Centre effects Evaluation of centre and period effects in allogeneic haematopoietic stem cell transplantation in France F Mesnil 1,3 , J-P Jouet 1,2 , P Tuppin 1 , J-P Vernant 2,3 and J-L Golmard 3 1 Etablissement Franc ¸ais des Greffes, Paris, France; 2 Socie´te´ Franc ¸aise de Greffe de Moelle et de The´rapie Cellulaire, Paris, France; and 3 HoˆpitalPitie´-Salpeˆtrie`re,Paris,France Summary: We evaluated the effect of individual and collective factors on the outcome of allogeneic haematopoietic stem cell transplantation (HSCT) at 35 French centres. Individual factors included patient and transplantation characteris- tics. Collective factors were related to the period and centre in which HSCT was performed. Two centre factors were studied: centre experience (ie number of HSCT performed during the study period) and the type of centre (paediatric or adult). All patients receiving a first allogeneic HSCT in France between 1st January 1993 and 31st December 1997 were included in the study. The follow-up period ended on 31st December 1997. The final sample included 2756 subjects. We analysed overall survival (OS) and transplant-related mortality (TRM). Prognostic factors were identified by univariate and multivariate analysis, using Cox models. We found that centre experience had no significant effect on outcome. However, survival rates, whether determined on the basis of OS or TRM, were significantly higher in paediatric centres than in adult centres. Residual heterogeneity was found between adult centres. Survival rates were sig- nificantly higher for HSCT performed after 1st January 1996 than for those performed before this date. Bone Marrow Transplantation (2004) 34, 645–651. doi:10.1038/sj.bmt.1704622 Published online 9 August 2004 Keywords: haematopoietic stem cell transplantation; evaluation; centre effect; period effect; survival analysis; transplant related mortality The last 20 years have seen an increase in the use of allogeneic haematopoietic stem cell transplantation (HSCT) as an alternative to chemotherapy in patients with haematological malignancies. 1 Many European centres, including 36 in France, currently perform allogeneic HSCT. The widespread nature of this activity makes it necessary to evaluate outcomes. Outcome may depend on factors originating from the patient, the disease or aspects of transplantation. Once these factors have been taken into account, efforts should be made to determine the variability in outcome according to the centre at which transplanta- tion was performed or over time. We describe here a retrospective evaluation of allogeneic HSCT outcomes at French centres that was carried out by the Etablissement Franc¸ais des Greffes (EFG), the French Transplantation Agency. This organisation is responsible for recording and evaluating the outcomes of organ, tissue and cell trans- plantation in France. The objective of this evaluation was to determine the effects of centre and period in which HSCT was performed after adjusting for patient and transplantation characteristics. Centre effects have been widely investigated in various medical and surgical fields, 2–10 including organ transplanta- tion. 11–14 Several factors may account for such effects, including the facilities available, procedures used, and the skills and experience available at the centre. Most of the surveys published to date only analysed the influence of experience at the centre. In the field of allogeneic HSCT, four studies have been conducted to date, based on data from the European Group for Blood and Marrow Transplantation, 15,16 the International Bone Marrow Transplantation Registry 17 and the Japan Society for Haematopoietic Cell Transplantation. 18 Three of these studies concerned a specific population of patients with leukaemia receiving grafts from an HLA-identical sib- ling, 15–17 whereas the fourth included patients with various diagnoses and HLA disparities between donor and recipient. 18 One of these studies found that the observed centre effect could be accounted for, at least in part, by centre experience. 16 The other studies simply analysed the effect of centre experience on outcome. No overall consensus was obtained, as three of these studies found a correlation whereas the fourth did not. 15 Improvements in survival over time have been clearly demonstrated for organ transplantation. 19–22 Such improve- ments have also been shown for allogeneic HSCT, in various diseases: chronic-phase chronic myelogenous leukaemia, 23 acute or chronic myelogenous leukaemia, 24 acute lympho- blastic leukaemia in second complete remission, 25 and primary immunodeficiencies. 26 In all these studies, increases in survival were attributed to improvements in supportive care and graft-versus-host disease (GVHD) prevention. Received 4 August 2003; accepted 17 May 2004 Published online 9 August 2004 Correspondence: Dr F Mesnil, Etablissement Franc ¸ais des Greffes, 5 rue Lacue´e, 75012 Paris, France; E-mail: fmesnil001@efg.atrium.rss.fr Bone Marrow Transplantation (2004) 34, 645–651 & 2004 Nature Publishing Group All rights reserved 0268-3369/04 $30.00 www.nature.com/bmt