The drug given to stop stones coming back did not beat placebo — at any of three doses

Thiazide diuretics have been the standard medical answer to recurrent calcium stones for decades. They lower urine calcium, urine calcium drives calcium stones, and the logic has been strong enough that guidelines recommended them on evidence nobody described as good. In 2023 the trial that had not been done was done.

What NOSTONE found#

Published in the New England Journal of Medicine, NOSTONE randomised 416 patients with recurrent calcium-containing kidney stones, double-blind, to hydrochlorothiazide at 12.5 mg, 25 mg or 50 mg once daily, or to placebo once daily. The primary end point was a composite of symptomatic or radiologic recurrence, radiologic recurrence being a new stone on imaging or the growth of one already visible at baseline. Median follow-up was 2.9 years.

The trial was designed around a dose-response question, and the dose-response was absent: P = 0.66 for the relation between dose and event. Read down the rate ratio column and the pattern is not a drug working weakly. Two of the three doses have point estimates above 1.

And the drug was not free. Hypokalaemia, gout, new-onset diabetes mellitus, skin allergy and a plasma creatinine exceeding 150 percent of baseline were each more common among patients taking hydrochlorothiazide than among those taking placebo. A preventive treatment that does not prevent still causes.

What the wider evidence says two years on#

A network meta-analysis published in World Journal of Urology in 2025 pooled nine randomised trials totalling 999 patients. Three regimens did separate from placebo on recurrence: chlorthalidone 50 mg daily with an odds ratio of 0.18 (95% CI 0.04 to 0.88), hydrochlorothiazide 50 mg daily at 0.52 (0.29 to 0.93), and trichlormethiazide 4 mg daily at 0.26 (0.10 to 0.68).

That reads like a rescue until the rest is included. No dose-dependent effect appeared when hydrochlorothiazide 50 mg was compared with 12.5 mg or 25 mg, or chlorthalidone 50 mg with 25 mg. Trichlormethiazide 4 mg produced more adverse effects than placebo with an odds ratio of 49.96 and a confidence interval running from 1.78 to 1,402.80 — an interval that width is a warning about the evidence, not a measurement of the harm. The authors' conclusion was that current evidence does not support routine prescribing of these drugs for stone prevention and that head-to-head trials are urgently needed.

The line of defence that is left#

If thiazides work by lowering urine calcium, then a trial that enrols recurrent stone formers regardless of their urine calcium is testing the drug on people it was never mechanistically expected to help. That argument has a 2026 measurement behind it.

Writing in the Journal of Urology, a group examined 634 patients who had a 24-hour urine collection for stone disease, were newly prescribed a thiazide, and completed a second collection between 30 and 180 days later. Higher baseline urine calcium was associated with greater absolute and percentage reductions in 24-hour urine calcium, and higher thiazide doses with greater reductions still.

What this means for a Korean stone patient#

The 2024 European Association of Urology guideline update did not abandon metabolic evaluation after NOSTONE; it made the case for it sharper. Every stone should be analysed, every patient should have baseline metabolic screening, and calcium-stone formers at high risk of relapse should have extensive screening with two 24-hour urine collections before targeted therapy. On the trial evidence, the 24-hour urine is now the more defensible half of that pairing and the tablet is the less.

How often the two are ordered together in Korean practice is not something any study cited here measures, and it is the gap worth naming rather than filling with an assumption. Korea has national claims data detailed enough to answer it — the same source that has already been used to count every lithotripsy and every scan — and no published analysis has asked how many Korean patients on preventive therapy ever had the urine test that is supposed to select them for it.

For the patient in front of a urologist, the practical reading is narrower than the headline. A thiazide is not a proven way to stop calcium stones returning. It remains a reasonable option in a patient with genuinely high urine calcium who has been counselled about hypokalaemia, gout and glucose, and it is a poor default for a recurrent stone former whose urine has never been measured.