When to Restart: RAAS Inhibitor Timing After an Episode of AKI
A single patient recovering from AKI, on chronic ACE-inhibitor therapy for proteinuric kidney disease that was held on admission. The disagreement is not whether to resume it eventually, but whether today's creatinine trend is confirmation enough to do it now.
C.B., a 68-year-old retired civil engineer, has spent his retirement doing exactly what he planned — long bicycle tours with his wife, the most recent through a stretch of rural roads that left him, five days ago, with a stomach bug he initially blamed on a gas-station sandwich. He has no diabetes and no cardiac disease; his only real chronic diagnosis is the proteinuric kidney disease itself, followed by the same nephrologist for nine years and, by his own account, the one part of his health he has actually taken seriously. Three days of significant diarrhea and poor oral intake later, he arrived at the hospital profoundly volume-depleted, and his admission labs showed his creatinine had risen from a stable outpatient baseline of 1.4 mg/dL, itself the product of longstanding proteinuric CKD attributed to prior hypertensive nephrosclerosis, to 2.6 mg/dL — a real acute-on-chronic injury, not simply his known disease finally being noticed. His lisinopril, which he has taken for eleven years specifically for its proteinuria-reducing effect, was held on admission, a standard and appropriate response to volume depletion superimposed on a RAAS-inhibited kidney.
Four days of resuscitation later, his creatinine has fallen to 1.6 mg/dL — close to, though not yet fully back at, his known baseline, and trending the right direction with each daily check. Nobody on rounds disputes that his proteinuric CKD still needs the drug long-term; the argument is entirely about which day. The observational literature the team will reach for — Brar's Alberta cohort of more than 46,000 post-AKI patients — found lower mortality among those who resumed a RAAS inhibitor within six months of discharge, but it also found those same patients hospitalized more often for renal causes, principally recurrent acute kidney injury and hyperkalemia. Both halves of that finding describe him, and they point opposite ways, because they are the same drug mechanism seen from two ends: dilating the efferent arteriole lowers the intraglomerular pressure that is slowly destroying his kidney, and is also precisely what makes a kidney four days into recovery vulnerable if that recovery is less complete than 1.6 suggests. The cohort cannot tell him which end he is on today, because it never compared people who waited.
Rounds, hospital day five, watching the trend line
I want to restart his lisinopril today, at half his prior dose, and retitrate as an outpatient. His creatinine has fallen for four consecutive days, he is clinically euvolemic, and every day his proteinuric kidney disease goes untreated is a day of unopposed intraglomerular pressure — the exact injury the drug exists to prevent. Observational data on RAASi discontinuation after AKI, including the Brar et al. cohort in JAMA Internal Medicine, found that patients who never resumed RAAS blockade after an AKI episode had meaningfully worse long-term outcomes, including higher mortality, than those who restarted. I will name the other half of that paper before you do: the same cohort found more hospitalization for renal causes among resumers, mostly recurrent AKI and hyperkalemia. I do not think that overturns the mortality signal, but it is real and it is his risk too.
I am not arguing against ever restarting it, and I take the discontinuation data seriously — and I appreciate you naming the renal-hospitalization arm rather than making me find it, because that arm is my whole argument. That cohort compared people who restarted to people who never did, at some point after discharge; it never compared either group to people who waited a week for a value fully at baseline. He is close to his baseline, not there yet, and restarting a drug that dilates his efferent arteriole while his kidney is still actively recovering risks re-triggering the same hemodynamic vulnerability that made this admission happen in the first place.
The observational data support resuming eventually, which we agree on; they do not specifically answer the question of whether today, at 1.6 rather than his known 1.4, is the right day, and I do not think it is fair to read that study as settling today's timing question.
That is a fair distinction, and I will concede the discontinuation literature does not directly address inpatient restart timing at a not-quite-baseline value — it is closer to the outpatient-follow-up question than to today's. Given that, I am willing to compromise on dose and setting rather than timing alone: start at a low, quarter-strength dose today with a recheck in forty-eight hours, rather than either his full prior dose today or waiting for a value that may take another week to fully normalize.
This splits the difference honestly rather than picking a side arbitrarily — it respects the discontinuation-risk argument by not waiting indefinitely, and the recovery-vulnerability argument by not restarting at full strength on a still-recovering kidney.
Agreed: lisinopril restarted today at a reduced dose, with creatinine and potassium rechecked in forty-eight hours before any further uptitration, rather than either an immediate full-dose restart or an open-ended hold.
The compromise was explicit and named as such by both physicians — not a default splitting of the difference, but a genuine response to the nephrologist's own concession that the discontinuation-outcome literature does not resolve today's specific timing question, only the broader eventually-restart-or-not question neither side actually disputed.