Background: Diagnosis of testicular torsion in children is challenging, as clinical presentation and findings may overlap with other diagnoses.
Objective: To define the clinical and ultrasound criteria that best predict testicular torsion.
Materials and methods: The records of children hospitalized for acute scrotum from 1997 to 2002 were reviewed. The clinical and ultrasound findings of children who had a final diagnosis of testicular torsion were compared with those of children who had other diagnoses (torsion of the testicular appendix, epididymitis, and epididymo-orchitis).
Results: Forty-one children had testicular torsion; 131 had other diagnoses. Stepwise regression analysis yielded three factors that were significantly associated with testicular torsion: duration of pain < or =6 h; absent or decreased cremasteric reflex; and diffuse testicular tenderness. When the children were scored by final diagnosis for the presence of these factors (0-3), none of the children with a score of 0 had testicular torsion, whereas 87% with a score of 3 did. The ultrasound finding of decreased or absent testicular flow had a sensitivity of 63% and a specificity of 99%. Eight of ten children with testicular torsion and normal or increased testicular flow had a coiled spermatic cord on ultrasound.
Conclusion: We suggest that all children with acute scrotal pain and a clinical score of 3 should undergo testicular exploration, and children with a lower probability of testicular torsion (score 1 or 2) should first undergo diagnostic ultrasound. Because the presence of testicular flow does not exclude torsion, the spermatic cord should be meticulously evaluated in all children with acute scrotum and normal or increased testicular blood flow.