K. Müller-Breitkreutz¹, T. Evers¹ , R. Perry² for the Working Group on Quality Assurance of the European Plasma Fractionation Association *
1. European Plasma Fractionation Association, Amsterdam, The Netherlands
² Scottish National Blood Transfusion Service, Edinburgh, Scotland
* Members of the Working Group are listed in the Appendix

Despite the universal application of virus inactivation and removal technologies during the preparation of plasma products (1,2), the exclusion of infectious donations before further processing and testing remains the first key step for the safety of plasma products and the primary determinant of safety for untreated blood components. Modern selection and screening technologies have reduced the risk of virus transmission to very low levels, but there remains a very small but quantifiable risk of donations escaping routine detection, in particular during the ‘window phase’ of a blood donor infection when the newly infected donor lacks a serological response (3-5). The magnitude of the residual risks, which should be reduced as much as possible, are determined primarily by the rates of relevant infections (HIV and hepatitis B (HBV) and C viruses (HCV) ) in blood and plasma donors. Although risk reduction strategies are designed to recruit donors from low risk populations and to exclude individual donors found to have risk factors, the effectiveness of such policies has until now neither been monitored systemically nor been the subject of standardised reporting. The need for such systems is also underlined by the observation that newly emerging and yet unknown viruses often enter the population through high risk individuals (6).

Not for profit¹ plasma fractionators in Europe believe that the standardised measurement and reporting of such epidemiological data is a key issue for public health. Accordingly, under the auspices of the European Plasma Fractionation Association (EPFA), they are working on a surveillance programme for the monitoring and reporting of such data for their unpaid blood and plasma donors. This report presents the results of a retrospective study whose aim was to gain an overview of the results of donor screening from 1990 to 1996.

Data on the results of routine screening for infection markers in European unpaid blood and plasma donations were collected retrospectively from EPFA member organisations in Belgium, Denmark, Finland, France, Germany, the Netherlands, Switzerland, and the United Kingdom. Since the number of organisations increased during the years, so the overall donor population did not remain identical during the study. The reported data covered nearly all voluntary unpaid donations in the different countries (apart from Germany, whose data covered Lower Saxony and North Rhine Westfalia only) ; most of the donations were of whole blood. The results represent the collection and testing of several million donations from first time and repeat donors each year between 1990 and 1996 for antibody to HIV-1 and 2 (anti-HIV-1 and 2), hepatitis B surface antigen (HBsAg), and antibody to HCV (anti-HCV). The numbers of first time donations included all "potential" donors who were tested for the first time for these viral markers and should reflect the prevalence of the virus markers in the population recruited for donation. Data on repeat donations included all donations of repeat donors. Data on positive donations that could be provided retrospectively did not always contain clear results of confirmatory assays, but the specimens were at least repeatedly reactive in screening assays. The reported data may therefore overestimate the rate of true positive donations.

Data were analysed separately for donations derived from first time and repeat donors.

Weighted marker rates: To gain an overview on the occurrence of viral markers in the whole donor population, the marker rates of the participating organisations were pooled and weighted by the reciprocal weight of the exact 95% confidence intervals using the rate of positive donations and the total number of donations (7). This calculation takes into account the fact that participating organisations represented different numbers of donors and that data obtained from small populations (especially if the marker rates were low) were subject to larger errors. To measure any significant tendency of the weighted virus marker rates over time, the data were analysed by linear regression.

Rates of individual organisations: The virus marker rates of each participating organisation were expressed as minimum, maximum, and median rates for each year. A more detailed analysis was performed to determine whether trends seen in the calculations with the weighted marker rates were common for the majority of the organisations or caused by few organisations that represented a large percentage of the donor population. For this purpose, the data sets derived from the individual organisations were compared. Additionally, mean values of the viral marker rates of the individual organisations (not weighted) were calculated and analysed by linear regression.

Weighted marker rates: The weighted marker rates, which reflect the rates of positives found in the whole donor population, have decreased significantly since 1990. The rates among repeat donations were about 10 times lower than among first time donations (table 1).

