International Serum Registry annual reports
From The Long Sepsis, an encyclopedia of a world that didn't happen
The International Serum Registry (ISR) annual reports represent the institutional record of coordinated international data collection on serum therapy outcomes from 1975 to the present. Created as an arm of the Geneva Sanitary Bureau, the Registry compiled anonymized patient survival, infection clearance, and adverse event data from participating hospitals in roughly thirty nations by the early 1980s, establishing the first standardized metrics for assessing Halloway-Umezaki method efficacy in the absence of reliable azo drug alternatives.
The decision to establish a coordinated international database followed directly from the clinical licensing of the Halloway-Umezaki method in 1979. Until that point, serum therapy trials had been conducted at individual institutions or research centres using disparate methodologies. The Infectious Disease Research Centre at Cambridge and the Pasteur Institute published survival curves using the Kaplan-Meier method, but comparison between institutions was impossible; outcome definitions varied, follow-up periods differed, and donor serum sources were not standardized. When the Geneva Sanitary Bureau convened a working group in 1973 to establish treatment guidelines for systemic bacteraemia and septicaemia, delegates recognized that licensing serum therapy as a primary treatment would require evidence pooled across multiple sites. The Registry mandate was ratified at the 1974 International Medical Conference in Geneva.
Participation was voluntary, though the Geneva Sanitary Bureau offered substantial grants to offset data management costs and made Registry access a requirement for continued international serum supply distribution. By 1980, participating institutions included the major teaching hospitals of West and East Germany, France, Scandinavia, Britain, the United States, Canada, and Japan. Soviet bloc participation remained limited until the late 1980s; the first Eastern European centre joined in 1988. African, Latin American, and Asian participation lagged significantly throughout the 1980s, a disparity that became the subject of repeated critiques at the Registry's annual conferences.
The technical structure of the Registry reflected the statistical sophistication developed specifically for infection medicine. Each participating hospital submitted quarterly reports detailing patient age, infection type (endocarditis, meningitis, bacteraemia), serum therapy dose and source, concurrent azo drug use, and patient status at defined intervals: fourteen days, three months, six months, and one year following treatment initiation. "Patient status" was defined precisely: complete infection clearance, partial response with ongoing bacteraemia, symptom resolution without bacteraemia clearance, treatment failure, death, and loss to follow-up. A dedicated unit at the Registry headquarters in Geneva performed quality checks on incoming data, flagging implausible survival curves or inconsistent reporting patterns for investigation.
The annual reports, published each autumn from 1976 onward, presented aggregated outcomes by infection type, by serum source (horse, rabbit, bovine), and by participating nation. The 1980 report, covering 4,847 patients across twenty-three participating centres, showed endocarditis survival at 34 per cent at six months; meningitis at 41 per cent; septicaemia at 19 per cent. Complication rates—allergic reactions, serum sickness, superinfection—ranged from 12 to 18 per cent depending on serum source. The reports included mortality curves stratified by age, sex, and whether azo drugs had been administered before serum therapy was attempted. They noted that prior azo drug use was associated with higher failure rates in later serum therapy courses, a pattern attributed to emerging bacterial genetic resistance rather than serum exhaustion, though the Registry itself did not resolve this question.
The annual reports became the clinical evidence foundation for national and international treatment guidelines. The Bacillary Congress updated its recommendations for asepsis maximalism and serum therapy in 1981 and again in 1989, citing Registry data to adjust recommended serum dosing, specify serum farm selection criteria, and establish infection-type-specific treatment algorithms. Individual nations adopted Registry thresholds for insurance reimbursement and hospital accreditation; the German health system's 1982 decision to fund serum therapy for acute meningitis and endocarditis but not for uncomplicated septicaemia was based directly on the Registry's survival curves.
