This term comes from medicine and usually applies to the use of one measure to predict another, usually life expectation/mortality, that takes longer to measure. Typical examples are levels of tumour antigens in the blood.
Arguably, all psychotherapy change/outcome measures are only proxy measures: clients don’t come wanting to have their self-report questionnaire measure score improved at the end of therapy, they hope for longer term change, increased resilience and ability to cope etc. There’s a desperate shortage of even relatively short term, three months to a year say, outcome measures in our field and almost no work on longer term follow-up.
Chapter 4, probably should have been in Chapter 10 too!