WarMED

Wartime medical education and disruptions to physician labour supply: impacts on physician quality, the health transition, and innovation

War disrupts the training of doctors. Curricula are compressed, clinical rotations shortened, teachers conscripted, and whole cohorts of students pushed through medical school under conditions nobody would have chosen. Those cohorts then practise medicine for the next forty years.

WarMED asks what that did — to the quality of the physicians produced, to the health of the populations they served, and to the pace at which new medical knowledge spread.

Three questions

Physician quality. Does the character of a doctor's training show up decades later in their patients' outcomes? Medical education is one of the most expensive human-capital investments any society makes, and yet how much the specific content and duration of it matters is remarkably hard to establish, because training is not normally assigned at random. War comes close to assigning it.

The health transition. What happens to a population's mortality when the supply of physicians is abruptly cut, and then as abruptly restored? This bears directly on a long argument in economic history about how much of the historical decline in mortality is attributable to doctors at all, as against sanitation, nutrition and public health.

Innovation. Does disrupting medical education slow the diffusion of new treatments — or can an unusual cohort, trained under pressure and often in wartime medicine itself, sometimes accelerate it?

Why it is a data problem first

The questions are old. What makes them answerable now is individual-level data that has not previously existed in usable form: who trained as a physician and when, under which curriculum, where they subsequently practised, and what happened to health outcomes in the places they served. That information survives in printed medical directories, medical school and licensing records, and vital statistics — but as archival material, never assembled into a single linked dataset.

Building it is where BDAD contributes: transcription of printed and handwritten sources at scale, structured extraction, and probabilistic linkage of individual physicians across records that were never designed to be joined — the same methods used on printed statistical tables and census enumerator sheets.

In the unit's wider work

WarMED belongs to a continuing strand on the historical determinants of population health, running from the 1918 pandemic studies — school closures and mortality in Sweden, the V-shaped recession in Denmark — through the CIHNR cohort to the present. The shared design is to find a sharp historical disruption, construct the individual-level data around it, and use it to identify effects that aggregates cannot reveal.