A Third of Overdose Hospitalizations Come With Kidney Injury. Two-Thirds of Those Are Severe.

Kidney failure is not what people picture when they picture an overdose. The image is respiratory — blue lips, a stopped chest, a naloxone kit. But a report the Centers for Disease Control and Prevention published Thursday in its weekly bulletin adds a different anatomy to the picture: among 1,297,847 hospitalizations involving a drug overdose between 2016 and 2025, 448,136 — 34.5 percent — carried a diagnosis of acute kidney injury.

That is roughly one in three. And in 273,093 of those cases, or 60.9 percent, the hospital course was severe by the study's own definition: acute dialysis, respiratory compromise, shock or hypotension, or death before discharge.

The authors, Sola Han, Fang Xu, Hui Xie, Gery Guy and Meda Pavkov, work across two CDC divisions — Diabetes Translation, and Overdose Prevention. That pairing hints at the argument. Kidney injury after overdose has been treated as a nephrology curiosity, a footnote to rhabdomyolysis. The report reframes it as an overdose-outcome measure, in the same family as a death certificate or an emergency department visit.

What the MMWR actually counted

The analysis drew on the Premier Healthcare Database, an all-payer hospital discharge system covering about 1,400 hospitals and health systems, roughly a quarter of American inpatient admissions. Overdose hospitalizations were flagged by ICD-10-CM codes for opioid, cocaine or psychostimulant poisoning, present on admission; follow-up encounters and sequelae were excluded. Acute kidney injury was flagged by its own code range, N17.0 through N17.9. Patients with end-stage kidney disease or chronic dialysis dependence were removed, which matters: otherwise the population would have been salted with people whose kidneys had already failed years earlier.

Severity was not a lab value. It was a bundle — dialysis, respiratory compromise, shock or hypotension, or in-hospital death — pulled from procedure and diagnosis codes.

Two structural limits belong up front, because they shape every number that follows. The unit of analysis is a hospitalization, not a person, so someone admitted three times contributes three times. And the database is a convenience sample of participating hospitals, weighted with projection factors to national estimates.

The age gradient runs backwards

The headline splits cleanly in two directions, and the split is the most interesting thing in the report.

Kidney injury itself rose with age, as it does in every other medical context: 28.2 percent of overdose hospitalizations among patients 24 and younger, 33.8 percent among those 25 to 44, 35.4 percent among 45 to 64, 37.4 percent among those 65 and older. All differences significant at p<0.001. Older bodies have less renal reserve, more heart failure, more diabetes, more prescriptions that stress the kidney.

Severity ran the other way. Among patients who had both an overdose and kidney injury, the severe-course share was 68.1 percent in the youngest group and 56.1 percent in the oldest. The 25-to-44 group sat at 62.5 percent, the 45-to-64 group at 61.0.

The authors offer a partial explanation rather than a confident one: drug mix differs by age, and in patients hospitalized with rhabdomyolysis — muscle breakdown flooding the blood with myoglobin, which the kidney cannot filter — illicit drug use has been associated with younger age, worse kidney injury and the need for dialysis. That is plausible and unproven here. The database carries no dose, no route, no adulterant list and no timing.

What the age inversion does establish is a burden argument. A 68-year-old who survives a severe overdose has perhaps fifteen years of kidney risk ahead. A 22-year-old has fifty. When kidney injury is a marker for later chronic kidney disease and failure — and it is, which is why the report treats it as both an acute complication and a forecast — the arithmetic of years at risk runs against the young.

The sex split is smaller and stranger

Men had more kidney injury: 38.1 percent against 29.4 percent for women. Women had slightly more severe courses when injury occurred: 61.9 against 60.4 percent.

A 1.5-point gap is small, and the report does not oversell it. The authors point to prior inpatient work reporting lower adjusted odds of kidney injury among women hospitalized for opioid overdose, and to a separate national cohort in which female sex and shock or hypotension were risk factors for severe opioid-related adverse events. They then write, plainly, that evidence of sex-specific patterns of hemodynamic instability in overdose-related kidney injury remains limited. It reads as a finding held at arm's length, which is the right posture for a 1.5-point difference in a claims-based dataset.

The practical instruction is nevertheless concrete: clinicians might watch female overdose patients more closely for severe complications once kidney injury appears, because the overall injury rate is lower and therefore less expected.

