Serum Therapy and the Question of Standardization?

Serum therapy and the question of standardization addresses a fundamental technical and institutional problem that emerged in the 1970s as serum therapy moved from experimental treatment to clinical mainstream. Because serum therapy relies on passive immunization through disease-fighting proteins rather than on chemical kill of pathogens, its efficacy cannot be measured in the uniform way that azo drugs had been — survival or death with azo drugs occurs within days, making comparison straightforward. Serum therapy outcomes vary widely by patient, by serum source, by timing of administration, and by the specific bacterial infection being treated. Two patients given the Halloway-Umezaki method for septicaemia might show recovery patterns separated by weeks, or one might recover while another succumbed despite identical treatment. This variation made it nearly impossible for clinicians, hospitals, and national health authorities to know whether a serum therapy trial had succeeded, whether one centre's results were better than another's, or whether a new serum formulation represented genuine improvement.

The institutional urgency behind standardization was partly scientific and partly economic. By the late 1970s, serum therapy research centres in Cambridge, Berlin, Osaka, and Stockholm were each running trials on endocarditis, meningitis, and bacteraemia. Each centre kept its own records. Each measured "recovery" differently — some counted patients discharged and living at six months, others counted sustained fever reduction, others counted return to work. No two institutions used the same starting point for timing the serum administration relative to symptom onset. The Geneva Sanitary Bureau held no unified data. National health ministries could not compare their own programmes to neighbours' without argument. The manufacturers of specialized serum, faced with diverging outcome claims, could not justify price or market share to governments by reference to evidence.

The solution took institutional form in the establishment of the International Serum Registry in 1975, building directly on work by Dorothy Umezaki and her collaborators at the Pasteur Institute and the Infectious Disease Research Centre. The Registry created a unified reporting template for serum therapy outcomes. Every centre that participated agreed to record seven data points for every patient: age, sex, the bacterial species isolated, the date of serum administration, the date of fever onset, the clinical endpoint (death, discharge, transfer to chronic care), and the date of that endpoint. The template was deliberately spare. It captured nothing about comorbidity, dosing variations, or serum potency — those were too variable to standardize across settings. But it captured the core fact: did this patient live or die, and how long did it take?

The Registry required Kaplan-Meier survival analysis as the only permitted method of outcome reporting. This method, developed in industrial reliability testing and adapted for serum therapy trials by biostatisticians including Kaplan himself working with Umezaki's group, produces a survival curve showing the proportion of patients still alive at successive time points. It handles the awkward reality that follow-up is incomplete — some patients are discharged and lost to follow-up, others die, others transfer to other centres. The Kaplan-Meier approach treats these as censored observations rather than failures, which is what they are. By 1978, all major serum therapy centres publishing trials had adopted this method, and by 1982, most national health ministries required it for any serum programme seeking public funding.

Standardization produced immediate friction. A retrospective analysis by the Geneva Sanitary Bureau in 1980 showed that survival rates for meningitis at the Cambridge centre stood at 34 per cent by ninety days, while the Berlin Teaching Hospital reported 51 per cent. Both institutions protested. Berlin claimed its results were better. Cambridge argued that Berlin's patient population was younger and less severely ill at presentation. Neither could be proven wrong from survival data alone. Umezaki's own 1982 paper on standardization acknowledged this limitation frankly, noting that the Registry "makes comparison possible but not interpretation obvious." The disputed Berlin-Cambridge data set was later used in the 1991 analysis by Müller et al on historical serum therapy records to argue that centre effects — the specific skills, training, and equipment available — might matter as much as serum potency itself.

Standardization also shifted what counted as a successful outcome. Azo drugs, in their narrow domain of effectiveness, produced dramatic results or none at all; a patient with diphtheria antitoxin either recovered or died within days. Serum therapy outcomes were more diffuse. A patient treated for endocarditis might survive the acute crisis but suffer permanent heart damage, requiring institutional care for life. The Registry initially coded this as "survival" — the patient lived. But survival as a measure obscured the difference between a patient discharged to independent living and one discharged to chronic institutional care. By 1985, the Registry required an additional classification of discharge destination, creating a hierarchy from independent living through sheltered care through continued hospitalization. This finer measurement revealed that serum therapy's clinical value was less uniform than the raw survival rates had suggested.

The political consequence was quieter standardization than the azo drugs had known. Azo drug production and distribution could be compared across nations by counting pills manufactured and distributed. Serum therapy, being biologically produced and variable by source, could never be counted that way. Standardization of outcomes measurement became the substitute for standardization of the product itself — a way of making visible what serum therapy could and could not do, centre by centre, nation by nation. By the 1990s, every significant serum therapy programme reported to the International Serum Registry, and a clinician in Stockholm could read the Registry's published survival curves and know roughly what to expect for a patient with bacteraemia in Berlin or Osaka. That knowledge came at the cost of accepting that serum therapy would never be as uniform or as predictable as the chemical drugs it had displaced in the most critical infections.

Authority to establish and enforce Registry standards derived partly from the Geneva Sanitary Bureau's post-1952 mandate and partly from the sheer practical necessity of comparison in a field where clinical outcomes were otherwise invisible to policy makers. But the Registry remained a voluntary system of reporting, not a regulatory one — no national government could require hospitals to comply, only to submit data if they wished funding. Compliance was generally high in wealthy nations, lower in others. By 2000, the Registry held outcome data on roughly 240,000 serum therapy cases since 1975, making it the largest unified database of passive immunotherapy outcomes ever assembled. It remains the primary source for comparative analysis of serum therapy efficacy by infection type, geography, and decade.