ALT cut-off value for blood donor screening should be established for servicemen separately.

Dear Sir, Recently, the Chinese new standard GB 18467-2011 Health examination criteria of blood donors was put in practice from 1 July 2012. This new version made some amendments, such as the health consulting, donation intervals and volume. For the plasma alanine transaminase (ALT) screening test, requirements changed from <40 IU L−1 to ‘Meets related requirements’. There are still some controversies about whether it is suitable or even necessary for donor blood ALT screening since the 1980s (Silverstein et al., 1984; Khedmat et al., 2007). Owing to the low specificity of ALT activity and the application of nucleic acid test (NAT), the ALT screening of blood donors seems valueless. But in China and some European countries, it is recommended or even required by law (Brinkmann et al., 2003). Usually, the upper limit of ALT activity for blood donors are determined according to the upper limit of the reference intervals for normal person (from 1·5 to 3·0 times) (Brinkmann et al., 2003). In our department, for the old version of GB 18467-2001 required ALT < 40 IU L−1, just fit the upper limit of the reference interval which we used to make it as the cut-off value for blood donor screening. As military hospital was operated and serviced by the PLA army, our department was authorised to collect whole blood from servicemen. In the past few years, we examined the plasma ALT activity of both servicemen and local donors. Results showed that from 2005 to 2007, the discard rate for simply elevated ALT activity of service and local donors were 7·24% (1553/21453) and 3·51% (504/14351), respectively (χ2 = 220·588, P = 0·000). The elevated ALT activity of serviceman may be due to their strenuous physical exertion. The reference interval of plasma ALT activity of the army men needs to be established separately. In 2008, 1720 servicemen were involved in the investigation about plasma ALT activity after obtaining their informed consent. Three millitres of venous blood was collected using lithium heparinate; plasma was separated and analysed within 4 h of blood draw by ADVIA2400 (Semens, Germany). Reference interval was established according to standard Clinical and Laboratory Standards Institute (CLSI) CA28-2 and the work done by Chan et al. (2008). All data were processed by spss 11.0 using non-parametric program for descriptive. The distribution of plasma ALT activity (Fig. 1) does not follow-up the normal