CHICAGO – There is considerable variability and poor documentation of severity assessments used for inclusion criteria and baseline severity evaluations in randomized, controlled pediatric atopic dermatitis (AD) trials, results from a systematic review showed.
“It is important for clinicians and investigators to recognize that these differences may limit our ability to reproduce trials, interpret individual studies, and compare results between studies of similar target populations for severity,” lead study author Rishi Chopra, MS, said in an interview in advance of the World Congress of Pediatric Dermatology. “Moreover, this heterogeneity should be considered when retroactively pooling results for meta-analyses of pediatric atopic dermatitis randomized, controlled trials.”
Rishi Chopra
In an effort to evaluate the documentation and characterize the severity assessments used in inclusion criteria and baseline evaluations for randomized, controlled trials of pediatric AD internationally, the researchers performed a systematic review of relevant studies contained in the Cochrane Library, Embase, LILACS, GREAT, MEDLINE, and Scopus databases during 2007-2016. Inclusion criteria were RCT with a pharmacological intervention and any comparison with a control group, children, and males or females. In all, 89 studies met the inclusion/exclusion criteria. Most (70.8%) were studies of pediatric populations aged 0-17 years, and almost 17% were studies of infants aged 0-1 years. The most common target populations were mild-moderate AD (31.5%), moderate-severe AD (18.0%), or undefined (36.0%).
Mr. Chopra and his associates found that the most commonly used severity indices were Scoring AD (SCORAD) in 29.2%, Body Surface Area (BSA) in 16.9%, and global assessments in 13.4%, while the most common assessments of baseline severity were SCORAD in 43.8%, global assessments in 20.2%, Eczema Area and Severity Index in 17.9%, BSA in 14.6%, and visual itch in 13.5%. Only 85.4% of studies recorded the severity assessments used for recruiting the predefined target population and only 76.4% of studies documented baseline severity.
There was considerable heterogeneity across studies, as 16 unique assessments were used as inclusion criteria and 34 assessments were used to evaluate baseline severity. “In addition, even within an individual study, there was substantial discordance in their use as only 71.2% of studies used the same assessments for inclusion and documenting baseline disease severity,” Mr. Chopra said. “Altogether, this multidimensional lack of documentation and heterogeneity of inclusion criteria and baseline severity assessments limits our ability to assess whether the recruitment methods for patients were adequate and confirm whether the intended target population for severity was successfully enrolled.”
He acknowledged certain limitations of the study, including the fact that it may not be generalizable to nonpharmacological interventional trials or noninterventional studies. “In addition, we could only conduct an analysis for studies that provided adequate documentation of inclusion criteria and baseline severity,” Mr. Chopra said. “Thus, those studies that did not provide this information were left out. Nevertheless, across all studies, the uniformity and concordance between assessments likely are even more negatively impacted. It should also be noted that lack of documentation of assessments for inclusion criteria and baseline severity does not imply lack of their utilization. Finally, it is important to acknowledge that AD’s diverse phenotype and relapsing and remitting course may result in the unavoidable heterogeneity of severity assessment use. This may actually help to capture a broader range of disease and improve the external validity of results.”
He reported having no financial disclosures.