Clinical Prediction Rules for Identifying Children With Testicular Torsion: A Multicenter Prospective Study.

OBJECTIVES

To validate clinical scores [Testicular Workup for Ischemia and Suspected Torsion (TWIST), testicular torsion (TT) score, Artificial Intelligence-based Score (AIS), Boettcher Alert Score (BALS)] when evaluating children under 18 with non-traumatic testicular pain in the emergency department. Our secondary objective was to create and compare a new TT score [Testicular Emergency Score for Torsion (TEST) score].

METHODS

This was a multicenter prospective study in 21 Spanish pediatric emergency departments between 2020 and 2022, including 903 children 3 months to 18 years old with non-traumatic unilateral testicular pain, of them 93 TT (10.3%). To create a new score, the sample was randomly divided into derivation and validation set.

RESULTS

The performance of the TWIST, TT score, AIS, and BALS was good, and the proportion of patients correctly classified as low risk was 37.9%, 52.7%, 30.3%, and 28%, respectively. The TEST score included the following predictors of TT identified by multivariable logistic regression analysis: age, duration of pain, nausea/vomiting, testicular volume increase, testicular elevation, induration, and absence of cremasteric reflex. TEST score had a higher area under the receiver operating curve (area under the curve) and correctly classified in the low-risk group of 63.6% of the patients.

CONCLUSIONS

Although TWIST, TT score, BALS, and AIS scores showed a good performance, the TEST score identifies a larger group of low-risk patients suitable for safe management without Doppler ultrasound.