Microscopic Colitis: A Taper, a Culprit Drug, and Which One to Address First
A single patient with new-onset microscopic colitis on a long-term NSAID, testing whether stopping the likely culprit drug or starting budesonide should come first when both are genuinely justified.
Ingrid W., a 67-year-old woman and retired postal clerk, has had chronic, watery, non-bloody diarrhea for the past three months, up to eight episodes a day, enough that she's stopped attending her Tuesday morning water aerobics class and started mapping every errand around bathroom access first. A colonoscopy with biopsies, obtained specifically because her diarrhea had a grossly normal-appearing mucosa despite the severity of her symptoms, came back showing lymphocytic microscopic colitis. She's been on naproxen twice daily for knee osteoarthritis for the past four years, a regimen that's kept her mobile enough to still walk her neighbor's dog most mornings, and has no other new medications, no recent infections, and no family history of inflammatory bowel disease.
Microscopic colitis has a real, well-documented association with several drug classes, and NSAIDs are among the most consistently implicated — a signal reproduced across case-control studies and reflected in the AGA's own guideline on microscopic colitis (Nguyen et al., 2016), which names drug discontinuation as a reasonable first step. Budesonide, meanwhile, has the strongest trial evidence of any therapy for inducing remission here: Miehlke and colleagues randomized patients with lymphocytic colitis specifically — Ingrid's histologic subtype, not the collagenous variant most of the older literature studied — and found budesonide markedly superior to placebo within six weeks. Both facts are true at once, and they don't automatically tell the group which move comes first: stopping a genuinely likely culprit drug before adding a new one, or starting the therapy with the strongest direct evidence right away given how much her daily life has already narrowed around her symptoms. What the decision actually hinges on isn't a disagreement about either fact, but a detail neither fact alone settles — whether her knee osteoarthritis has anywhere else to go if the naproxen stops. Ingrid herself raised the sequencing question unprompted, before the group could: her knee pain is real and has genuinely limited her mobility in the past, and she wants to understand honestly whether stopping her arthritis medication is being proposed because it's clearly the right move, or because it's the more convenient one for a diagnosis nobody had heard of until this week.
Clinic, new diagnosis, choosing what to change first
I'd stop the naproxen first and observe for a few weeks before starting budesonide. NSAIDs are one of the most consistently implicated drug classes in microscopic colitis, and the AGA guideline names stopping a suspected culprit drug as a reasonable first step rather than an afterthought. If discontinuing it alone resolves her symptoms, we'll have identified and removed the actual cause rather than starting a corticosteroid she may not have needed.
I'd start budesonide now, alongside stopping the naproxen, rather than sequentially. She's had eight episodes a day for three months and has already stopped an exercise routine she genuinely valued. Budesonide has the strongest trial evidence of any therapy for inducing remission in microscopic colitis — Miehlke's randomized data in lymphocytic colitis specifically, her own subtype, showed response within six weeks — waiting to see whether drug discontinuation alone works, if it doesn't, means real additional weeks of suffering that budesonide could have shortened.
I think the real variable that decides this isn't which piece of evidence is stronger — it's whether stopping naproxen is actually a clean, cost-free move for her, which depends on whether her knee osteoarthritis has anywhere else to go. If it doesn't, and she has no adequate alternative for her joint pain, stopping the NSAID outright isn't really a neutral diagnostic step, it's trading one quality-of-life problem for another while we wait and see. I'd want that conversation had directly with her before deciding, rather than assuming discontinuation is free.
Agreed, once Ingrid confirmed scheduled acetaminophen adequately controlled her knee pain in the past during a prior NSAID interruption: naproxen discontinued and budesonide 9mg daily started the same day rather than sequentially, with an 8-week reassessment for taper planning and explicit instructions that any recurrence after the eventual taper should prompt reconsideration of whether naproxen discontinuation alone would now be sufficient, since that specific question was never isolated by starting both changes together.