Table 1: Rates of anti-HIV 1 + 2 positive donations

First time donations:  anti-HIV 1 + 2 positives

Year

Pooled data, weighted  (positive per
100 000)

Rates found in individual organisations (positive per
100 000)

Total number of donations  x 10⁶ (No of participating organisations)

^{Repeat donations:} anti-HIV 1 + 2 positives

Pooled data, weighted  (positive per
100 000)

Rates found in individual organisations (positive per
100 000)

Total number of donations  x 10⁶ (No of participating organisations)

Rates in individual organisations: A broad range of virus marker rates were seen for first time and repeat donations among the participating organisations (table 1). The marker rates declined particularly for those organisations that started with high marker rates in 1990 (data not shown). These organisations represented a large percentage of the whole donor population and, therefore were the main cause for the decreasing trend in the weighted marker rates (see above). This is also reflected by the fact that the median rates remained stable and low over the years (table 1). On the other hand, the mean anti-HIV rates decreased slightly but significantly with time (data not shown). Furthermore, the range of marker rates became smaller year on year: in 1990, marker rates of first time donations ranged between 2.7 and 46.7 per 100 000 compared with 0.0 to 12.2/100 000 in 1996. For repeat donations, rates were 0.3 to 4.4/100 000 in 1990 and 0.0 to 1.3/100 000 for 1996.

Taken together, these facts indicate a common trend towards a stable low rate of anti-HIV positivity in first time and repeat donations among the organisations.

Weighted marker rates: Results of anti-HCV screening were available for 1991 to 1996. The weighted rates of positive donations from first time and repeat donors decreased significantly, a trend that was substantially more pronounced among repeat donations than in first time donations (table 2). Consequently, the difference between the marker rates of first time and repeat donations became greater with time: the anti-HCV rate in repeat donations was 3.3 times lower than in first time donations in 1990 and 22.6 times lower in 1996.

First time donations:  anti-HCV  positives

Year

Pooled data, weighted  (positive per
100 000)

Rates found in individual organisations (positive per
100 000)

Total number of donations  x 10⁶ (No of participating organisations)

Repeat donations:  anti-HCV  positives

Pooled data, weighted  (positive per
100 000)

Rates found in individual organisations (positive per
100 000)

Total number of donations  x 10⁶ (No of participating organisations)

Rates in individual organisations: Marker rates among first time donors decreased particularly in those organisations that started with high marker rates in 1991 (data not shown). Most organisations reported stable low rates of anti-HCV among first time donors during the later years. This was also reflected by the median (table 2) and mean values (data not shown), which remained stable over time (table 2). Thus, the decreasing slope of the weighted rates among first time donors did not reflect the tendency of all organisations that took part in the study, but was caused by results from a few organisations. In contrast to the first time donations, significant decreases in anti-HCV marker rates were detectable not only for the weighted but also for the unweighted data (mean values) among repeat donations, showing that the tendency to lower marker rates of repeat donations was present in results from all organisations.

It should be noted that the screening assays were less sensitive when the study began, and that this presumably resulted in an underestimate of the true viral marker rates. Two new generations of anti-HCV enzyme immunoassays, which were more sensitive in detecting chronic infections, were introduced in 1992 and in 1993/94 (the latter generation being only slightly more sensitive than the former (8)). On the other hand, the data on HCV collected between 1991 and 1993 were assumed to contain a higher rate of false positives and thus overestimated the true rates because confirmatory assays were not available before 1993. It was therefore impossible to draw firm conclusions from the data collected before 1994.

However, the trends towards lower anti-HCV rates described were also obvious in the years after 1993, in particular among repeat donations (table 2). In conclusion, a clear trend to lower marker rates in repeat donations was seen in all organisations.

Weighted marker rates: Linear regression analysis showed that weighted virus marker rates between 1990 and 1996 for both first time and repeat donations decreased slightly.

