The risk of transmitting HCV, HBV or HIV by blood transfusion in Victoria
Gordon S Whyte and Helen F Savoia
Readers may print a single copy for personal use. No further reproduction or distribution of the articles should proceed without the permission of the publisher. For permission, contact the Australasian Medical Publishing Company
Journalists are welcome to write news stories based on what they read here, but should acknowledge their source as "an article published on the Internet by The Medical Journal of Australia <http://www.mja.com.au/>".
Abstract - Introduction - Methods - Incident rates - Risk estimation - Results - Hepatitis B virus - Hepatitis C virus - HIV - Discussion - Acknowledgements - References - Authors' details
-
©MJA1997
Schreiber et al. recently reported estimates of the risk of
transfusing blood collected during the infectious "window period"
(the time between a non- reactive blood donation and a repeat blood
donation confirmed positive by the same test) at five United States
blood centres.2 We used a
method similar to that of Schreiber et al. to estimate the incidence of
HBV, HCV and HIV in repeat Victorian blood donors and the risk of
collecting blood infective for the viruses but seronegative by
screening tests. Window-period collections are responsible for
most transmissions of these three viruses.
Patients attending for autologous, directed or therapeutic
donations were excluded, as were donors returning for repeat testing
or counselling only. Plasma donors were not analysed because they are
selected from repeat donors, there are no seroconversions and plasma
undergoes virucidal procedures in fractionation.
In the study by Schreiber et al., seroconversion intervals all lay
within a three-year period.2
In contrast, in our study, seroconversion intervals
were not required to commence in the same period, but were taken back to
the year of the introduction of the screening test for each virus that
was used in 1994-1995 in order to increase the number of
seroconversions available for study. Therefore, the study period
was different for different tests, but required that the reactive
donation was given between March 1994 and December 1995.
The HBV window period is thought to be 59 days (range, 37-87),5 so the previous donation of each of
the two seroconvertors could have been falsely negative for a total
window period of 118 days (range, 74-174) in 6.22 million
person-weeks, or 2.71 per million donations. The risk of giving blood
infective for HBV (i.e., in the window period) was therefore 2.71 per
million donations (range, 1.70-4.00).
Of the 358 332 repeat donations in the 22 months from March 1994, the
first donation of 349 226 interdonational intervals was given after
December 1991, calculated as 165 weeks before the end of January 1995.
The intervals represent 8 221 189 person-weeks, giving an incidence
rate of HCV of 1.89 per 100 000 person-years.
The HCV window period for second generation antibody tests is
considered to be 82 days (range, 54-192).3,6 The risk of donating blood
infective for HCV but seronegative was therefore 246 days (range,
162-576) in 8 221 189 person-weeks, or 4.27 per million donations
(range, 2.82-10.01).
There were 347 076 interdonational intervals after July 1992,
calculated as 134 weeks before January 1995. The donations represent
7 951 347 person-weeks, or 152 911 person-years, giving an incidence
rate of HIV of 1.31 per 100 000 person-years.
The HIV window period for second generation tests is considered to be
22 days (range, 6-38).7 The
risk of donating blood infective for HIV but seronegative was
therefore 44 days (range, 12-76) in 7 951 341 person-weeks, or 0.79 per
million donations (range, 0.22-1.37).
HBV: Schreiber et al. argued that the true risk of a
seronegative donation which is nevertheless infective is higher
than that identified by HBV surface antigen because only 42% of HBV
incident infections persist to be detected by the HBV surface antigen
assay.2 Application of this
adjustment to the risk in Victoria yields a window-period risk of 6.45
per million donations (range, 4.05-9.52). The comparable figure in
the United States is 15.83 per million (range, 6.82-31.97)2 and, in France, 8.45 per million
(range, 2.8-25.2).8
In our study, the unadjusted incidence of HBV in Victorian repeat
volunteer donors was 1.67 per 100 000 person-years, comparable to the
unadjusted incidence in the Australian general population of 2.4 per
100 000 person-years.9 The
similarity of the two figures suggests that the critical factors for
community transmission of HBV have not been identified well enough to
assist in donor selection. During the study period, each time they
donated blood donors signed a form stating that they had not engaged in
male-to-male sex or used intravenous drugs.
