Sir
Lambert and Goldacre (BMJ, Nov 21, p 1429) estimate that about 17% of
qualifiers from British medical schools in 1983 and 1988 are not now
working in the NHS. Their calculation, using the capture / recapture
method, depends on the assumption that the probabilies of responding to
either of the two inquiries (MCRG and DoH) are statistically independent
of each other. Suppose, however, that there is a link, ie that a person
answering one inquiry is more likely to respond to the other, and the
converse. A two fold difference in this propensity would increase the
estimated number of additional doctors in the NHS from 147 to 294 and
the participation rate from 83% to 87%. This is a less worrying proportion
and indicates the advisability of including a sensitivity analysis by
varying the basic assumptions as well as the confidence limits for the
error of the method.
GH Hall (Late Chairman, Central Manpower Committee)
--
GH Hall MD, Ballamount, Broadclyst,Devon EX5 3BJ.PH:01392 461515