Table 1: Baseline Characteristics
Learn how to build and present the baseline characteristics table — the first table in nearly every clinical paper — correctly and completely.
You'll learn
How to build, format, and interpret baseline characteristics tables the way journals expect.
Use this when
You are writing the first table of any clinical or epidemiological paper.
What Table 1 Shows and Why It Matters
Table 1 describes who was in your study before any intervention or outcome measurement. Its purpose is to allow readers to assess (1) how representative the sample is of the target population, and (2) whether comparison groups are balanced at baseline.
💡 Table 1 is a description, not a test
Table 1 describes baseline characteristics — it does not test whether groups are different. Reporting p-values in Table 1 for randomized trials is considered inappropriate by most journals and statisticians, because randomization cannot produce a "significant" baseline imbalance by chance alone. Describe the groups; do not test them.
What to Report and How to Format It
| Variable type | Summary statistic | Format |
|---|---|---|
| Continuous, normal | Mean ± SD | 42.3 ± 8.1 years |
| Continuous, skewed | Median [IQR] | 6.2 [3.4–11.8] days |
| Binary / nominal | n (%) | 84 (38.2%) |
| Ordinal (small range) | Median [IQR] or n (%) | Grade 2 [1–3] |
- ●Test normality before choosing mean vs median — or use the Shapiro-Wilk test, or a histogram
- ●Always report the denominator: 84/220 (38.2%), not just 84 (38.2%)
- ●For grouped studies, present one column per group plus a total column
- ●List the most clinically important variables first: age, sex, primary diagnosis, key comorbidities
- ●Report missing data as a footnote or separate row
Standardised Mean Differences for Matching Studies
In propensity score matched or weighted studies, the standardised mean difference (SMD) replaces p-values as the balance measure. An SMD < 0.10 (or sometimes < 0.20) indicates adequate balance.
📖 SMD formula
SMD = (mean₁ − mean₂) / pooled SD. An SMD of 0 means perfect balance; > 0.10 suggests meaningful imbalance that may confound the analysis.
Building Table 1 in VibeResearch
The Table Builder automatically selects mean±SD for normally distributed continuous variables, median[IQR] for skewed variables, and n(%) for categorical variables. You can add a grouping variable to produce side-by-side columns.
- 1.Upload your dataset and go to Analyze → Table Builder
- 2.Select all baseline variables (age, sex, comorbidities, etc.)
- 3.Optionally add a group variable (e.g., treatment arm)
- 4.Download the formatted table or copy for Word
Practice with your own dataset
Generate Table 1 from your dataset.
Required variables
- • Any dataset — all variables appear as rows
- 1.Upload your dataset
- 2.Go to Analyze → Table Builder
- 3.Select all baseline variables you want to describe
- 4.Optionally specify a group variable
- 5.Review the output and download or copy for Word
Trusted sources behind this lesson
Further reading
Read next
STROBE: Reporting Observational Studies
Learn how to apply the STROBE checklist to ensure your cohort, case-control, or cross-sectional study is reported completely and transparently.