Letter to the Editor 218 Int J Sports Med 2010; 31: 218· DOI 10.1055/s-0030-1248327 G. Banfi, L. Drago, G. Lippi: Analytical Variability in Athletes Haematological Testing The interesting paper by Mørkeberg et al. [8] has prompted us to make further considerations on the importance of the preanalytical and analytical phases in antidoping testing. Remarkable differences between various instruments used for analyzing hematological samples have already been demonstrated in a previous paper published in this journal four years ago [10]. In this paper, the variation of hemoglobin (Hb) measured with differ- ent haematological instruments in professional cyclists was extremely wide and analytically unacceptable in the clinical prac- tice. Such a variability might be attributed to different sources. Circadian rhythm for reticulocytes (Ret); presence of hemoconcen- tration or hemodilution; period of training and/or competition sea- son. The biological variability of several hematological parameters depends on the different involvement in training program and psy- chophysical stress; this is particularly evident in cycling, where Hb values tend to decrease during long-lasting and demanding stage races [8, 11], but also in other sports [1], so that the comparison of values between sports disciplines characterized by different inten- sities and duration of training might be unreliable [2]. The accuracy of data is not due to the instrument itself, but to a variety of other issues that include maintenance, calibration, quality assessment of results by internal (IQA) and external pro- grams (EQA). Although an internal quality control method for hematological instruments is usually performed, there is no mention whether this procedure was followed in the study of by Mørkeberg et al. [8] A simple quality control scheme, which is valid also for spot samples and thereby suitable for laboratory analysis of specimens collected from sportsmen, is the Bull algo- rithm, which is based on previous averages of values of a single parameter. The quality control data represents the “analytical passport” of a single instrument, overwhelming the remarkable technological differences. The Union Cyclistique Internationale (UCI) has decided to use Sysmex instruments in order to standardize and harmonize the data among accredited laboratories. This can be seen as a correct decision, but the criteria of choice were not clearly explained. For example, some of the authors of the paper published in the Int J Sports Med [10] have published elsewhere that Ret values (x10 9 /L) measured on a Sysmex XT2000 and tested at two differ- ent temperatures during a period of 72 h from the drawing, decreased after 4, 8, 24 and 48 h at room temperature, whereas the decrease was fairly less significant at 4 ° C. The values after 72 h were both similar to the baseline, but this was an artefact of the hematological system [9]. Conversely, the stability at 24 h was later studied using a Siemens Advia120 on the blood of 25 male professional road cyclists. The samples were collected in the morning and the measurements were either performed within 2 h from venipuncture or 24 h thereafter, after storage at 4 ° C. The decrease was statistically significant, but clinically negligible, so that it was concluded that the samples can be tested until 24 h after drawing, even for legal purposes [5]. In the paper by Mørkeberg et al. [8], the mean values of Hb measured on Beckman Coulter LH750 and Siemens Advia 120 were identical (15 g/dL), but that measured on the Sysmex KX21 was different (14.5 g/dL). Moreover, the Ht values showed a remarkable difference: the authors correctly commented that the practical and empirical rule of Ht/Hb ratio was not met by some measurements. This could be attributed to the different estimation mode of Ht. The Sysmex KX21 has a lower level of technology compared with the other two: the instrument rec- ommended by the UCI is Sysmex XE2100 (or XT2000). However, the highest discrepancy is recorded for Ret: 0.94 on Coulter, 1.05 on Advia, and 0.85 on Sysmex. This is possibly due to the differ- ent technology used for enumerating Ret: methylene blue in Coulter, auramine O in Advia, and oxazine 750 in Sysmex R-500 [3]. The analytical variability of Ret is crucial for evaluating their own measurement, but also for the OFF-model score [4]. After the embarrassing evidence of a poor inter-laboratory agreement of isoelectric focusing testing for detecting blood doping with erythropoietin in urine [7], accurate and reliable haematological results are mandatory to maintain credibility and feasibility of antidoping testing [6]. Foremost in this process is the implementation of a suitable quality control scheme which should be issued and supported by sports federations and agen- cies, and made available to all accredited or recommended labo- ratories, along with specific educational programs for antidoping professionals aimed at reducing the uncertainty during the total testing process. Moreover, the key role of laboratory profession- als in evaluating and interpreting data, organizing quality con- trols programs (or proficiency testing) and accrediting laboratories should be acknowledged. Giuseppe Banfi (Milan), Lorenzo Drago (Milan), Giuseppe Lippi (Parma) References 1 Banfi G, Del Fabbro M, Mauri C, Corsi M, Melegati G. Haematological parameters in elite rugby players during a competitive season. Clin Lab Haematol 2006; 28: 183–188 2 Banfi G, Del Fabbro M. Behaviour of reticulocyte counts and immature reticulocyte fraction during a competitive season in élite athletes of four different sports. Int J Lab Hematol 2007; 29: 127–131 3 Banfi G. Reticulocytes and related parameters in sports medicine. Sports Med 2008; 38: 187–211 4 Gore CJ, Parisotto R, Ashenden MJ, Stray-Gundersen J, Sharpe K, Hopkins W, Emslie KR, Howe C, Trout GJ, Kazlauskas R, Hahn AG. Second-gen- eration blood tests to detect erythropoietin abuse of athletes. Haema- tologica 2003; 88: 333–344 5 Lippi G, Salvagno GL, Solero GP, Franchini M, Guidi GC. Stability of blood cell counts, hematologic parameters and reticulocytes indexes on the Advia A120 hematologic analyzer. J Lab Clin Med 2005; 146: 333–340 6 Lippi G, Banfi G. Preanalytical and analytical issues in indirect haema- tological testing. J Appl Physiol 2008; 105: 1990 7 Lundby C, Achman-Andersen NJ, Thomsen JJ, Norgaard AM, Robach P. Test- ing for recombinant human erythropietin in urine: problems associated with current anti-doping testing. J Appl Physiol 2008; 105: 417–419 8 Mørkeberg JS, Belhage B, Damsgaard R. Changes in blood values in elite cyclist. Int J Sports Med 2009; 30: 130–138 9 Robinson N, Mangin P, Saugy M. Time and temperature dependent changes in red blood cell analytes used for testing recombinant eryth- ropoietin abuse in sports. Clin Lab 2004; 50: 317–323 10 Robinson N, Schattenberg L, Zorzoli M, Mangin P, Saugy M. Haemato- logical analysis conducted at the departure of the Tour de France 2001. Int J Sports Med 2005; 26: 200–207 11 Schumacher YO, Jankovits R, Buttermann D, Schmid A, Berg A. Hematologi- cal indices in elite cyclists. Scand J Med Sci Sports 2002; 12: 301–308 Dr. Giuseppe Banfi IRCCS Galeazzi, Direzione Scientifica Via Galeazzi 4 20161 Milano Italy Tel.: + 39/266214850 Fax: + 39/266214806 giuseppe.banfi1@unimi.it