Incidence of the Bertillon and Gompertz effects on the outcome of clinical trials

TL;DR

This paper examines how demographic factors like marital status and age distribution, specifically the Bertillon and Gompertz effects, significantly influence clinical trial outcomes and proposes correction methods to improve reliability.

Contribution

It identifies the impact of demographic variables on trial results and offers practical correction strategies based on Bertillon and Gompertz laws.

Findings

Marital status can alter death rates by over 100%.

Age distribution in the elderly significantly affects death counts.

Corrections based on demographic data improve trial accuracy.

Abstract

The accounts of medical trials provide very detailed information about the patients' health conditions. In contrast, only minimal data are usually given about demographic factors. Yet, some of these factors can have a notable impact on the overall death rate, thereby changing the outcome and conclusions of the trial. This paper focuses on two of these variables. The first is marital status; this effect, which will be referred to as the Bertillon effect, may change death rates by over 100%. The second is the age of the oldest patients; because of the exponential nature of Gompertz's law, changes in the distribution of ages in the oldest age group can have dramatic consequences on the overall number of deaths. It will be seen that randomization alone can hardly take care of these problems. Appropriate remedies are easy to formulate however. First, the marital status of patients as well as the age distribution of those over 65 should be documented for both study groups. Then, thanks to these data and based on the Bertillon and Gompertz laws, it will become possible to perform appropriate corrections. Such corrections will notably improve the reliability and accuracy of the trial's conclusions, especially for trials which include a large proportion of elderly subjects.