Rates in individual organisations: HBsAg screening results differed more widely than for the other viral markers because one organisation (which could provide data only for 1994 to 1996) reported much higher rates than the other organisations (see also the high maximum value of the HBsAg rates among the repeat donations for 1994 to 1996 - table 3). Another organisation started with high HbsAg rates in 1990 and had markedly reduced the rates over time (data not shown). Most of the other organisations reported much lower, stable rates which can also be seen in the median values: they ranged between 74 and 104/100 000 for first time donations and 0.8 and 1.8/100 000 for repeat donations overall. Taken together, HBsAg rates among first time and repeat donors were stable or decreased over time.

First time donations:  HBs Ag  positives

Year

Pooled data, weighted  (positive per
100 000)

Rates found in individual organisations (positive per
100 000)

Total number of donations  x 10⁶ (No of participating organisations)

Repeat donations:  HBs Ag  positives

Pooled data, weighted  (positive per
100 000)

Rates found in individual organisations (positive per
100 000)

Total number of donations  x 10⁶ (No of participating organisations)

We have presented data on the rates of infection markers of more than 50 million blood and plasma donations from unpaid donors in Europe between 1990 and 1996. The marker rates among first time donors suggest that the prevalence of anti-HIV, anti-HCV, and HBsAg in the donor population recruited for donation was stable or decreased with time.

Among repeat donations, a general trend to low marker rates with time was observed for anti-HIV and HBsAg, less pronounced for HBsAg rates. The most pronounced effect was observed for anti-HCV rates, for which all participating organisations substantially reduced the rates of positives during the study. This result may indicate that the effectiveness of donor selection strategies for HCV among all organisations improved.

A recent study from the European Centre for Epidemiological Monitoring of AIDS, detected decreasing anti-HIV rates in donations in the European Union during the same observation period as the EPFA study (9). Rates of anti-HIV in repeat donations were about 11 times lower than in first time donations. These observations are consistent with our findings. Direct comparison of the virus marker rates of the EPFA donations with other studies (10,11) is difficult because there is no internationally uniform way of collecting and expressing such data. It must be recognised that the reported rates among repeat donations did not represent seroconversion rates (incidences) because donations from repeat donors that were tested for the first time with an assay of a new generation (with enhanced sensitivity) were included in the data. Positive results in these cases reflect prevalence rather than incidence. To perform valid and comparable studies, clear and uniform definitions of the donors as well as for the different ways of estimating residual risks are needed (12,13). In order to draw firm conclusions about residual risks of virus transmission by blood and plasma products, future studies should ideally measure: (i) the rate of confirmed seropositivity among first time and candidate donors as a measurement of prevalence in the donor population recruited by the donation system (6,13), (ii) the incidences in first time and repeat donors and interdonation interval of seroconverting donors to estimate the risk resulting from one or more ‘window’ donations (11,13), and (iii) the total rate of confirmed positive donations to estimate the ‘error-risk’ of releasing a positive unit due to errors (14). Future EPFA studies will try to take all these different parameters into account.

The need for a surveillance system to collate and analyse screening data was recently highlighted by the European Commission (15,16) and the United States (6). The development and adoption of such a standardised system for data collection and interpretation is vital for providing evidence on which to base future options to enhance the safety of blood and plasma products. The surveillance system should be linked to the evaluation of risks associated with paid donations (6,17) and of methods for donor selection (15). The present study identified differences between the viral marker rates of individual organisations. These differences, when analysed in the light of donor recruitment and selection strategies, could help to develop the most effective policies. In conclusion, current data suggest that the risk of infectious donations from voluntary unpaid donors entering the blood/plasma supply is low and decreasing. Further improvements in donor surveillance will provide transparency of the residual risk estimates and ensure that the effectiveness and benefits of current and future safety initiatives will be based on scientific evidence.

¹ ‘Not-for-profit’ means that the organisations have no shareholders with financial interest or participation in surpluses. Any surpluses are either retained as reserves or reinvested in projects, research or other activities in the general public interest.

Particular thanks go to the blood banks who provided the data reported in this article and to Dr WP Schaasberg (Amsterdam) for statistical advice.

