Comparison between two severity scoring scales commonly used in the evaluation of rotavirus gastroenteritis in children.

INTRODUCTION: Two severity scales, the Vesikari 20-point scale, and the Clark 24-point scale, were recently used to evaluate rotavirus vaccines. Each was used to evaluate a different vaccine. However, the two scales differ and have not been previously compared using the same patients.
OBJECTIVE: To compare the Vesikari 20-point and the Clark 24-point diarrhoea severity scales.
RESULTS: 987 gastroenteritis episodes in children < 3 years old were evaluated. Of these, 30.7% were rotavirus-positive. Of the 987 studied episodes, 755 (76%) were defined as severe by the Vesikari scale, vs. only 262 (27%) by the Clark scale (P < 0.001). All episodes defined as severe by the Clark's scale were also severe by the Vesikari scale. In contrast, only 262 (35%) of the Vesikari severe cases, were defined as severe by the Clark scale, while 475 (63%) and 18 (2%) were defined as moderate and mild, respectively by the Clark scale. By both methods, mean scores were slightly higher among Jewish than among Bedouin children, among females than among males, among hospitalised than among non-hospitalised children, and markedly higher in the 6-3 months old than in other age groups. The mean scores were also higher in the rotavirus-positive samples than in the rotavirus-negative samples by both scales. Transforming the Vesikari 2-category scale into a 3-category scale by further subdividing the severe category into two parts, improved the correlation between the two systems. However, still 155/344 (45%) found in the upper Vesikari severity score (> or = 16) were classified as only "moderate" by the Clark severity scale.
CONCLUSION: The two evaluation scales differ largely in their definition of severe cases. This may affect the ability to compare the effectiveness of various vaccines if studies do not use the same severity of scales.