I have two samples of survival data. One has known date of exposure
(same as date of entry to observation), while in the other it is
possible that the recorded date of entry is later than the true date.
What I'd like to do is estimate the delay from time of exposure to
entry date for the second sample (possibly with allowance for
covariates) and also establish whether the actual survival distribution
differs from the first sample (eg by estimating the hazard ratio). The
point is that both effects may be operating & I want either a "best
estimate" of the two, or alternatively to show that assuming the true
survival distributions are the same, then the estimated delay is
implausibly long.
I've considered a simple approach by which I compare observed survival
between the samples, but condition on varying delays for the first
sample. The problem here though is that as the delay increases, the
sample size gets smaller and the precision of the hazard ratio
decreases.
Any thoughts on this please?
Paul Silcocks MSc, BM BCh, FRCPath, FFPH, CStat
Medical Advisor
Trent Cancer Registry
Fulwood House
5 Old Fulwood Road
SHEFFIELD S10 3TG
Tel: 0114 226 3563 (direct line)
Fax: 0114 226 3561
E-mail: [log in to unmask]
website: www.trentcancer.nhs.uk
DISCLAIMER: This e-mail and any files transmitted with it are
confidential and intended solely for the use of the individual or
entity
to whom they are addressed. Any views or opinions expressed are those
of
the author and do not represent the views of Trent Cancer Registry or
Sheffield Teaching
Hospitals NHS Foundation Trust unless otherwise explicitly stated. The
information contained in this e-mail may be subject to public
disclosure
under the Freedom of Information Act 2000. Unless the information is
legally exempt from disclosure, the confidentiality of this e-mail and
your reply cannot be guaranteed.
|