Paul MeierBMJ 2011; 343 doi: https://doi.org/10.1136/bmj.d5507 (Published 31 August 2011) Cite this as: BMJ 2011;343:d5507
- Bob Roehr
They are everywhere in biomedical research today, those serrated line graphs of mortality and survival known as Kaplan-Meier curves. They are the product of a complex algorithm that takes into account missing and censored data and translates them into an elegant visual display of information, simultaneously capturing the arc of the group studied and the ragged jolts of individual change, often a death.
Paul Meier was half of the collaboration that created the algorithm, first published in 1958 as “Nonparametric estimation from incomplete observations” in the Journal of the American Statistical Association (1958;53:457-81, doi:10.2307/2281868).
His partnership with Edward L Kaplan was essentially a shotgun marriage arranged by an editor who did not want to publish two similar but somewhat conflicting papers. He told the pair to work it out and produce a single paper. It took four years of further refinements before the collaborative document was ready.
Publication caused barely a ripple, just 25 citations in the ensuing decade. But as computing power grew and became cheaper, researchers …