A Short NSAID Course Before a Scheduled Knee Replacement
A woman with mild-moderate CKD wants real relief from an osteoarthritis flare before surgery three months out. The instinct to withhold any NSAID from any CKD patient runs into the actual variable that changes her risk — and it isn't her CKD stage alone.
A knee replacement three months out is the fixed point everything else in Patricia N.'s case gets measured against. She is a 62-year-old woman who has kept a half-acre vegetable garden productive every summer since she retired from bookkeeping six years ago, work she's had to scale back sharply this season as her right knee osteoarthritis has worsened. In the meantime she has tried acetaminophen with only partial relief and a topical NSAID that helps somewhat but not enough to let her kneel at her raised beds the way the season requires. She is asking directly for a short course of an oral NSAID to get her through the growing season before surgery resolves the problem for good.
Patricia's CKD is stage 3a, eGFR 52, attributed to longstanding but well-controlled hypertension, and she takes no other medication that touches her kidneys — no diuretic, no RAAS inhibitor, nothing that would compound an NSAID's known effect of afferent arteriolar vasoconstriction and reduced prostaglandin-mediated renal blood flow. That combination matters more than her CKD stage does. The strongest evidence against NSAIDs in kidney disease is about acute injury rather than gradual decline, and it is concentrated in one specific combination: Lapi and colleagues, working through nearly half a million antihypertensive users in the UK Clinical Practice Research Datalink, found that adding an NSAID to a diuretic plus an ACE inhibitor or ARB — the "triple whammy" — raised the rate of acute kidney injury by 31%, rising to 82% in the first thirty days. The finding that actually describes Patricia is the one that got less attention: in that same analysis, NSAIDs added to either agent alone, as a double combination, carried no increased rate at all. She is not on a double combination. She is on neither leg of it, which places her outside even the arm of that study that came back clean.
One flare, three months, and a medication list that isn't the classic risk stack
I think a short, defined course is reasonable here. Her CKD is mild-moderate, her blood pressure and volume status are both fine, and she's asking for real relief to get through a defined window before surgery resolves the underlying problem. That's a legitimate clinical goal, and I don't think "she has CKD" alone should be a hard stop.
NSAIDs lower GFR through afferent arteriolar vasoconstriction regardless of what else someone is taking, and I'd add that Lapi's null result for dual therapy has been challenged: Dreischulte and colleagues, defining acute kidney injury by an actual creatinine rise rather than by hospital coding, found dual combinations carried much the same elevated risk as triple. Cohort data has also linked relatively short courses to measurable decline. She's already stage 3a — any further drop moves her closer to territory that starts to matter, and there are alternatives, even if they're less effective, that don't carry the same mechanism.
I'm not saying the risk is the same for everyone with CKD — I'm saying it's real even in someone whose disease looks mild, and "mild" isn't the same as "safe."
The specific interaction that drives most of the strongest evidence here — the "triple whammy" — is an ACE inhibitor or ARB, a diuretic, and an NSAID together, each removing a different renal compensatory mechanism at once. In Lapi's data that combination ran a rate ratio of 1.31, and 1.82 in the first month. Patricia isn't on either of the other two legs of that stack. I take Dreischulte's creatinine-based reanalysis seriously and it's why I won't call this risk-free — but it moves her from "outside the risk group entirely" to "at the low end of it," not into it. That's not a reason to treat NSAIDs as risk-free for her, but it does change what the same evidence base actually predicts. Shortest effective course, lowest effective dose, creatinine checked before starting and again at two weeks, explicit stop criteria if her numbers move — and this specific plan wouldn't be reasonable at all if she were on a RAASi or diuretic.
Agreed: a two-week naproxen course started at the lowest effective dose, creatinine checked before starting and repeated at two weeks, with an explicit plan to stop immediately if her renal function moves in the wrong direction.
Explicitly documented as case-specific, not a general policy: the same plan would not have been offered had Patricia been on a RAAS inhibitor or diuretic, a distinction the team agreed to state clearly in her chart so it isn't read later as a blanket "NSAIDs are fine in stage 3a CKD" precedent.