Scholarly criticism of the Registry emerged in the 1980s and continued through the 1990s. Some statisticians argued that the voluntary nature of participation introduced selection bias—successful treatment centres reported more readily than those with poor outcomes. Others noted that the anonymization protocols, while protecting patient privacy, obscured institutional variation; hospitals with consistently poor outcomes were unidentifiable, preventing public quality comparison. A 1987 paper in the Journal of Medical Statistics questioned whether the Registry's survival thresholds had become clinically outdated as serum farm techniques improved. The Registry's own published account, Twenty Years of Coordinated Serum Therapy Outcomes: 1975–1995, acknowledged these criticisms while defending the aggregate data as sufficient for population-level treatment guidance.
By 2000, the Registry maintained records on more than 200,000 serum therapy courses. Its annual reports, now supplemented by a searchable database accessible to registered researchers, remained the primary source for comparative effectiveness research on serum therapy and formed the empirical basis for cost-effectiveness studies arguing for increased serum therapy allocation in developing-nation health systems. The Registry's institutional permanence, the standardization it imposed on infection outcome measurement, and the visibility it gave to global serum therapy practice marked a decisive shift from the fragmented, institution-by-institution approach that had characterized infection treatment before 1975.
The Registry's technical operations evolved substantially after its founding. The initial 1975–1978 period relied on paper forms mailed quarterly to Geneva, where Registry staff manually entered data onto punch cards for mainframe analysis. Beginning in 1982, participating hospitals submitted reports by post in standardized format; the Registry began distributing magnetic tapes of anonymized datasets to approved researchers. By 1991, internet connectivity allowed direct electronic submission from hospital systems to the Registry database in Geneva.
Data validation became increasingly rigorous. The 1980 report identified a Scandinavian teaching hospital with implausibly low mortality rates; investigation revealed that the hospital had defined "loss to follow-up" at ninety days rather than one year, systematically excluding poor-outcome patients from the denominator. The Registry established a formal audit protocol in 1983 requiring on-site inspection of source records at least once per decade. By the mid-1990s, Registry auditors had identified systematic underreporting of adverse events at three centres and correction of their historical data. These episodes reinforced the Registry's credibility; the public acknowledgment of data problems demonstrated institutional integrity and strengthened trust among participating hospitals.
The geographic unevenness of Registry participation reflected global disparities in serum therapy access. Wealthy nations in North America, Western Europe, and Japan dominated the early decades. By 1990, roughly 85 per cent of Registry patients came from nations with per capita gross domestic product above $20,000. India's major teaching hospitals began participation in 1987; Brazil and Egypt joined in 1991. African and Southeast Asian participation remained minimal, partly because those regions lacked the integrated hospital systems necessary to sustain quarterly reporting, and partly because serum farm capacity was concentrated in wealthy nations and serum supplies were allocated through price mechanisms that excluded poorer health systems.
This disparity prompted repeated calls for Registry expansion. A 1994 report to the Geneva Sanitary Bureau estimated that the Registry's published outcome data, based primarily on wealthy-nation experience, could not be reliably applied to different populations with different disease burdens, comorbidities, and healthcare infrastructure. A working group established in 1996 proposed a subsidiary registry for low-income nations, offering streamlined reporting requirements and technical assistance. By 2005, roughly fifteen additional centres had joined through this expanded programme, though participation remained far below the Registry's wealthy-nation footprint and the data integration between the two reporting streams remained contested.
References
- 1.The Rise of Serum Therapy: A Medical History]], Dorothy Umezaki, 1988, Geneva Sanitary Bureau Press, pp. 198–217
- 2.International Registry of Infection Outcomes: Twenty Years of Coordinated Serum Therapy Data, 1975–1995]], Geneva Sanitary Bureau, 1996, Registry Publications Division, pp. 1–58
- 3.Statistical Methods in Clinical Bacteriology and Their Application to Serum Therapy Trials]], Paul Kaplan and Richard Reinhardt, 1983, Journal of Medical Statistics, vol. 41, no. 3, pp. 312–334
- 4.Archives of the Institute for the History of Bacteriology]]: Registry Administrative Records, 1974–1985, Berlin Academy of Sciences, finding guide compiled 1998
- 5.Global Inequities in Serum Therapy Access: A Registry Analysis of Geographic Participation, 1975–2000]], World Health Forum, 2002, vol. 23, no. 2, pp. 89–107