Cutpoints for mild, moderate, and severe pain on the VAS for children and adolescents - What can be learned from 10 million ANOVAs?

Cutpoints that classify pain-intensity into mild, moderate and severe levels are widely used in pain-research and clinical practice. At present there are no agreed-upon cutpoints for the visual analogue scale (VAS) in pediatric samples. We applied a method based on Serlin's procedure that was previously only used for the 0-10 NRS to empirically establish "optimal" cutpoints (OCs) for the VAS and used bootstrapping to estimate the variability of these thresholds. We analyzed data from the German Health Interview and Examination Survey for Children and Adolescents (KIGGS) study and defined OCs both for parental-ratings of their childrens' pain and adolescents' self-ratings of pain-intensity. Data from 2,276 children (3-10years; 54\% female) and 2,982 adolescents (11-17years; 61\% female) were analyzed. OCs were determined in a by-millimeter analysis that tested all possible 4851 OC-combinations, and a truncated analysis were OCs were spaced five mm apart resulting in 171 OC-combinations. The OC-method identified two different OCs for parental ratings and self-report - both in the by-millimeter and truncated analysis. Estimating the variability of the by-millimeter analysis we found that the specific OCs were only found in 11\% of the samples. The truncated analysis showed however that cutpoints of 35:60 are identified as optimal in both samples and are a viable alternative to separate cutpoints. We found a set of cutpoints that can be used both parental-ratings of their childrens' pain and self-reports for adolescents. Adopting these cutpoints greatly enhances the comparability of trials. We call for more systematic assessment of diagnostic procedures in pain-research.