Accepted answer
The operative variable is not episode count. It is whether your intake exceeds your losses, plus a short list of features that reclassify an episode regardless of how many there have been. One episode with the wrong features is a clinical problem; eight episodes over three weeks with maintained fluid intake and normal urine output is a tolerability problem. That is why the labels are vague: the count is genuinely not the discriminator.
The framework
Three questions, in order.
1. Can you keep fluid down? This is the whole question, and it is close to binary. If you can drink and retain a few hundred millilitres over an hour or two after an episode, you have a symptom. If every attempt at fluid comes back, you have a developing volume-depletion problem, and the clock on that is hours to a day, not weeks.
2. Is urine output maintained? The cheapest objective measurement available to you. Passing urine at roughly normal frequency, pale to mid-yellow, is reassuring. Not having passed urine in eight to twelve hours, or passing a very small volume of dark concentrated urine, is not. Note that this is more informative than thirst, because thirst signalling is itself altered on this class and several people report drinking less without noticing.
3. Does the episode have any of the reclassifying features? Listed below. These override the count entirely.
Features that make a single episode significant
- Abdominal pain that is severe, constant, or radiates to the back or right shoulder. Pancreatitis classically radiates through to the back; biliary pain to the right shoulder tip or between the shoulder blades. Both occur on this class. Pain out of proportion to the vomiting is the pattern to notice.
- Fever, rigors or jaundice. Suggests cholecystitis or cholangitis rather than a drug effect.
- Vomiting that begins abruptly after a period of stability at an unchanged dose. The drug did not change; something else did.
- Vomiting of blood, or material that looks like coffee grounds. Emergency, without qualification. Repeated forceful vomiting can also cause an oesophageal mucosal tear.
- Complete absence of bowel movements plus distension plus vomiting. That combination raises obstruction or ileus, both of which have been reported on this class, and it is managed very differently from nausea.
- Faecal odour to the vomitus, or vomiting of material eaten more than about eight hours previously. Retention or obstruction rather than emesis.
- Confusion, marked dizziness on standing, palpitations, or inability to stand. Circulatory consequences of volume depletion.
- Reduced urine output as above.
- Any of this in someone taking a diuretic, an ACE inhibitor, an angiotensin receptor blocker, an SGLT2 inhibitor, a sulfonylurea, insulin, lithium, metformin or an NSAID. Those drugs change the arithmetic substantially, which is the subject of the answer below.
What would a clinician look at
Beyond examination: postural blood pressure, capillary refill, mucous membranes, and a basic panel comprising urea and creatinine with electrolytes, sometimes with bicarbonate, chloride and glucose. Also lipase or amylase if pancreatitis is on the list, liver enzymes and bilirubin if biliary disease is, and imaging if either is genuinely suspected. Nothing you can do at home replaces that. What you can do at home is track the three questions above, which is what determines whether the panel is warranted.
Where your own pattern sits
Three episodes in ten days, all in the 24-48 hour post-dose window, all followed by improvement, with normal function in between, is the classic phase-locked pattern. That is the expected profile. The specific things I would watch, given that pattern:
- Whether the count is rising. Three in ten days is a symptom. Three per week rising to daily is a trajectory, and trajectories are more informative than snapshots.
- Whether it persists into the second half of the dosing week. Loss of the phase-locking is the single most useful sign that something has changed.
- Your total fluid intake, measured rather than estimated, for a week. Most people are surprised, and the direction of the surprise is always the same.
- Whether it survives the next four weeks at the current dose. If you are within a fortnight of a dose step it will very likely improve on its own; if you have been at this dose for two months it will not.
Worth naming the failure mode this framework exists to prevent. It is not people ignoring dramatic symptoms; it is people normalising a slow slide. Vomiting twice a week for two months, drinking a bit less each week because drinking makes it worse, feeling increasingly tired and attributing that to the deficit, and arriving at a genuine acute kidney injury without ever having had a day that felt like an emergency. Every step in that sequence feels tolerable. The endpoint is not. That is why the criterion should be an objective one you check on a schedule rather than a judgement you make when you feel bad.
edited 25 Dec 2024 by dmitri_savchuk — corrected a unit error in the worked example
Urine output over thirst is the correct emphasis. Thirst is unreliable on these drugs and almost nobody is told that. – e_dziedzic 2 months ago 8The slow-slide description is exactly what happened to me. Nine weeks, no single bad day, eGFR of 41 at the end of it. – nine_point_nine 15 days ago 7Vomiting food from the previous evening is a genuinely different sign and I had no idea it meant something separate. – claudia_ferrante 5 months ago add a comment