Advertisement
American Journal of Kidney Diseases

Plasma fibrinogen and dialysis

Published:February 28, 2005DOI:https://doi.org/10.1053/j.ajkd.2004.12.025
      To the Editor:
      In a recently published meta-analysis, Goldwasser et al concluded that plasma fibrinogen level is markedly greater in peritoneal dialysis than hemodialysis patients.
      • Goldwasser P.
      • Feldman J.G.
      • Emiru M.
      • Barth R.H.
      Effect of dialysis modality on plasma fibrinogen concentration A meta-analysis.
      Unfortunately, the investigators had to deal with reports not entirely suitable for meta-analysis. In order to minimize confounding factors such as the time of publication or diabetes mellitus (DM), they needed to stratify the studies into 2 major sets and then the second set into 2 more subsets including patients with or without DM. The final sub-meta-analysis was performed on only 4 studies with 82 peritoneal dialysis and 96 hemodialysis patients, all without DM. Other relevant confounding factors such as age, infections, cause of end-stage renal disease, and therapies could not be analyzed. To test the investigators’ hypothesis, we also analyzed plasma fibrinogen levels in all patients from our dialysis unit. No difference was found between peritoneal dialysis and hemodialysis patients with or without DM (Table 1).
      Table 1Plasma Fibrinogen Levels in Our Population of 67 Dialyzed Patients
      Plasma Fibrinogen, mg/dL
      nMean ± SDRange
      PD all13471 ± 105326–626
      PD with DM2390–494
      PD without DM11476 ± 112326–626
      HD all
      P = not significant v PD all.
      54461 ± 146215–817
      HD with DM15483 ± 121307–737
      HD without DM
      P = not significant v HD with DM and v PD without DM.
      39452 ± 292215–817
      NOTE. To convert fibrinogen in mg/dL to μmol/L, multiply by 0.0294.
      Abbreviations: PD, peritoneal dialysis; HD, hemodialysis.
      low asterisk P = not significant v PD all.
      P = not significant v HD with DM and v PD without DM.
      An increasing number of meta-analyses, often carrying arguable messages, are being published. It is also difficult to avoid biases at the time of the statistical analysis resulting from the lack of randomization, assay methods, and so on, of the original reports. A relevant step forward in the field of meta-analysis could be to open a discussion to set rules for reports that eventually can be meta-analyzed.
      ALL LETTERS TO THE EDITOR MUST BE SUBMITTED ONLINE VIA EDITORIAL MANAGER (http://ajkd.edmgr.com). Letters should be in response to an AJKD article, and that article should have appeared no more than 6 months previously. The title must be different from that of the original article. Letters must not exceed 250 words (excluding references, maximum number 10) and contain no more than 1 figure or table. Letters are subject to editing and abridgment without notice and there is no guarantee that your letter will be published. Submitting the letter constitutes your permission for its publication in any current or subsequent issue or edition of AJKD, in any form or media, now known or hereafter developed.

      Reference

        • Goldwasser P.
        • Feldman J.G.
        • Emiru M.
        • Barth R.H.
        Effect of dialysis modality on plasma fibrinogen concentration.
        Am J Kidney Dis. 2004; 44: 941-949

      Linked Article

      • Effect of dialysis modality on plasma fibrinogen concentration: A meta-analysis
        American Journal of Kidney DiseasesVol. 44Issue 6
        • Preview
          Background: Concentrations of plasma fibrinogen, a vascular risk factor, tend to be greater in patients on peritoneal dialysis (PD) than hemodialysis (HD) therapy, like concentrations of serum cholesterol, lipoprotein(a), and transthyretin, despite the substantial loss of protein during PD. Worse vascular outcome has been noted in PD patients compared with HD patients in several studies. Methods: In this study, the mean difference in plasma fibrinogen levels (PD − HD) was quantified by means of meta-analysis of mean differences found in 12 cohorts with both PD and HD patients (set 1; N = 630) by using a fixed-effects model and meta-analysis of mean fibrinogen values reported in 30 cohorts of patients on a single dialysis modality (set 2; 8 PD cohorts, 22 HD cohorts; N = 2,096) by using a mixed model.
        • Full-Text
        • PDF
      • In reply
        American Journal of Kidney DiseasesVol. 45Issue 4
        • Preview
          Sepe et al express concern about potential confounding among the studies in our meta-analyses and call for discussion of the rules to select reports based on their finding only a 10-mg/dL (0.3-μmol/L) difference in mean plasma fibrinogen between the 13 peritoneal dialysis (PD) and 54 hemodialysis (HD) patients in their small, select sample—less than any value contained in set 1 (67 to 343 mg/dL [2.0 to 10.1 μmol/L]).1
        • Full-Text
        • PDF