — | 0.62 |
| Estimated number needed to treat | — | — | 43 |
| All-cause mortality | 7.3% | 6.9% | No significant difference |
| Major cardiovascular events | 9.7% | 9.5% | No significant difference |
| Recorded six-month safety outcomes | Similar | Similar | No clear safety signal |
The kidney outcome combined several serious developments, including dialysis restarting, progression to end-stage kidney disease, or a recorded estimated glomerular filtration rate below 15 mL/min/1.73 m².
The absolute difference was 2.3 percentage points. The adjusted hazard ratio of 0.62 suggests a 38% lower relative hazard of an adverse kidney event in the earlier-treatment group. The estimated number needed to treat was 43, meaning that earlier treatment would correspond to one fewer adverse kidney event for every 43 patients if the observed relationship proves causal.
Why Earlier Treatment Could Help
SGLT2 inhibitors began as glucose-lowering medications, but their benefits extend well beyond blood sugar. They can reduce pressure within the kidney’s filtering system, slow chronic kidney disease progression, and improve heart-failure outcomes in appropriate patients.
These effects make the post-AKI period clinically important. A patient recovering from severe kidney injury may also have diabetes, hypertension, protein in the urine, chronic kidney disease, or heart failure. Each condition influences the others. Delaying treatment for one problem may therefore weaken protection across the entire metabolic, cardiovascular, and kidney system.
However, timing matters. Starting too early during active instability may worsen dehydration or low blood pressure. Waiting automatically for several months may postpone meaningful kidney protection. A balanced approach focuses less on the calendar alone and more on whether the patient has reached genuine clinical stability.
What Stability Should Mean Before Starting
An SGLT2 inhibitor should not begin simply because 30 days have passed. The clinical picture should guide the decision. Before considering treatment, clinicians should confirm several points:
- Blood pressure and circulation have stabilized.
- The patient can eat and drink reliably.
- Vomiting, diarrhea, fever, or severe infection has resolved.
- Kidney function has stopped rapidly worsening.
- The patient no longer requires dialysis.
- Fluid status appears balanced rather than depleted or severely overloaded.
- The medication fits the patient’s kidney, diabetes, or heart-failure indications.
- The patient understands when illness, fasting, or surgery may require a temporary medication hold.
Active shock, ongoing volume depletion, evolving kidney failure, or substantial ketoacidosis risk should lead to delay. The same caution applies when a patient faces prolonged fasting, major surgery, or an acute illness that prevents normal food and fluid intake.
Different Patients May Need Different Timelines
The study supports earlier reassessment, but it does not replace individualized care. Several clinical variations show why the same hospitalization may lead to different decisions:
- Case 1 — Stable Kidney Recovery: A 62-year-old man with type 2 diabetes stops dialysis before leaving the hospital. His blood pressure remains stable, his appetite returns, and his kidney function reaches a steady level. Starting an SGLT2 inhibitor within the first month may offer meaningful kidney protection with close monitoring.
- Case 2 — Persistent Volume Depletion: A 70-year-old woman continues to experience diarrhea, poor intake, dizziness, and low blood pressure after discharge. Although she may benefit from an SGLT2 inhibitor later, her current condition favors hydration and stabilization before treatment.
- Case 3 — Kidney Disease With Heart Failure: A 66-year-old man has diabetes, chronic kidney disease, proteinuria, and heart failure. His clinician reviews his diuretic dose, blood pressure, weight, and kidney trend soon after discharge. Once these factors stabilize, earlier SGLT2 treatment may support both kidney and heart health.
- Case 4 — Evolving Kidney Failure: A 58-year-old woman leaves the hospital without dialysis, but her kidney function continues to decline. She develops swelling, nausea, and changing electrolyte levels. Her care team delays SGLT2 treatment while nephrology evaluates whether she may need dialysis again.
- Case 5 — Recovery Before Planned Surgery: A 64-year-old man stabilizes after AKI but expects major surgery within several weeks. Because surgery introduces fasting and metabolic stress, his physicians coordinate the start and temporary interruption of therapy rather than following a rigid timeline.
What the Study Cannot Prove
Investigators designed the analysis to resemble a clinical trial, and propensity matching balanced many measurable differences. However, this approach cannot control every factor.
Patients who started earlier may have recovered more smoothly, received better specialist follow-up, or appeared healthier in ways that electronic records could not capture. The 90-day landmark design also required participants to remain alive, off dialysis, without a transplant, and without recorded kidney function below a critical threshold. It therefore excluded patients who died or deteriorated during those first 90 days.
Both groups eventually received an SGLT2 inhibitor, so the study compared earlier treatment with later treatment rather than treatment with no treatment. It also evaluated new users, which means it does not directly answer when to restart an SGLT2 inhibitor that a patient had already taken before hospitalization.
Finally, similar recorded safety outcomes do not prove equal safety for every patient. A randomized prospective trial must confirm whether earlier initiation itself causes better kidney outcomes.
The Takeaway: Stabilize First, Then Reassess Promptly
This study meaningfully challenges the assumption that every patient should wait several months after severe AKI before starting an SGLT2 inhibitor. Once circulation, fluid balance, oral intake, and kidney function have stabilized, the recovery period may offer an opportunity for protection rather than a reason for indefinite hesitation.
The smartest strategy avoids both extremes. Clinicians should not start treatment during active physiological instability, but they should also avoid delaying it solely because AKI recently occurred. Thoughtful follow-up can identify the point when caution and prevention begin to support the same goal.
At our practice in Montgomery County, Maryland, and through virtual care, we evaluate kidney recovery within the broader context of diabetes, blood pressure, heart health, medications, and fluid balance. This individualized approach helps determine not only whether an SGLT2 inhibitor may help, but also when the patient’s recovery has become stable enough to use it safely and meaningfully.
References:
- Chou YT, Chen JY, Jiang ZH, et al. Early sodium-glucose cotransporter-2 inhibitor use after acute kidney injury in type 2 diabetes. JAMA Network Open. 2026;9(8):e2631231. doi:10.1001/jamanetworkopen.2026.31231.
About Dr. Gerti Tashko, MD
Dr. Gerti Tashko, MD, is an endocrinologist in Montgomery County, Maryland. He is uniquely quadruple board-certified in endocrinology, lipidology, hypertension, and obesity medicine. His practice delivers root-cause-focused metabolic and endocrine care, available virtually and in person. He uses advanced diagnostics, personalized nutrition, and preventive medicine to improve long-term health.
