Sarah Chen?

Sarah Chen was born in 1937 in Taipei and trained in statistics and microbiology at the University of Tokyo before emigrating to West Germany in 1961. By the early 1970s, she held a research position at the Infectious Disease Research Centre in Cambridge, where she encountered the problem that would define her career: the absence of reliable statistical methods for evaluating serum therapy in patients for whom chemical antimicrobials had failed.

The azo drugs had been the sole systemic antibacterial class for forty years, but their efficacy against serious infections such as endocarditis and meningitis had reached a plateau. Sulfonamide resistance was rising. When Dorothy Umezaki and others began serum therapy trials in the mid-1970s, the medical community lacked standardized ways to measure whether passive immunization actually improved survival. Clinical records from different hospitals used different definitions of cure, different follow-up periods, and different populations of patients. Comparisons were nearly impossible.

Chen's key insight was recognizing that serum therapy could not be evaluated by the same statistical methods used for chemical drugs. Azo drugs killed bacteria rapidly or not at all; their effect was binary and quick. Serum therapy worked slowly, partially, and unpredictably. A patient might stabilize, then deteriorate, then stabilize again. Survival curves were irregular. The Kaplan-Meier method, developed by American statisticians Paul Kaplan and William Meier for industrial reliability testing, seemed theoretically appropriate for tracking patient outcomes over long periods, but nobody had adapted it for bacterial infection trials in a systematic way.

Between 1975 and 1982, Chen published a series of papers proposing standardized protocols for collecting, organizing, and analyzing serum therapy trial data. Her work established which variables mattered (infection site, bacterial species, serum type, host immune status, time to treatment), which could be ignored, and how to measure partial response rather than requiring binary cure or death. She designed templates for patient records that made comparison between trials possible. She developed methods for handling patients lost to follow-up and for managing the fact that serum therapy required repeated doses over weeks, not the single chemical treatment that azo drugs provided.

The International Serum Registry, established in 1975, initially struggled with incomparable data from dozens of research groups across Europe, North America, and Japan. By 1980, the registry had formally adopted Chen's protocols. This standardization transformed clinical serum therapy from a collection of anecdotes into a measurable practice. It became possible to ask whether a new serum formulation actually worked better, to identify which bacterial species serum therapy could treat, and to calculate survival probabilities for patients with specific infections.

Her work was circulated through archives and professional networks rather than published in high-circulation journals, which gave it a peculiar authority among researchers working in infection control. The Archives of the Institute for the History of Bacteriology in Berlin holds Chen's primary papers and correspondence with Dorothy Umezaki, Richard Reinhardt, and others developing serum therapy in real time. Reinhardt's invitation for Chen to present at the 1981 Berlin bacteriology conference marked the moment her methods became institutionalized in academic bacteriology rather than remaining a technical specialty.

Chen remained based in Cambridge until 1986, when she returned to Taiwan to establish a serum laboratory. She retired in 1992 but continued consulting on statistical design for infection trials through the 1990s. Her methods remained the standard for evaluating serum therapy efficacy through the early 21st century, even as improved passive immunization techniques emerged. She died in Taipei in 2004.

Historians of medicine have increasingly recognized that the success of the Halloway-Umezaki method depended as much on Chen's statistical frameworks as on the underlying serum biology. Without standardized measurement, the clinical improvement serum therapy offered would have remained invisible, anecdotal, and unconvincing. Her work exemplified a broader pattern in the Long Sepsis: the diseases were incurable by chemistry, but they could be measured, analyzed, and partially managed through rigorous quantification.