Dr B Flan, Laboratoire Francais du Fractionnement et des Biotechnologies, Lille, France;
Dr A Gardi, Zentrallaboratorium Blutspendedienst SRK, Bern, Switzerland;
Dr A Hoburg, DRK Blutspendedienst Institut Hagen, Hagen, Germany;
Mrs BW Knudsen, Statens Serum Institut, Copenhagen, Denmark;
Dr J Koistinen, Finnish Red Cross Blood Transfusion Service, Helsinki, Finland;
Dr R Laub, Departement Central de Fractionnement, Croix-Rouge de Belgique, Brussels, Belgium;
Dr H Mohr, DRK-Plasmaverarbeitungsgesellschaft, Springe, Germany;
Dr R Perry (chairman), Scottish National Blood Transfusion Service, Edinburgh, Scotland;
Dr C vd Poel, for the Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam, The Netherlands;
Dr T Snape, Bio Products Laboratory, England.

Author for correspondence:

Dr Konstanze Müller-Breitkreutz
European Plasma Fractionation Association
PB 9190
1006 AD Amsterdam, The Netherlands.
Telephone: +31-20 512 3561
Fax: +31-20 512 3559
email: T.Evers@epfa.nl

1. Suomela H. Inactivation of viruses in blood and blood plasma products. Transfus Med Rev 1993; 7: 42-57.

2. Williamson LM, Allain JP. Virally inactivated fresh frozen plasma. Vox Sang 1995; 69: 159-65.

3. Sloand EM, Pitt E, Klein HG. Safety of the blood supply. JAMA 1995; 274: 1369-73.

4. Sloand EM. Viral risks associated with blood transfusion. Photochem Photobiol 1997; 65: 428-31.

5. Dodd RY. Infectious risk of plasma donations: relationship to safety of intravenous immune globulins. Clin Exp Immunol 1996; 105(Suppl 1): 31-4.

6. United States General Accounting Office: Blood supply. Transfusion-associated risks. Washington DC: GAO/PEMD-97-2, 1997.

7. Sachs L. Applied statistics. New York: Springer-Verlag, 1982.

8. Dow BC, Follett EAC, Davidson F, Calmann M, Lee SR. Comparison of second- and third-generation hepatitis C virus enzyme-linked immunosorbent assays in a low-risk blood donor population. Transfusion 1996; 36: 477-8.

9. European Centre for Epidemiological Monitoring of AIDS. HIV prevalence in blood donations in the European Union. Second quarterly report 1997; 39: 44-8.

10. Riggert J, Schwartz DWM, Uy A, Simson G, Jelinek F, Fabritz H, et al. Risk of hepatitis C virus (HCV) transmission by anti-HCV-negative blood components in Austria and Germany. Ann Hematol 1996; 72: 35-9.

11.Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ. The risk of transfusion transmitted viral infections. N Engl J Med 1996; 334:1685-90.

12. Commission of the European Communities: Proposal for a Council Recommendation, COM(97) 605, Brussels, 17. November 1997.

13. Müller-Breitkreutz K, van der Poel CL, Barbara JAJ, Noel L, Evers T, Dodd RY. Methods for donor surveillance and risk assessment: a workshop report. Presentation at the ISBT Congress Oslo/Norway 1988; Vox Sang (abstract) in press.

14. Schwartz DWM, Simson G, Baumgarten K, Fabritz H, Riggert J, Neumeyer H, et al. Risk of human immunodeficiency virus (HIV) transmission by anti-HIV-negative blood components in Germany and Austria. Ann Hematol 1995; 70: 209-13.

15. Department of Health, Ireland. Conclusions and Recommendations. Colloquium on Blood Safety and Self-sufficiency: An Agenda for the European Community. Adare, County Limerick: September 1996.

16.Gouvras G, Delaney FM. The Dutch presidency and the health interests of the European citizen. Eurohealth 1997; 3: 15 -7.

17. Dormen RE. Paid-versus-volunteer blood donation in the United States: a historical review. Transfus Med Rev 1995; 9: 53-9.