Item 36 — Demographic and clinical characteristics of cases in each partition and dataset

Specify the demographic and clinical characteristics of cases in each partition and dataset. Identify sources of potential bias that may originate from differences in demographic or clinical characteristics, such as sex distribution, underrepresented racial or ethnic groups, phenotypic variations, or differences in treatment.

Explanation and Elaboration: Key characteristics include age, sex, race/ethnicity, and disease presence. Depending on the clinical context, additional relevant characteristics may include known risk factors, comorbidities, disease severity, genetic markers, laboratory test values, prior treatments, and socioeconomic indicators. These characteristics should be summarized using appropriate descriptive statistics (e.g. mean with standard deviation or median with interquartile ranges) and presented separately for each dataset. Tabular presentation (typically in Table 1) is standard and should include stratification by dataset.

Examples

Table 1: Demographic Characteristics for Training, Validation, and Test Sets. [Table 1 and its caption (32), CC-BY 4.0]

Table 2: Patient and Imaging Characteristics of Skellytour Dataset, with Data for Total, Training, and Test Datasets. [Table 2 and its caption (14), CC BY 4.0]