Background and purpose

Uncomplicated ureteral stone is a frequent diagnosis in the emergency department among adults with acute flank or abdominal pain, yet imaging—particularly computed tomography—carries cost, throughput implications, and ionizing radiation exposure. Clinicians often need a structured way to estimate how likely a ureteral stone is before or alongside imaging, and to identify presentations where alternative pathology remains plausible.

The STONE score is a clinical prediction rule derived and prospectively validated by Moore and colleagues and published in The BMJ in 2014. It converts five routinely available history and bedside findings into an integer score from 0 to 13. Higher scores correspond to a higher proportion of patients with ureteral stone on confirmatory imaging in the original cohorts, while very low scores identify groups in whom stone was uncommon and important alternative diagnoses were more relevant to consider.

The score is intended as a decision support tool: it does not replace clinical judgment, repeat examination when symptoms evolve, or institutional protocols for imaging and follow-up.

Derivation and validation at a glance

The rule was developed using multivariate modeling in a retrospective derivation sample of adults undergoing non-contrast CT for suspected uncomplicated kidney stone, then assessed in a separate prospective validation cohort. Investigators retained five predictors that carried the most predictive weight and could be operationalized as simple categorical fields suitable for busy emergency practice.

In both derivation and validation, patients were grouped into three bands on the total score—roughly aligned with low, moderate, and high estimated probability of ureteral stone. In the published work, the high-probability band was associated with ureteral stone in a large majority of patients, while the low-probability band had single-digit stone prevalence; the moderate band sat near even odds. These proportions are population-level estimates and will shift with local prevalence, case mix, and referral patterns.

Domains of the score

Each domain below is scored independently; points are summed for the total STONE score.

Sex

Ureteral stone is more common in male patients in many epidemiologic series; the original model assigns 2 points for male sex and 0 points for female sex. This does not imply that women cannot have stones—rather, the weight reflects the relative frequency of stone as the explaining diagnosis in the derivation population among those sent for CT.

Duration of pain

Symptom timing is scored in three tiers from onset (or the patient’s best estimate) to evaluation: pain present for less than six hours earns 3 points, six to twenty-four hours earns 1 point, and pain present for more than twenty-four hours earns 0 points. Shorter durations were more consistent with an acute obstructive ureteral calculus presentation in the derivation data.

Race category (as defined in the original study)

The published rule uses a binary race variable from the original cohort: non-Black versus Black, with 3 points assigned for non-Black and 0 points for Black. This variable reflected associations observed in the specific derivation sample and care setting; it is widely discussed in later literature as a limitation for generalizability, because baseline stone risk, healthcare access, imaging thresholds, and how race is recorded vary across systems. Teams applying the score should be transparent about this limitation when interpreting results for individual patients and when comparing performance across hospitals or countries.

Nausea and vomiting

Gastrointestinal symptoms are common with severe visceral pain from ureteral obstruction. The score stratifies: no nausea or vomiting is 0 points, nausea without vomiting is 1 point, and vomiting (whether or not nausea is also present) is 2 points. When vomiting is reported, the highest applicable category in this block should be used.

Microscopic hematuria

Hematuria is a classic—but not universal—feature of ureteral stone. The rule assigns 3 points when microscopic hematuria is considered present and 0 points when absent. In practice, sites should align this field with a defined standard: for example, positive blood on urine dipstick and/or red blood cells at or above a locally agreed threshold on microscopy. Dipstick specificity can be affected by myoglobinuria, recent instrumentation, infection, and other factors, so integration with the rest of the presentation matters.

Interpreting the total (0–13)

The total score places the encounter into one of three broad risk bands used in the primary publication:

• 0–5 points (low probability band): In the original cohorts, ureteral stone was found in roughly 8–9% of patients in this range. A low score therefore supports heightened attention to alternative causes of pain (including conditions that cannot be excluded without targeted testing) and may inform shared decisions about the urgency and modality of imaging, always conditional on vitals, comorbidity, pregnancy status, and evolving examination findings.
• 6–9 points (moderate probability band): Stone prevalence in published cohorts was near half of patients. This band is inherently the most “equivocal” clinically: neither stone nor non-stone can be inferred reliably from the score alone.
• 10–13 points (high probability band): Ureteral stone was present in roughly 88–90% of patients in the original reports, and acutely important alternative CT findings were uncommon in this stratum—though not impossible. High scores may support pathways that emphasize confirmation, symptom control, and disposition in line with local guidelines, while still maintaining safety netting for atypical features.

External validation studies have reported variable discrimination and calibration compared with the Yale-era cohorts, which is expected when disease prevalence, imaging thresholds, and patient demographics differ. The score is best viewed as one structured input among many—not as a standalone rule-in or rule-out test.

How to use this calculator on CalcMD

Select the option that best matches the patient’s presentation for each domain. The tool sums points automatically and displays the total, a component-wise breakdown, and the corresponding probability band text. Reset the form when moving to a new patient to avoid carryover.

Remember that the STONE score applies to the suspected uncomplicated ureteral stone context for which it was studied—typically adults in the emergency setting being evaluated for renal colic—not to every patient with abdominal pain, fever, sepsis, pregnancy, or known complex urologic history without adaptation.

Limitations and cautions

• Not a substitute for bedside assessment. Tachycardia, fever, pregnancy, immunosuppression, solitary kidney, intractable pain, or inability to tolerate oral intake may mandate imaging or admission regardless of score.
• Stone can exist at any score. Low scores reduce probability in population data but do not exclude ureteral calculus or other emergencies.
• Race as a predictor is sociotechnically and statistically contested; some centers have explored scores that omit or replace this term. Until local evidence supports an alternative, users of the original STONE score should document that the race item reflects the published model and may not transport equally.
• Hematuria absent does not rule out stone; similarly, dipstick-positive blood has other causes.
• Pediatric populations, anticoagulation, chronic kidney disease, and patients already known to have stones or stents may not match the derivation cohort.

Educational use only. Clinical decisions must remain the responsibility of the treating clinician and should follow current guidelines, institutional pathways, and patient preferences.