Why kidneys fail after an overdose

The mechanisms are well described and all of them are downstream of the drug event rather than mysterious. Rhabdomyolysis is the classic: prolonged immobility on a floor, seizures, hyperthermia, muscle breakdown, myoglobinuria, tubular injury. Ischemia is the second: hypotension and hypoxia from respiratory depression cut renal perfusion, and the kidney is an oxygen-hungry organ. Substance-specific toxicity is the third, varying by compound.

The report cites the oldest and starkest of the supporting studies, a 1988 New England Journal of Medicine series in which acute renal failure appeared in 33 percent of patients hospitalized for cocaine intoxication with rhabdomyolysis. Almost four decades later, the national burden of the same pathway has not been quantified in this way until now.

Kidney injury is also not cheap. The report cites an estimated $1,795 increase in cost per hospitalization when acute kidney injury is present, rising to $11,016 in additional cost when dialysis is required. Multiply the second figure by the severe-course count and the order of magnitude is a nine-figure annual line item that no one budgeted for.

The trend turned in 2022, and the report won't say why

Kidney injury as a share of overdose hospitalizations climbed from 2016 to 2022 at an annual percent change of 2.69 percent, then fell from 2022 to 2025 at −5.11 percent. Severe courses kept rising until 2023 (+1.82 percent) and declined after (−2.89 percent). The decline in severity was not statistically significant among patients younger than 45, or among women.

The turning points line up with the arc of the overdose epidemic — fentanyl saturation, then a plateau, then the first national decline in overdose deaths in years — but the report explicitly refuses the causal inference. These are adjusted percentages of hospitalizations, not counts of people, and the denominator moves when the numerator does. If naloxone saturation since 2019 means fewer overdoses become hospitalizations at all, the surviving hospitalizations are the worse ones, and the rate can rise while the population-level burden falls.

The severity measure has a definitional problem

That word — compromise — deserves scrutiny. Respiratory compromise is one of four criteria in the severe-course bundle, alongside dialysis, shock and death. In an opioid overdose, respiratory depression is not a complication of kidney injury. It is the overdose. A patient whose breathing slowed enough to require a monitored bed and who also happened to be coded with acute kidney injury will be counted as a severe course regardless of what their creatinine did.

So 60.9 percent should be read as "these hospitalizations contained at least one marker of physiologic instability," not "60.9 percent of these patients were in kidney failure." The dialysis component of the bundle is unambiguous. The rest is a compound indicator, and the report's own framing — "severe inpatient course" as a single construct — flattens that.

This does not undo the finding. It does mean the most quotable number in the report is the softest one.

The measurement floor is probably below the truth

Kidney injury here is a coded diagnosis, not a laboratory threshold. Creatinine has to be drawn, the result has to rise, a clinician has to document it, and a coder has to abstract it. Every one of those steps leaks. The authors name this themselves: diagnosis codes rather than lab criteria "might have resulted in an underestimate."

The comparison that makes the undercount legible is the base rate. Acute kidney injury occurs in roughly 20 percent of hospitalized adults generally. Thirty-four and a half percent in overdose hospitalizations is about 1.7 times that — and the true figure is probably higher, because a patient admitted for an overdose at 3 a.m. may never get a full renal workup.

What a treatment system is supposed to do with this

The report's recommendations are modest and mostly about continuity: recognize kidney injury early, link patients to kidney follow-up after discharge, and prevent the next overdose through better access to opioid addiction treatment and, for the stimulant-involved hospitalizations that make up a meaningful share of this cohort, stimulant treatment.

The follow-up piece is where the gap is widest. KDIGO, the international kidney guideline body, recommends assessing kidney function within three months of an acute injury. Nationally, follow-up among Medicare beneficiaries after acute kidney injury was 13.6 percent in 2023 — below even the Healthy People 2030 target of 16.4 percent.

Thirteen point six percent. That is the number to hold next to 34.5. The acute event is being counted; the chronic consequence is not being tracked. For a 24-year-old discharged after a fentanyl overdose with a creatinine that spiked and a follow-up appointment that never happened, the hospitalization is the beginning of a kidney disease story that will be told a decade later in a dialysis unit.

Whether the second read of this report is hopeful depends on which series you look at. Deaths are falling. Overdose hospitalizations with kidney injury have been falling since 2022. But severity among the youngest and among women has not measurably improved, and the handoff to kidney care barely exists.

The CDC's own summary puts the opportunity plainly: because the exposure is identifiable and potentially avoidable, some of this kidney injury might be preventable. That is a narrower claim than it sounds. It puts the intervention on the prevention side — naloxone, treatment access, not using alone — and leaves the nephrology side to catch what prevention misses. Right now it catches about one in seven.