HCV: A current estimate of the incidence of HCV in Australia is 7.6 per
100 000 person-years.10 The
estimate has been considered unreliable because of the unlikeliness
that mild cases would be detected, although most of the individuals
tested were more likely to be at high risk. Locarnini et al.
hypothesised that if the number of incident cases were
underestimated by a factor of three, and that 75% were intravenous
drug users, then the true rate could be extrapolated to 22.2 per 100 000
per year.11 In our study, the
10-times-lower incident rate of HCV in repeat donors of 1.89 per 100
000 person-years is evidence of the low-risk behaviour of repeat
volunteer blood donors.
The risk of transmission of HCV by blood transfusion in Victoria in the
window period was 246 days in 8 221 189 person-weeks, or 1 in 234
000 donations (range, 100 000-355 000). The comparable United States
figure is 1 in 103 0002 and, in
France, 1 in 223 000.8
HIV: The incidence of HIV in Australia is
thought to be 480 per year from 1993, or 2.7 per 100 000 person-years.12 Our
study shows that repeat Victorian blood donors have an incidence of
HIV of 1.31 per 100 000 person-years. The
limited reduction in the incidence of HIV in repeat volunteer donors
is evidence of an increasing proportion of seroconversions caused by
activity not identified as high risk. The risk of collecting a
seronegative but HIV-infected donation in the window period is 1 in
1.27 million, similar to the calculation by Dax et al.1
The comparable United States figure is 1 in 493 0002 and, in France, 1 in 571 000.8
The incidence rate of HBV and HIV in regular blood donors is
comparable to that of the general population. This suggests that
donor assessment is ineffective in repeat donors, presumably
because those who contract HBV or HIV do not regard themselves as at
risk by the criteria applied by the blood bank. The incidence rate of
HCV is lower for regular blood donors than the general population. The
relative effectiveness of HCV discrimination presumably reflects
the lack of experimentation by regular donors with intravenous
drugs.
The risk of window-period transmission of HBV, HCV and HIV in Victoria
is low and compares favourably with overseas figures. The risk is
probably overestimated for HIV because of the long seroconversion
intervals. Care should be exercised when generalising from these
figures because of the small number of seroconversions. However, the
medical community and the general public should be reassured by this
evidence that the blood supply is very safe.
No reprints will be available from the author. Correspondence: Dr G S Whyte, PO Box
354, South Melbourne, VIC 3205.
©MJA 1997
<URL: http://www.mja.com.au/>
© 1997 Medical Journal of Australia.
Abstract
Objective: To report the incidence rate of hepatitis
B virus (HBV), hepatitis C virus (HCV) and HIV in Victorian repeat
blood donors and to derive the residual risk of transmission of the
viruses by screened blood transfusion.
Design: The interval from the previous whole blood
donation was extracted retrospectively from Victorian Red Cross
Blood Bank records for each of the 358 332 repeat donations given
between March 1994 and December 1995. Records of repeat donors found
positive for the viruses in this period were traced to the previous
seronegative donation and accepted if screened by the same test. For
each virus, the number of previous donations screened by the same test
was calculated and the sum of all donation intervals used to derive the
incidence of infection in the repeat donor population. Published
intervals after infection (when a donation can be infective although
seronegative) were used to calculate the risk of release of a
seronegative unit which would be infective.
Participants and setting: Homologous blood donors
at the Red Cross Blood Bank of Victoria.
Main outcome measures: Incidence rate of HBV, HCV and
HIV in regular blood donors and risk of infective donations being
seronegative.
Results: The incidence of infection in repeat donors
was: HBV: 1.67 per 100 000 person-years; HCV: 1.89 per 100 000
person-years; and HIV: 1.31 per 100 000 person-years. The risk of a
seronegative repeat donation being infective was: HBV: 2.71 per
million donations (adjusted to 6.45 to account for viraemias which
remain seronegative); HCV: 4.27 per million donations; and HIV: 0.79
per million donations.
Conclusion: The risk of transmitting HCV, HBV or HIV
by repeat blood donors is low and compares favourably with overseas
data. Repeat donors have an incidence rate of HIV and HBV comparable to
that of the general population, but the incidence rate of HCV is lower
for repeat donors than in the general population.
