Prevention of Recurrent Nephrolithiasis in Adults and Children: A Systematic Review

31-study review finds "low" strength-of-evidence for nearly every kidney stone prevention strategy, and zero data on surveillance imaging

Journal: Annals of Internal Medicine | Published: 2026-03-24 | Type: Systematic Review | PMID: 41871357 Authors: Asher GN et al. (RTI–University of North Carolina Evidence-based Practice Center; UNC School of Medicine, Chapel Hill) Funding/COI: Funded by the Patient-Centered Outcomes Research Institute and AHRQ (contract 75Q80120D00007/75Q80124F32010). COI not listed.

Summary

This RTI-UNC systematic review synthesized 31 studies on preventing recurrent kidney stones across diet, drugs, and imaging. The headline finding is unremarkable but clinically important: almost everything commonly recommended — more water, low-sodium/low-protein diets, thiazides, alkali therapy, allopurinol — has only low strength of evidence behind it. The review found no studies at all evaluating surveillance imaging strategies, a conspicuous gap given how routinely imaging is ordered. Evidence in children is nearly nonexistent; 28 of 31 studies enrolled adults only.

Claims

• Increased fluid intake, a diet high in calcium but low in protein and sodium, thiazides, alkali therapy, and allopurinol may reduce recurrence in adults with calcium oxalate or calcium phosphate stones — all rated low SOE
• No significant difference between selective (targeted by stone type/metabolic workup) and empirical pharmacotherapy — low SOE
• Acetohydroxamic acid (AHA) may reduce stone growth in infection-related stones (low SOE), but evidence on preventing recurrent stones was insufficient; AHA probably increases adverse events (moderate SOE — the highest confidence rating in the entire review)
• Lemon juice associated with minor adverse events; thiazides and allopurinol showed no increase in serious adverse events (low SOE)
• Zero studies evaluated imaging surveillance strategies
• 3 of 31 studies included children; evidence for pediatric populations is essentially absent

Study Quality

26 of 31 included studies were RCTs, which is the right design for prevention trials. Dual independent risk-of-bias assessment and strength-of-evidence grading were done, following standard systematic review methodology for an AHRQ-commissioned report. The PROSPERO registration (CRD42024617257) confirms prospective protocol registration, reducing post-hoc outcome switching risk. Data extraction used a one-reviewer/one-checker model rather than full dual extraction, which is a common but acknowledged shortcut that introduces modest risk of missed errors.

The fundamental problem isn't the review's methodology — it's the underlying evidence base. When 26 RCTs produce almost exclusively low-SOE conclusions, the trials themselves are the problem: small samples, short follow-up, heterogeneous populations, inconsistent outcome definitions. The reviewers correctly report what the evidence supports rather than overstating it.

Red Flags

• Low SOE across the board — "may reduce" is doing a lot of work in the conclusions; this is not a confident evidence base
• No imaging studies at all — a common clinical intervention with zero RCT or NRSI evidence reviewed
• Children nearly unrepresented — 28/31 adult-only studies; pediatric nephrolithiasis is rising in prevalence and clinicians have almost no trial data to work from
• Funding source not listed in the paper metadata; the abstract credits PCORI and AHRQ (government payers, not industry) — this is actually a strength, but the metadata gap is worth noting
• English-only and ≥30 participants/group exclusion criteria could miss relevant smaller trials, particularly in pediatric populations where large trials are rare
• COI disclosure absent from the data provided

Strengths

• Commissioned by AHRQ as part of the USPSTF evidence infrastructure — designed to be rigorous and policy-relevant
• Covers both adults and children in scope (even if pediatric data was sparse)
• Prospectively registered on PROSPERO
• Searched through December 2025 — unusually current for a systematic review
• Explicitly rates SOE rather than burying uncertainty in vague language
• The moderate-SOE finding on AHA adverse events is a clinically useful negative result

Verdict

This is a competent, properly registered systematic review from a credentialed evidence-based practice center — and its most useful contribution is the clarity of its uncertainty. The honest message: the standard advice for calcium stone prevention (drink more water, eat less salt and protein, take a thiazide) rests on low-quality evidence. That doesn't mean the advice is wrong, but it means the field is running on plausibility and clinical tradition more than solid trial data. The zero-study finding on imaging surveillance is the most actionable gap identified; any urologist ordering serial CT scans for recurrence monitoring is doing so without RCT support. For pediatric stone disease, the evidence vacuum is stark. This paper is worth reading not for reassurance, but as an honest accounting of how thin the evidentiary foundation actually is.