There are large amounts of evidence that suggests that low amounts of infectivity are present in the blood of a TSE infected animal. The actual amount appears to vary according to both the strain of disease and the species infected.¹
Initially it was done by inoculation of blood from one animal into another and seeing if the recipient animal developed disease. The finding was that this could be shown, but it required that a large proportion of an animal's blood would be needed to cause the transmission of disease. Rohwer et al at his group in the USA found that white blood carried 80% of the infectivity but that, although platelets carried little, there was still some in red cell fractions and plasma - the blood transfusion. Many research groups around the world have inoculated blood samples or fractions into animals and transferred disease. The chance of transmission appears to be low and hence the level of infectivity present in blood seems to be low also. Actual attempts to measure infectivity levels have suggested 5 IU/ml (Diringer) or 40 IU/ml (Brown).
There is no proof of transfer in blood happening in humans from sCJD²

What can be said is that it is extremely difficult to carry out retrospective research to somehow work out if blood transfer of CJD has taken place in the past. The first reason is that it tended to be younger people that donated blood (generally less than 40 years old) and older people that received it.

With sCJD it is possible to calculate the number of people that would be expected to have received infected blood and yet lived a further 20 years. This represented less than 5% of the CJD cases that were seen (in the UK this would be less than 2 per annum) ³

The thing about vCJD is that a much larger proportion of the donors would have been infected and within a specific time of developing symptoms. When this calculation is done again it is seen that 10% of cases of vCJD would be expected to be as a result of blood transfusion. If accepted, these calculations undermine the methodology and conclusions of Esmonde's work.

The work in mice shows that infectivity is likely to appear in blood starting early in the incubation period at exceptionally low levels and rises to a standard level and then remains similar until shortly before symptoms appear, when it appears to rise again. Dr Brown's excellent research suggests infectivity appears at around 50% of the incubation period, Some of this is guesswork but it would be reasonable to take this position.

A further worry is with vCJD: should we think of this as being a greater problem with blood transmission infection? Part of problem in answering this, is that we do not know how much infection there is in the blood. There is also a concern that in vCJD more infection appears to be found in other peripheral tissues than is found with sCJD, and hence blood may also carry more infectivity.

1 Holada K, Vostal JG, Theisen PW, MacAuley C, Gregori L, Rohwer RG. Scrapie infectivity in hamster blood is not associated with platelets. J Virol. 2002 May;76(9):4649-50.
Brown P. The risk of blood-borne Creutzfeldt--Jakob disease. Dev Biol Stand. 2000;102:53-9.

2 King SM, Watson H, Heurter H, Ricketts M, Elsaadany S. Notifying patients exposed to blood products associated with Creutzfeldt-Jakob disease: theoretical risk for real people. CMAJ. 1998 Oct 6;159(7):771-4.
Ricketts MN. Is Creutzfeldt-Jakob disease transmitted in blood? Is the absence of evidence of risk evidence of the absence of risk? CMAJ. 1997 Nov 15;157(10):1367-70.
Ricketts MN, Cashman NR, Stratton EE, ElSaadany S. Is Creutzfeldt-Jakob disease transmitted in blood? Emerg Infect Dis. 1997 Apr-Jun;3(2):155-63.
Esmonde TF, Will RG, Slattery JM, Knight R, Harries-Jones R, de Silva R,
Matthews WB. Creutzfeldt-Jakob disease and blood transfusion. Lancet. 1993 Jan 23;341(8839):205-7.

3 (Dealler S et al, 1996)