Introduction
There are no current Australian estimates of the risks of
transmission by blood of hepatitis B virus (HBV) or hepatitis C virus
(HCV), although the theoretical risk of HIV transmission has been
estimated to be less than 1 in 903 000.1
Accurate estimates of the risk of
transfusion-transmitted viral infections are important data for
the risk-benefit analysis of homologous blood transfusions and in
assessing the cost-effectiveness of new screening tests or methods
of donor assessment.
Methods
A glossary of terms is shown in the Box. The donation interval was
extracted from Victorian Red Cross Blood Bank records for all repeat
whole blood donations given in Victoria in the 22-month period 1 March
1994 to 31 December 1995. This period was chosen because an algorithm
was finalised in March 1994 to decide whether a donor reactive to a
second generation HCV screening test was truly positive, liable to
transmit HCV and liable to the sequelae of infection.3
The repeat donor population screened by the same test was calculated
to match the seroconversion study period for each disease by assuming
that all repeat donors in the study period gave blood on 31 January 1995
(midpoint of the study period). For HBV, the number of donors was found
whose previous donation was less than 53 weeks before (January 1994);
for HCV, the interval was 165 weeks (December 1991); and, for HIV, the
interval was 134 weeks (July 1992).
Incident rates
The incident rate was calculated as the number of incident cases
(i.e., the number of seroconversions) divided by the sum of the
interdonational intervals, in person-years, as described by Busch
et al.,3 of all the donors in
the study period.
Risk estimation
To derive the residual risk of transmission of each virus, the number
of seroconversions were multiplied by the reported window periods
before seroconversion, expressed in fractions of a year. The product
is the probability that a seroconverting donor gave an infectious
unit of blood during the window period that was not detected as
seropositive by the screening tests currently in use and could
therefore have been given in a blood transfusion.2
Results
Repeat whole blood donors gave 358 332 donations in which the
interdonational interval lay between March 1994 and December 1995.
Half of the interdonational intervals were 12-15 weeks, with none
less than 12 weeks. Ninety per cent of interdonational intervals were
less than 54 weeks.
Hepatitis B virus
There were two seroconversions in the interval covered by the same
screening test. There were 325 534 interdonational intervals after
January 1994, calculated as 53 weeks before the end of January 1995,
representing 6 221 761 person-weeks. The incidence rate of HBV was
therefore 2 in 6 221 761 person-weeks, or 1.67 per 100 000
person-years. Interdonational intervals for the two
seroconvertors were 77 and 178 days.
Hepatitis C virus
There were three seroconverting whole blood repeat donors whose
seronegative donation was after December 1991 and whose second
(reactive) donation was between March 1994 and December 1995.
Interdonational intervals for the three seroconvertors were 96, 651
and 1369 days, respectively.
HIV
There were two seroconverting whole blood repeat donors whose
seronegative donation was after July 1992 and whose reactive
donation was between March 1994 and December 1995. Interdonational
intervals for the two seroconvertors were 279 and 223 days.
Discussion
The effect of modifying the model used by Schreiber et al.3 depends on the length of the
interdonational intervals of the study population compared with the
intervals for seroconvertors. If long interdonational intervals
are characteristic of seroconvertors, there will be a progressive
overestimation of incidence in our model compared with that of
Schreiber et al. This is because the total population of intervals is
skewed strongly towards 12-15 weeks, and long intervals are
under-represented in this study.
Acknowledgements
We wish to thank John Butler, Christine Carroll, Phil Keily and Tony
Chan at the Red Cross Blood Bank Victoria for data collation and
processing, and John McNeil of Monash University for critical review
of the manuscript.
References
(Received 5 Dec 1996, accepted 16 April 1997)
Authors' details
Red Cross Blood Bank, Southbank, VIC.
Gordon S Whyte, FRACP, FRCPA, Director;
Helen F Savoia,
MB BS, Registrar.
E-mail: gwhyte @ rcbbv.org.au
Received 22 December 2024, accepted 22 December 2024
- Gordan S Whyte
- Helen F Savoia