Drug X Clinical Trial Analysis — Efficacy & Safety Dashboard
Dose-response and adverse event analysis on a simulated trial dataset
Data Analysis / Clinical Data
A statistical analysis of a simulated 1,200-patient clinical trial across three dosage groups, examining efficacy scores and adverse event rates, and presenting the dose-response and safety trade-off as a readable dashboard.
- Simulated dataset built for portfolio demonstration purposes.
Dashboard — Efficacy & Safety
Total Patients
1,200
Dosage Groups
3
Placebo / Low Dose / High Dose
Overall Adverse Event Rate
21.5%
Statistical Significance
p < 0.001
Efficacy and adverse events
Dose-response
Average efficacy score per dosage group (0–100)
Adverse event rate
% of patients with any adverse event
Adverse event severity breakdown
% of patients per severity, by dosage group
Placebo
Mild 6.6% · Moderate 2.2% · Severe 0.7% · No event 90.5%
Low Dose
Mild 12% · Moderate 6.9% · Severe 1% · No event 80.1%
High Dose
Mild 16.8% · Moderate 11.1% · Severe 7.3% · No event 64.8%
Key findings
- Efficacy increases significantly with dose (ANOVA, p < 0.001): Placebo 49.6 → Low Dose 62.0 → High Dose 72.5
- Adverse event rate also increases significantly with dose (Chi-square, p < 0.001): 9.5% → 19.9% → 35.2%
- Severe adverse events rise roughly 10x from Placebo (0.7%) to High Dose (7.3%)
- Interpretation: High Dose shows the strongest efficacy but the steepest safety trade-off — a narrow therapeutic window worth flagging for further clinical evaluation
Statistical tests: one-way ANOVA (efficacy across dosage groups), chi-square test of independence (adverse event rate across dosage groups). Data cleaned for duplicates, label inconsistencies, missing values, and invalid entries prior to analysis. Dataset is fully synthetic.
01 — The Problem
Trial data on its own does not answer the question a reviewer actually asks: does the benefit of a higher dose justify its safety cost? Efficacy and adverse events have to be read together, not in separate tables.
02 — Objective
Clean a simulated trial dataset, test whether efficacy and adverse event rates differ significantly across dosage groups, and present both together so the therapeutic window is visible at a glance.
03 — Approach — what I did
- Cleaned the dataset for duplicates, label inconsistencies, missing values and invalid entries.
- Calculated mean efficacy score per dosage group.
- Calculated the proportion of patients with any adverse event, and the severity breakdown, per dosage group.
- Tested efficacy differences across groups with a one-way ANOVA.
- Tested adverse event rate against dosage group with a chi-square test of independence.
- Built a dashboard view pairing the dose-response curve with the safety profile.
04 — Tools used
05 — Process
06 — Output
- Overview statistics for 1,200 patients across three dosage groups.
- Dose-response chart of mean efficacy score per group.
- Adverse event rate and severity breakdown charts per group.
- A plain-language findings panel with the therapeutic window interpretation.
07 — What I learned
- How to read efficacy and safety as a single trade-off rather than two separate results.
- How to choose the right test for the data type — ANOVA for continuous scores, chi-square for categorical event counts.
- How to present statistical output so a non-statistical reader can act on it.
08 — Limitations
- The dataset is fully synthetic and built for portfolio demonstration. No conclusion here applies to any real compound.
- The analysis covers group-level differences only — no covariates, subgroup analysis or time-to-event modelling.
09 — Skills demonstrated
10 — Project evidence
Evidence for this project is the dashboard reproduced in full on this page: overview statistics, charts, findings and the methodology note.