How High to Correct Anemia in a Patient With a Prior MI
A CKD patient with disabling anemia remembers feeling far better years ago at a hemoglobin target current guidance no longer allows. Her own history of a heart attack sits her inside the exact population where correcting that high produced real harm in trial data.
"I can't get to my own mailbox without stopping twice" is how Grace L. described her mornings at today's visit. She is a 66-year-old woman who spent most of her career as a court reporter, work she loved for the way it kept her mind moving fast even when her body sat still — a contrast to where she finds herself now. Her CKD, stage 5 but not yet requiring dialysis, has left her with a hemoglobin of 9.2, and she remembers a stretch eight years ago, before current anemia guidelines tightened, when a different physician kept her hemoglobin closer to 12 to 13 — a period she describes as the last time she felt genuinely like herself. She survived a myocardial infarction four years ago and has known coronary artery disease on a recent stress test, findings that predate today's visit and that the team already has in hand.
That cardiac history places her directly inside the population three major trials studied when they tested whether correcting anemia higher actually helps. Besarab and colleagues' Normal Hematocrit Trial enrolled dialysis patients specifically with congestive heart failure or ischemic heart disease, randomizing them to a normal versus a lower hematocrit target — and stopped early after the higher-target group showed more deaths and nonfatal MIs. CHOIR, in earlier-stage CKD, found the same direction: its higher-target arm had significantly more composite cardiovascular events than the lower-target arm. CREATE, run in the same era, is the more careful comparison — it did not show excess cardiovascular events, it simply failed to find any benefit from full correction, and more of its high-target patients ended up needing dialysis. Between them the two trials rule out the thing Grace is hoping for rather than establishing harm at every target. Grace's own remembered improvement at a higher number is real to her — but it comes from precisely the population, and precisely the target range, those trials found genuinely dangerous.
A hemoglobin target request built on a memory the trial data speaks to directly
Target 10 to 11, not 12 to 13. Besarab's trial didn't study a general dialysis population — it specifically enrolled patients with congestive heart failure or ischemic heart disease, exactly Grace's profile, and the higher-target arm was stopped early for more deaths and nonfatal myocardial infarctions. CHOIR found significantly more composite cardiovascular events in its high-target arm in earlier CKD; CREATE didn't show excess events so much as no benefit at all from chasing a normal hemoglobin, with more of its high-target patients reaching dialysis. This isn't a population-average risk being applied to her by extrapolation — it's the population the harm was actually demonstrated in.
I'm not disputing the trial data, and I share the concern given her MI history. But she is describing a real, specific, remembered functional difference — not a vague preference — and quality of life matters as its own outcome, not only as a proxy for cardiovascular risk. A rigid cap that ignores what she's telling us about her own experience risks treating a hard trial endpoint as the only thing that counts.
I take the population match seriously — I'm not arguing for 12 to 13, only that "conservative" shouldn't mean dismissing what she's reporting without first seeing whether we can get her closer to feeling better within the range the evidence actually supports.
There's an intervention here nobody's tried yet: her ferritin is 68 and saturation 18%, both below where iron optimization typically stops mattering. Correcting those first could raise her hemoglobin and symptoms meaningfully without pushing the ESA dose — or the target — anywhere near the range the trials found dangerous in her exact population. Start there, target 10 to 11 with ESA support, and treat any conversation about pushing higher as one to have only after we know what iron optimization alone actually bought her.
Agreed: IV iron started first, ESA initiated targeting hemoglobin 10-11, and Grace's own reported functional status tracked explicitly at each visit alongside the lab value rather than treating the number as the whole story.
Not agreed: what happens if she reaches 11 and still reports meaningful functional limitation — the cardiologist would revisit the target conversation directly with her at that point; the nephrologist would hold firm at the guideline range regardless. Both agreed to have that conversation only once iron optimization's own contribution is actually known, not before.