When was the first recorded use, or the first recorded intention of use of hospital readmissions as a summary statistic?
A hospital readmission (also called a rehospitalization) is defined as a hospital admission that is unplanned, but related to a recent hospital discharge. For example, a person who returns to a hospital (and is readmitted) after surgery with an infection would be considered a hospital readmission.
The earliest reference I can find to the collection of a hospital readmission statistic is from Pearl, R. (1921) Modern methods in handling hospital statistics. The Johns Hopkins Hospital Bulletin 32: 185, in which he describes collecting readmissions with a punched card system (shown and described below). I'm not sure if Pearl ever implemented his system while working at Hopkins, but he at least intended to do so.
The fourteenth is a split single column field It records the following information... (c) Whether this case represents the first admission of the patient to the hospital, or whether it is a readmission.
Despite this early reference, I strongly suspect that others have collected these data in previous work. I have investigated a few publications from the likely suspects, including Florence Nightingale, William Farr, and Major Greenwood, all of whom were pioneers in collecting hospital statistics, but I have yet to find anything that specifically mentions hospital readmissions.
Note that when I say "statistic" I don't mean a test statistic in particular, even a simple count would suffice. Even the term "hospital" I am very loosely defining here: if anyone found a clinic or any other place of medical care that measured something similar to hospital readmissions, I would be interested, even if the source was ancient.