A vaccine that is effective in a trial may be useless in a village, and the reason is rarely biological.
Most vaccines lose outside a narrow temperature range, and the chain that maintains it has many points of failure.
Studies that placed monitors inside shipments found in a substantial share of them.
Freezing is the more common error and the less suspected one.
Staff trained to fear heat set refrigerators too low, and several vaccines are destroyed by cold rather than by warmth.
The damage is invisible, since a frozen looks identical to an intact one after it .
Vaccine monitors address this at a cost of a few cents.
A small label changes colour irreversibly when heat exposure passes a threshold, and a health worker reads it without training.
Their introduction allowed vaccines to leave the deliberately for the last stage of a journey.
Controlled temperature chain protocols now permit several days outside refrigeration, which changes what a campaign can reach.
A health worker can carry doses on a motorbike to that a refrigerated vehicle cannot reach.
The remaining constraint at that scale is electricity rather than equipment.
A clinic with an unreliable supply cannot hold stock, so it orders small quantities frequently and wastes staff time on collection.
Solar refrigerators with direct drive, which need no battery, changed that in several programmes.
Battery-based units failed within three years because nobody had budgeted for replacement, which is the ordinary way such programmes end.
Every element of this is unglamorous and none of it appears in the discussion of vaccine development.
A billion doses produced and a village unreached are not a scientific failure.
They are a failure that a research budget cannot repair and a maintenance budget can.