An application of the theory of probabilities to the study of a priori pathometry.—Part I

Ronald RossProceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character 1916 · DOI

This 1915 paper by Ronald Ross—the Nobel laureate who discovered that mosquitoes transmit malaria—is a foundational text in mathematical epidemiology. In it, Ross moves beyond simply observing disease statistics to building the first rigorous "a priori" mathematical models of how infections spread through a population.

TL;DR

Ronald Ross establishes the mathematical foundations of epidemiology by using differential equations to model disease dynamics. He demonstrates that the characteristic "bell-shaped" curve of an epidemic can emerge naturally from the exhaustion of susceptible individuals, challenging the contemporary belief that pathogens must "weaken" for an outbreak to end.

What Problem Does This Paper Address?

In the early 20th century, the study of epidemics was largely "a posteriori"—researchers looked at past data and tried to find curves that fit the numbers. However, this didn't explain why diseases behaved differently. Why do some diseases like tuberculosis remain at a steady "flat rate," while others like the plague flare up and vanish, or like measles, strike in periodic waves?

Ross argues that to understand these patterns, we must work from first principles (a priori). He tackles a specific debate of his time: whether epidemics end because the pathogen itself loses its "infectivity" over time, or because the population simply runs out of susceptible people to infect.

Methods

Ross introduces what he calls the "Theory of Happenings." He treats the transition from "unaffected" to "affected" as a mathematical event that can be modeled using a system of differential equations.

Key components of his model include:

  • Population Partitioning: He divides a population ($P$) into affected ($Z$) and non-affected ($A$) groups.
  • Variation Elements: He accounts for birth rates, death rates, and migration for both groups, noting that these rates often differ (e.g., the infected group has a higher mortality rate).
  • Independent vs. Dependent Happenings: This is the paper's most critical distinction.
  • Independent happenings (like accidents) occur at a constant rate regardless of how many people are already affected.
  • Dependent happenings (infectious diseases) occur at a rate proportional to the number of people already affected ($h = cZ/P$).
  • Reversion: He includes a "reversion-element" ($r$), representing recovery or the loss of immunity.

By setting up these equations, Ross can solve for the proportion of the population affected over time ($t$).

Datasets

This is a purely theoretical paper. Ross does not use a specific dataset for training or evaluation, though he references the work of Dr. William Farr (1866) and Dr. John Brownlee, who studied cattle plague and zymotic diseases. Ross positions his work as the mathematical "standard" against which future statistical examinations should be measured.

Key Insights and Ideas

  • The Exhaustion of Susceptibles: Ross’s most profound "aha moment" is proving that you don't need a biological change in the germ to explain the end of an epidemic. As the number of healthy people ($A$) decreases, the "happening-element" naturally hits a ceiling and then declines.
  • The S-Curve and the Bell Curve: He shows that while the total number of infected people follows an S-shaped (logistic) curve toward a limit, the rate of new cases follows a symmetrical bell-shaped curve. This matches the visual data of many real-world epidemics.
  • The Progeny Factor: Ross points out a crucial asymmetry: the children of infected parents are usually born healthy. This constant "injection" of susceptible individuals into the population is a key driver of long-term disease dynamics.
  • Generality of the Model: Ross suggests his "Theory of Happenings" isn't just for doctors. He argues it applies to any "happening" in a population, including the spread of political opinions, trade union membership, or even evolution.

Key Results

  • Mathematical Proof of Epidemic Symmetry: Ross derives equations showing that under constant conditions of infectivity and recovery, an epidemic will naturally rise and fall in a nearly symmetrical fashion.
  • The Steady State ($L$): He defines the limit ($L$) of an endemic disease—the point where new infections exactly balance recoveries and births.
  • Analytical Framework: He provides the explicit integrated solutions for the proportion of affected individuals ($x$) over time, allowing researchers to calculate the "tangential" (the rate of change) at any point in an outbreak.

Limitations and Future Work

Ross acknowledges several limitations in this "Part I" paper:

  • Constant Parameters: He assumes that birth, death, and infection rates remain constant for the sake of the initial proof, though he admits they likely fluctuate in reality.
  • Theoretical Scope: The paper is intentionally "academical." Ross notes that applying these equations to real-world epidemics requires much more work (which he planned for Part II).
  • Pathogen Evolution: While he proves that a "weakening germ" isn't necessary to explain an epidemic's end, he doesn't rule out that it could happen. He suggests his model should be used as a baseline to see if observed data deviates from it, which would then indicate that the pathogen's infectivity is indeed changing.

Summary updated

Abstract

Abstract Prefatory.—It is somewhat surprising that so little mathematical work should have been done on the subject of epidemics, and, indeed, on the distribution of diseases in general. Not only is the theme of immediate importance to humanity, but it is one which is fundamentally connected with numbers, while vast masses of statistics have long been awaiting proper examination. But, more than this, many and indeed the principal problems of epidemiology on which preventive measures largely depend, such as the rate of infection, the frequency of outbreaks, and the loss of immunity, can scarcely ever be resolved by any other methods than those of analysis. For example, infections diseases may perbaps be classified in three groups: (1) diseases such as leprosy, tuberculosis, and (?) cancer, which fluctuate comparatively little from month to month, though they may slowly increase or decrease in the course of years; (2) diseases such as measles, scarlatina, malaria, and dysentery, which, though constantly present in many countries, flare up in epidemics at frequent intervals; and (3) diseases such as plague or cholera, which disappear entirely after periods of acute epidemicity. To what are these differences due? Why, indeed, should epidemics occur at all, and why sbould not all infections diseases belong to the first group and always remain at an almost flat rate? Behind these phenomena there must be causes which are of profound importance to mankind and which probably can be ascertained only by those principles of careful computation which have yielded such brilliant results in astronomy, physics, and mechanics. Are the epidemics in the second class of diseases due (1) to a sudden and simultaneous increase of infectivity in the causative agents living in affected persons; or (2) to changes of environment which favour their dissemination from person to person; or (3) merely to the increase of suscep­tible material in a locality due to the gradual loss of acquired immunity in the population there; or to similar or other causes? And why should diseases of tbs third class disappear, as they undoubtedly do, and diseases of the first class remain so persistently?—all questions which immediately and obviously present themselves for examination.
BibTeX
@inproceedings{ross1916application,
  author = {Ronald Ross},
  title = {{An application of the theory of probabilities to the study of a priori pathometry.—Part I}},
  booktitle = {Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character},
  year = {1916},
  doi = {10.1098/rspa.1916.0007},
}
Figure