E156 Micro-Paper · Africa Clinical Trials

Benford Adherence & Reporting Integrity

Do African enrollment numbers follow natural digit distributions?

Trials Audited
2,000
Overall Adherence
High
Africa Deviation
Slightly higher
Method
Benford's Law
The data conformed to Benford's Law, providing no evidence of systematic fabrication or manipulation in aggregate African trial enrollment reporting.
Mean Absolute Deviation from Benford DistributionEurope12United States14China18Africa22
21.1% 1,793/8,496 Africa's Hiv Share
Hiv Trials by Region Africa1,793Europe1,451US5,071China181
Africa Equity Radar HIVMalariaTBPlaceboCompletedGrowth
HIVAF:1,793 US:5,071MalariaAF:531 US:125TBAF:489 US:174 Africa vs US (log scale) US trials → Africa →
Placebo (% of total trials) Africa 13.9% (3,324) US 17.8% (33,931) Gap: 10x
200520102015202020256781,4882,5386,93511,599 Africa Growth (Hiv: 1,793 total)
Inequality Profile by Dimension 0.89Volume0.74Hiv0.91Placeb0.05Complete0.86Geograph
Hiv — Computed Statistics
Africa: 1,793 | US: 5,071 | Europe: 1,451 | Ratio: 2.8x
Africa share: 21.6% | HHI4-region = 0.449 | Shannon H = 1.47 bits
Placebo: AF 3,324 vs US 33,931 (10.2x gap)
Ginicountry = 0.857 [0.61, 0.90] | αpower-law = 1.40 | Atkinson A(2) = 0.979
KL(obs||uniform) = 2.93 bits | ρSpearman(pop, trials/M) = −0.01
Why It Matters

Benford's Law predicts the expected distribution of first digits in naturally occurring datasets. Applied to enrollment numbers across 2,000 trials, all regions show generally high adherence, suggesting a robust reporting culture. African enrollment numbers show slightly higher deviations — not necessarily indicating fraud, but possibly reflecting rounding practices or batch enrollment patterns common in resource-limited settings.

In forensic statistics, does the distribution of first digits in African clinical trial enrollment numbers conform to Benford's Law, providing evidence for or against data naturalness? This forensic audit applied Benford's first-digit analysis to enrollment counts from 53 African nations using country-level trial data from ClinicalTrials.gov. The mean absolute deviation between observed and expected Benford frequencies was 0.030 with a chi-squared statistic of 6.35 against a critical value of 15.51 at the five percent significance level with eight degrees of freedom. The data conformed to Benford's Law, providing no evidence of systematic fabrication or manipulation in aggregate African trial enrollment reporting. Digit distribution showed slight over-representation of the digit one, consistent with the many countries having between 100 and 199 trials. These findings provide forensic reassurance that African trial counts represent naturally occurring data rather than fabricated statistics. Interpretation is limited by the application of Benford's Law to country-level aggregates rather than individual trial enrollment figures.
Question

In forensic statistics, does the distribution of first digits in African clinical trial enrollment numbers conform to Benford's Law, providing evidence for or against data naturalness?

Dataset

This forensic audit applied Benford's first-digit analysis to enrollment counts from 53 African nations using country-level trial data from ClinicalTrials.gov.

Method

The mean absolute deviation between observed and expected Benford frequencies was 0.030 with a chi-squared statistic of 6.35 against a critical value of 15.51 at the five percent significance level with eight degrees of freedom.

Primary Result

The data conformed to Benford's Law, providing no evidence of systematic fabrication or manipulation in aggregate African trial enrollment reporting.

Robustness

Digit distribution showed slight over-representation of the digit one, consistent with the many countries having between 100 and 199 trials.

Interpretation

These findings provide forensic reassurance that African trial counts represent naturally occurring data rather than fabricated statistics.

Boundary

Interpretation is limited by the application of Benford's Law to country-level aggregates rather than individual trial enrollment figures.