Medical Education Library
Designing a scoring system for clinical trials can be complicated. Sample sizes are dependent on the delta, or change, in efficacy or variation in response, and the design of the score will affect how easy it is to detect a statistically meaningful difference. These choices are a critical part of the design of small studies, particularly if obtaining enough statistical power can be challenging. Additionally, it is easier to detect change in more homogenous populations where we expect a more consistent response. Hidradenitis suppurativa is not a particularly homogenous disease, which furthers the risk of designing a trial that cannot detect important differences. The PGA often is required by the US Food and Drug Administration and has the major advantage that it is easy to understand, but the categories can sometimes be too broad to detect change easily, and more granular data can provide the basis for more in-depth analyses. An ideal outcome measure is a simplified scoring system that assesses disease severity and responsiveness to treatment while accurately serving as a surrogate for patient-reported outcomes, such as the dermatology life quality index, visual analog scale for HS skin pain, the work productivity and activity impairment questionnaire (specific health problem), or the patient global assessment. Validation processes for outcome measures, such as the one that HiSCR underwent, are essential to ensure that the proposed scoring system has clinical meaningfulness to both the physician and patient.
A 2016 Cochrane review of interventions for HS included 12 randomized controlled trials that employed a total of 30 different outcome measures instruments. Because use of multiple scoring systems makes it difficult to compare analyses of treatment, the authors concluded that there was a need for improved validation of HS outcome measures for future clinical trials.7 Schmitt et al8 recognized that atopic dermatitis also was in a similar predicament; they noted that more than 20 outcome measures were employed to assess disease severity in clinical trials. The authors called this situation “a significant threat to evidence-based health care” and outlined the Harmonizing Outcome Measures for Eczema (HOME) research initiative’s methodology for creation of core outcome sets for any dermatologic disease. Their consensus process involved first identifying what to measure, termed outcome domains, followed by developing how to measure these domains through outcome measures instruments, which would be assessed for validity, reliability, sensitivity to change, and feasibility.8
Using the framework set forth by the HOME initiative and data from the 2016 Cochrane review,7 a recent review of all outcome measures instruments currently employed in HS found that 90% (27/30) were not validated.9 Even those that were validated still could not be fully recommended by the authors. The authors identified 10 potential outcome domains for measurement, including quality of life, pain, lesion count, PGA, patient global self-assessment, recurrence rate, overall satisfaction with treatment, impairment of function, cosmesis, and duration of recovery. They recommended a further consensus process to better define these outcomes.9
Measuring all of these variables seems daunting, but as the speed of HS research rapidly progresses, we would greatly benefit from employing a standard validated scoring system that captures both disease severity and activity. Several groups are working to improve our current tools, but we will need to move quickly to a common approach so we can better compare treatment effects and build an evidence base for treatment decisions. For now, the HiSCR is the most validated clinical trials instrument, but it may not be ideal for the clinical setting. In our practice, we grade all patients each visit with Hurley staging, the validated HS-PGA scoring system to track improvement in inflammatory lesions, and a 10-point pain scale to monitor disease activity and severity. We have found these tools to be quick and effective for measuring treatment response and would recommend employment of these scoring systems as a standard measure in clinical practice until further consensus is reached.
