 |
 |
I believe you use 2 markers, whereas we use 3. With typical markers and analysis CVs in the 3-5% region; 1, 2 and 3 markers produce likelihood ratio CVs in the region of 6%, 12% and 17% respectively (modelled data: Ann Clin Biochem (2001) 38:28-36). However, this varies considerably depending on the marker levels (one example from this paper shows a 95%CI for risk of 1:723 - 1:125 for one set of results using 3 markers). In an older study with AFP, t-hCG and uE3 CVs of 4%, 7% and 9% respectively, the likelihood ratio CV varied from 21% to 48% for a group of 5 selected samples (Observed data: BMJ (1991) 302:1275). This is of course accounted for in the calculation of the risk and is not evidence for using fewer markers.It is therefore inevitable that, as the number of markers increases, the incidence of "wrong" results will increase, and the usefulness of repeats declines. Incidentally, reducing the assay CVs reduces the likelihood ratio CV but has little effect on detection rate.If we all move to 3 or 4 markers as the UK National Screening Committee is favouring, repeating the risk calculation will not be very useful. You could repeat tests to ensure that the marker levels measured were correct, but if the medians, DR and IPR are OK, I see no reason to do this either.
----- Original Message -----From: [log in to unmask] href="mailto:[log in to unmask]">Kathryn BrownbillTo: [log in to unmask] href="mailto:[log in to unmask]">[log in to unmask]Sent: Monday, February 04, 2002 1:27 PMSubject: Ethical issue re Down's Syndrome ScreeningAs part of our quality assurance repertoire for the NTD/Down's antenatal
screening programme, if we become aware of a known Down's case that the
screening programme missed we would retrieve the original frozen sample for
re-analysis. We have found this to be a useful retrospective check of the
system and the results have always closely agreed. Therefore, the following debate is hypothetical, but our individual views are polarised and wouldwelcome mailbase readers views......
WHAT WOULD YOU DO IF A SPECIMEN ISSUED WITH A "LOW" DOWN'S RISK RESULT (say 1:2000, cut off 1:250) WAS RE-ANALYSED AND FOUND TO BE "HIGH" RISK OF DOWN'S SYNDROME (say 1:10) IN A KNOWN LIVE BIRTH TRISOMY 21 CASE (eg if a sample mis-match were to be discovered)?
- ?decide to inform the family as being ethically appropriate. Probable litigation,
however money would assist raising handicapped child. Probable media
circus and undermining of the screening programme.-?decide not to inform the family who have probably accepted the child and the
emotional/psychological sequalae that would ensue. Who knows how the
culture will change in, say, 10 years and the information that will have to be
released to the public domain. More media involvement if a suspected
concealment is brought to light.Do you feel that this retrospective quality assurance check of missed cases is appropriate? Or do you favour a front-end approach eg random audit of specimen reception, re-analysis of patient sample from a previous batch.
Minimizing the possibility of any error throughout the system and knowing that the approriate detection rate is achieved, is it better to avoid the potential ethical dilemma raised above.
Your opinions welcomed,
Kathryn Brownbill and John Martin, Royal Bolton Hospital
***************************************************************************
This message may contain information from Bolton Hospitals
NHS Trust which is confidential or privileged. If you are not the
intended recipient, please advise the sender immediately by
reply e-mail and delete this message and any attachments without
retaining a copy. Should you have any questions please reply to:
[log in to unmask]
***************************************************************************