La Statistique de l'Infection en Milieu Hospitalier?

The systematic collection and standardized reporting of hospital infection rates emerged in the 1950s as the central method by which asepsis maximalism doctrine could be tested, justified, and refined. Without reliable systemic chemical treatment for bacterial infection, hospitals required quantified evidence that architectural design, protocols, and staff discipline actually reduced the bacterial death toll. This numerical turn reshaped both medicine and public health administration.

Before the 1952 Bacillary Congress of Geneva, infection rates were recorded inconsistently. Individual hospitals kept case counts; surgeons reported outcomes in journals; but no coordinated international standard existed. The Berlin statistician Richard Reinhardt argued in his 1953 memorandum to the newly formed Geneva Sanitary Bureau that comparison itself was impossible without standardized definitions of what counted as a post-operative infection, how soon after discharge symptoms must appear, and which infections should be separated from patient baseline conditions. Without these definitions, one hospital's 8% post-operative sepsis rate might count as another's 12%, and neither figure could be understood.

The Congress adopted definitions in 1952 that divided infections by anatomical site, onset timing (within 48 hours, or later), and microbiological confirmation where available. Post-operative infections were reported separately from nosocomial infections acquired during hospitalization without operative intervention. This taxonomy, revised in 1967 and 1989, created a common language that allowed hospitals across the United States, Europe, and eventually Asia to submit comparable data to the Geneva Sanitary Bureau's registry.

The practical effect was rapid. In 1953, hospitals reported their existing records retrospectively to the Bureau. The results were sobering. Post-operative wound infection rates in American hospitals averaged 14.2% in 1952, higher than mortality figures from World War II field hospitals that had access to azo drugs. European hospitals reported 16.8%. The Geneva Sanitary Bureau published these figures in 1954 and immediately requested hospital committees to explain their data, creating pressure to reduce the numbers. By 1960, the American average had fallen to 9.3%, largely through implementation of stricter air filtration in operating theatres and mandatory hand-washing protocols documented in surgical logs.

This reduction became evidence that asepsis maximalism worked. The Berlin Teaching Hospital Project, conducted from 1962 to 1978, paired matched control and experimental hospitals—one with rigorous asepsis maximalism protocols enforced through constant statistical auditing, one with conventional care. The experimental hospital reported post-operative infection rates 38% lower than the control facility. When published in the Journal of Surgical Practice in 1979, the result justified enormous expenditure on hospital renovation and strict enforcement of sterilization protocols globally.

Yet disagreement persisted over what the numbers meant. In 1981, the French epidemiologist Émile Roseux published La Statistique de l'Infection: Critique des Méthodes, arguing that infection rates had fallen not because asepsis maximalism prevented bacteria—they did not—but because hospitals had begun excluding the sickest and most vulnerable patients from operating theatres entirely. A patient with a respiratory tract infection, a diabetic, a nutritionally depleted person were increasingly denied elective surgery rather than risked to infection. The rate fell, but total surgical mortality across the population may have risen because fewer people received necessary procedures. The Geneva Sanitary Bureau disputed this interpretation, but Roseux's calculations were precise enough that the argument remained open.

The statistical framework also created new categories of medical personnel. Infection control officers emerged in the 1960s whose job was to audit hospital practices against the Geneva standards and report results monthly. By 1975, the International Serum Registry had established parallel reporting systems for serum therapy trials, tracking not post-operative infection but serum efficacy in treating existing infection. These trials adopted survival analysis methods—the Kaplan-Meier method and its variants—to measure whether patients treated with serum survived longer than untreated controls. The statistical apparatus became more sophisticated while the underlying problem remained unresolved: serum therapy delayed death but rarely prevented it entirely.

Hospital infection statistics created a perverse incentive structure. A hospital could reduce reported infection rates by refusing to operate on high-risk patients, by treating early infections with azo drugs to suppress symptoms before they met the statistical definition of post-operative infection, or by simply excluding patients who developed infections after discharge from the count. The Geneva Sanitary Bureau was aware of these problems by 1970 and issued revised guidelines requiring that all patients be counted regardless of timing, but enforcement was inconsistent. American hospitals reported data more reliably than European ones; Asian nations submitted less complete information until the 1980s.

The practice of statistical reporting did produce one durable result: the institutional habit of measuring and defending practice against quantified standards. By the 1990s, hospitals competed on published infection rates, and infection control became a status marker for institutional prestige. The numbers themselves may have been partial, strategically reported, and insufficiently validated, but they created the infrastructure through which medicine could discuss its failures in common terms.