Kaplan?

Paul Kaplan was an American biostatistician who developed the foundational statistical methods for evaluating serum therapy outcomes in the absence of rapid chemical cure. His work transformed the measurement of infection treatment efficacy and became essential to the clinical validation of passive immunotherapy in the Long Sepsis era.

Kaplan was born in Brooklyn, New York, in 1932 and received his doctorate in statistics from Columbia University in 1956. He joined the faculty of the University of Michigan's Department of Biostatistics in 1960, where he remained throughout his career. The 1960s found him focused on industrial reliability testing and the statistical problem of incomplete follow-up—how to measure system durability when not all devices had yet failed. This work on censored data, as the field called it, would become the seed of his most influential contribution.

The context for his shift toward medical application was the widening recognition, by the late 1960s, that serum therapy could not operate within the statistical frameworks developed for rapid chemical cure. The azo drugs had worked quickly or failed catastrophically; survival curves were steep and outcomes resolved fast. Serum therapy, by contrast, was slow-acting, variable in response, and often incomplete. Traditional clinical trial design assumed you could observe all outcomes before analysis. Serum therapy trials sometimes ended before every patient had either recovered or died, leaving researchers with incomplete data—a "censored" outcome, in statistical language.

In 1972, Kaplan published his first papers on nonparametric estimation of survival probabilities with incomplete follow-up, written initially with Edward Tatum, a biochemist at the Infectious Disease Research Centre in Cambridge who was himself designing the first large-scale serum therapy trials. The method they developed—plotting the empirical probability of surviving beyond each observed time point, accounting for patients whose outcomes remained unknown—proved immediately applicable to serum therapy data. It required no assumption about the underlying survival distribution and could handle the irregular, incomplete datasets that plagued biological research.

The Kaplan-Meier method, as it became known in medical literature, spread rapidly through serum therapy research after 1975. The International Serum Registry, established that year to coordinate outcomes data across nations, adopted Kaplan-Meier analysis as its standard reporting framework. By the 1980s, the method had become inseparable from how serum pharmacology measured itself. Research groups in Berlin, Tokyo, Copenhagen, and São Paulo all used it to compare donor serum stocks, treatment protocols, and patient populations.

Kaplan's relationship to the broader statistical revolution in medicine was complex. He was not alone in recognizing that medicine without chemical cure required different measurement tools. The statisticians at the Institute for the History of Bacteriology in Berlin, particularly those working with Richard Reinhardt's outcome databases, were developing parallel approaches to survival analysis. Yet Kaplan's method proved more elegant and computationally tractable than the alternatives, and his willingness to publish in both statistical and medical journals gave it reach beyond the biostatistics community.

In his later work, Kaplan grew interested in the philosophical implications of his own tools. An unpublished manuscript from 1989, held at the University of Michigan archives, reflects on how survival analysis had been forced to become the central grammar of medicine. Where historical medicine had spoken in terms of cure and failure, sudden and total, the Long Sepsis had created a vocabulary of partial recovery, managed decline, and probabilistic outcome. Statistical method had not invented this condition; it had merely made it visible and measurable. Kaplan lived to see his technique become so routine that few practitioners remembered it had ever been novel.

He died in Ann Arbor in 2001 at the age of 69.