2.4 Statistical analysis
Descriptive statistics such as mean and standard deviation (SD) and
range, were reported for continuous variables, and frequency (%) for
categorical variables. Two group comparisons were conducted using
Wilcoxon rank sum exact tests for continuous data and Fisher’s exact
tests for categorical data. A Firth’s logistic regression model was used
to determine biopsy variables associated with residual melanoma at the
time of surgical excision. We employed a backward selection approach to
create a parsimonious model and applied Firth’s logistic regression to
mitigate bias in estimates, given the limited sample size. Kaplan-Meier
progression-free survival curves and log-rank tests were used to
evaluate ulceration and presence of tumor-infiltrating lymphocyte (TIL)
status at biopsy. Multivariable Cox proportional hazards model in
combination with backward elimination method on preselected variables,
was used to determine variables associated with progression-free
survival.