Mortality Outcomes in Dermatology: An Exploration of Core Outcome Sets and Cochrane Skin Systematic Reviews

Cochrane has been a trusted proponent of evidence-based medicine for over 20 years. Its dermatology-specific editorial team (Cochrane Skin Review Group) is the pre-eminent source of systematic reviews in dermatology [1]. Explicit standardized Cochrane review methods can minimize bias and maximize the reliability of reported outcomes, establishing benchmarks for decision-making. Mortality is one outcome where pronounced heterogeneity in reporting may affect its utility in clinical research. We therefore explored mortality outcome expression and execution in the Cochrane Skin portfolio and concurrently analyzed mortality in core outcome sets (outcomes that, at a minimum, should be measured in clinical trials) by searching dermatology studies registered in the COMET (Core Outcome Measures in Effectiveness Trials) database [2]. COMET contains text from core outcome sets publications, from which we extracted core outcomes and classified these according to the taxonomy developed by Dodd et al [3] for validated standardized annotation.

All Cochrane Skin Group reviews as of March 2021 were included and exported from the Cochrane Database of Systematic Reviews [1], allowing descriptive analysis and characterization of mortality reporting by category of mortality terminology (all-cause, cause-specific, infant/maternal, survival). All COMET database core outcome sets classified in the published “skin” research category as of August 23, 2021, were reviewed for reporting of mortality outcomes and categorized according to the mortality terminology previously described. Core outcomes specified in terms of “death” were included in the all-cause mortality category.

Of the 113 Cochrane Skin dermatology reviews, 13 reported mortalities as an outcome measure: 10 all-cause, 2 cause-specific, 5 survival, and 1 infant/maternal (Table 1).

Four reviews (4/13) reported more than one mortality outcome. More than one-third of the total reviews (5/13) were melanoma-related. Reviews of other dermatologic conditions reporting mortality included cutaneous squamous cell carcinoma (cSCC), nonmelanoma skin cancer, pemphigus vulgaris, pemphigus foliaceus, bullous pemphigoid, toxic epidermal necrolysis (TEN), necrotizing fasciitis, drug-induced skin rash, and topical steroids used during pregnancy. The time frame of mortality outcome reporting ranged widely, from 10 days to 10 years, but generally correlated appropriately with the condition (eg, 30 days for TEN capturing acute onset and progression vs 10-year survival for melanoma).

COMET database searches revealed 13 core outcome set studies of 13 skin conditions (Table 2); only 2 (15%) included mortality as a core outcome (survival for head and neck lymphatic malformations, death from cSCC).

Although limited in the number of studies appraised, our results illustrate substantial variability in the reporting and timing of mortality outcomes in Cochrane Skin reviews and COMET dermatology-related core outcome sets. Allowance of potentially unclear metrics (eg, “death”) and fluctuations in the time frame considered (especially within studies of a particular disease) may be detrimental to the downstream harmonization and generalizability of research findings. Guidelines to assist researchers during trial design and registration would encourage the selection of clear metrics and facilitate consistent outcome reporting at the later stages. Increased guidance and communication among stakeholders in this area, including further refinement of reporting guideline statements such as CONSORT (Consolidated Standards of Reporting Trials) [4] and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) [5], could promote much-needed standardization in mortality reporting, facilitating comparison across studies and helping decision makers effectively use dermatology research.