By the end of December 2002 the number of cases of confirmed
or probable variant Creutzfeldt-Jakob Disease (CJD) reported in the United
Kingdom (UK) had reached a total of 129 (
http://www.doh.gov.uk/cjd/stats/jan03.htm).
Quarterly analyses are performed that estimate the underlying trend in vCJD
incidence. These enable short term predictions for the expected number of
deaths in the next year as well as estimates of the total number of individuals
with symptom onset yet to be identified (
http://www.cjd.ed.ac.uk/vcjdq.htm).
Previous quarterly analyses of these data have indicated an exponentially
increasing trend (1, 2). The most recent analysis, however (table and
figure), showed
significant evidence of a departure from this pattern; the underlying incidence
was estimated to have reached a peak in 2000 of about 5 deaths and onsets
a quarter and then to have declined to 3.9 deaths per quarter (3.2 onsets)
at the end of 2002.
Table. Annual cases
by onset and death
| Year |
Onset |
Death |
| |
|
|
| 1994 |
8 |
0 |
| 1995 |
10 |
3 |
| 1996 |
11 |
10 |
| 1997 |
14 |
10 |
| 1998 |
17 |
18 |
| 1999 |
29 |
15 |
| 2000 |
24 |
28 |
| 2001 |
16* |
20 |
| 2002 |
0* |
17 |
| Total |
129 |
121 |
* Numbers of cases may rise in subsequent quarters due to reporting delay
A slowing down in the previously increasing trend means that short term
projections for the number of deaths in the next year as well as the numbers
of unobserved onsets are now much lower than in past analyses. An additional
12 cases are expected bringing the total number of cases with onsets by
the end of December 2002 to 141 (95% confidence interval 135 to 151). A
total of 13 deaths were predicted for 2003 (95% confidence interval 5 to
23). The analysis of the deaths data was reported in detail recently (3).
In addition, back calculation models which attempt to estimate the total
size of the epidemic are affected; a recent analysis based on deaths to
the end of 2001 gave a prediction of a total of 80 further deaths by the
year 2080 with a 95% confidence interval of 10 to 7000 (4). Revised estimates
using 2002 data, including upper confidence interval estimates, will be
much lower than those reported previously.
Observed (-o-)
and expected (-e-) quarterly incidence of vCJD onsets. Fitted quadratic
trend* (___) is given with its 95% confidence limits (...).
*includes adjustment
for delay from onset to diagnosis
It is encouraging that trends in deaths from vCJD are consistent with the
epidemic having reached a peak, but this should be interpreted with caution;
it is possible that the epidemic may have a very long tail with some cases
occurring many years after exposure; or we may see more than one peak to
the epidemic at a later date as individuals with non methionine homozygous
genotypes at codon 129 of the PRNP gene (which so far accounts for all cases)
are affected. There is also the theoretical possibility of secondary transmission
through human to human spread. Continued surveillance and research to investigate
these possibilities is therefore important.
The most recent monthly CJD statistics for the UK are available at http://www.info.doh.gov.uk/doh/IntPress.nsf/page/2003-0090?OpenDocument.