The Lancet, March 20: Deaths from variant Creutzfeldt-Jakob disease

[Guest guest]

Dear All,

I've not been able to read all the mail that came to CJD Voice and CJD

Blood in

my recent absence, but [similarly] am passing on the following research

article

from The Lancet, March 20, 1999, on the off-chance that you may have missed

seeing it.

Best wishes etc., Lynette.

http://www.thelancet.com/newlancet/reg/issues/vol353no9157/body.research97

9.html>http://www.thelancet.com/newlancet/reg/issues/vol353no9157/body.rese

arch979.html

Volume 353, Number 9157 20 March 1999

 

Deaths from variant Creutzfeldt-Jakob disease

R G Will, S N Cousens, C P Farrington, P G , R S G Knight, J W Ironside

See Editorial

Variant Creutzfeldt-Jakob disease (CJD) was first described and putatively

linked to bovine spongiform encephalopathy (BSE) in 1996.1,2 Since then,

evidence in support of a causative link has strengthened.3,4 BSE was first

formally identified in 19865 and in the following years the incidence of

confirmed cases increased rapidly, from about 60 cases in 1986, to more than

600 in 1987, and more than 3000 in 1988. The death rate from variant CJD per

year has remained approximately constant until recently. We report an

unusually

high number of deaths from variant CJD in the last quarter of 1998.

Up to March, 2, 1999, 39 deaths from variant CJD had been reported to the

National Creutzfeldt-Jakob Disease Surveillance Unit (table). From the

beginning of 1996 to the end of the third quarter of 1998, the number of

deaths

per quarter appears relatively constant, with the largest number of deaths

in a

single quarter, five, occurring in the first quarter of 1996. In the last

quarter of 1998 there were nine deaths. Given the total number of deaths (35)

observed over the 3 years, 1996-1998, the probability of observing nine or

more

deaths in a quarter, if the death rate from variant CJD was constant over the

period, is low (p=0·025). One possible explanation is that the underlying rate

of mortality has been increasing over time. If deaths have been occurring at a

constant rate from sometime in 1995 onwards, then the plot of cumulative

mortality (figure) should be linear (apart from any random variation).

However,

from the second quarter of 1995 to the end of 1998 there is some evidence of

non-linearity (p=0·03), compatible with an increasing mortality rate.

Restricting this analysis to the period up to and including the third quarter

of 1998 reveals no evidence of increasing mortality.

----------

Year Quarter        

  1st quarter 2nd quarter 3rd quarter 4th quarter Total

1995 0 1 0 2 3

1996 5 3 1 1 10

1997 4 3 1 2 10

1998 2 2 2 9 15

1999 1* ·· ·· ·· 1

*Data incomplete.

Mortality from variant CJD in UK by quarter

----------

The unusually high number of deaths from variant CJD occurring in the last

quarter of 1998 should be interpreted with caution. The recorded number of

deaths could have increased because of improvements in case ascertainment.

However, we consider that ascertainment since 1996 is likely to have been

almost complete over the age range of observed cases (about 15-55 years).

Furthermore, even if there had been under-ascertainment of cases, it seems

unlikely that there would have been a sudden improvement towards the end of

1998. The occurrence of variant CJD cases has been monitored closely since

1996

and this is not the first occasion on which statistical analyses to test for a

change in the mortality rate have been done. The p values quoted take no

account of sequential hypothesis testing and therefore overstate the real

statistical significance of this observation. This analysis takes no

account of

delays in reporting and confirmation of deaths from variant CJD to the

surveillance unit. However, 41% of the 39 cases were reported and diagnosed

within 1 month of death, 79% within 2 months, 97% within 5 months, and all

within 6 months. Thus any deaths from variant CJD that have already occurred

but that have yet to be reported or confirmed are only likely to affect the

numbers of deaths as far back as the third quarter of 1998. The identification

of additional deaths in the two or three most recent quarters would tend to

strengthen the evidence in favour of increasing mortality.

Cumulative mortality from variant CJD in UK

Although the number of deaths in the last quarter of 1998 was unusual, we do

not know if this level of mortality will be sustained. The number of variant

CJD deaths during the coming years will provide a clearer indication of

whether

the apparent increase in deaths towards the end of 1998 was a chance

observation or marks a change in the underlying mortality rate.

1 Statement by the UK Spongiform Encephalopathy Advisory Committee. House of

Commons, 20th March 1996, HMSO.

2 Will RG, Ironside JW, Zeidler M, et al. A new variant of Creutzfeldt-Jakob

disease in the UK. Lancet 1996; 347: 921-25.

3 Bruce ME, Will RG, Ironside JW, et al. Transmission to mice indicate

indicate

that 'new variant' CJD is caused by the BSE agent. Nature 1997; 389: 498-501.

4 Collinge J, Sidle KCL, Meads J, et al. Molecular analysis of prion strain

variation and the aetiology of 'new variant' CJD. Nature 1996; 383: 685-90.

5 Wells GAH, AC, CT, et al. A novel progressive spongiform

encephalopathy in cattle. Vet Rec 1987; 121: 419-20.

----------

National Creutzfeldt-Jakob Disease Surveillance Unit, Western General

Hospital,

Crewe Road, Edinburgh; Department of Infectious and Tropical Diseases, London

School of Hygiene and Tropical Medicine, Keppel Street. London WC1E 7HT, UK

(S N

Cousens e-mail: s.cousens@...s.cousens@...); and

Department of Statistics, The Open University, Milton Keynes