Portal de Periódicos da CAPES

Artigo Acesso aberto Produção Nacional Revisado por pares

BIBTEX RIS

Reply

2009; Lippincott Williams & Wilkins; Volume: 124; Issue: 1 Linguagem: Inglês

10.1097/prs.0b013e3181a83b0f

1529-4242

Daniela Francescato Veiga, Joel Veiga-Filho, Carlos Américo Veiga Damasceno, Yára Juliano, Lydia Masako Ferreira,

Healthcare cost, quality, practices

Sir: Despite its challenging nature, it was very gratifying to receive comments from anesthesiologists regarding our work. Different from plastic surgery, anesthesiology is prodigal in randomized controlled trials. The difficulties in performing randomized controlled trials in surgical issues, particularly in plastic surgery, have been largely discussed previously.1,2 The difficulty in obtaining an appropriate sample, the difficulty in standardizing surgical procedures, and technical differences between surgeons could be emphasized.3,4 As Dr. Mangano and colleagues appropriately pointed out, a sample of thousands of patients would be necessary to state the best antisepsis regimen before clean plastic surgery operations on the trunk. However, it would not be possible to perform the proposed study with such a sample size while minimizing bias attributable to multiple centers, multiple hospitals, multiple surgical teams, and multiple microbiology laboratories. Our trial was designed to attribute eventual differences in colony-forming unit counts and postoperative infection rates to the difference in the antiseptic agent used, by strictly standardizing surgical and laboratory procedures. Also, our sample size was sufficient to verify a statistical difference between the antiseptic agents regarding colony-forming unit counts. Obviously, the trial was not free of bias and had several limitations. For example, because of color differences between the antiseptics used, neither surgeons who scrubbed and collected the samples nor the laboratory technicians who plated the samples were blinded. With regard to statistical analysis, the Mann-Whitney test5 was used to assess the differences between the two groups (povidone-iodine and chlorhexidine) at preantisepsis, after 2 minutes, and at the end of surgery, regarding colony-forming unit counts, and not the Wilcoxon test, as mentioned by Dr. Mangano and colleagues. The Mann-Whitney test is used for independent samples, which was the case in our trial (two different groups of patients), whereas the Wilcoxon test is used for nonindependent samples. Furthermore, Friedman two-way analysis5 of variance was used to assess the differences among the three moments (preantisepsis, after 2 minutes, and at the end of surgery). This analysis was applied independently for each antisepsis regimen (Tables 1 and 2).Table 1: CFU Counts in the Povidone-Iodine GroupTable 2: CFU Counts in the Chlorhexidine GroupWe also used Spearman rank-order correlation5 to assess correlation between duration of surgery and skin colonization at the end of the procedure. The chi-square test5 and Fisher’s exact test5 were used to compare povidone-iodine and chlorhexidine groups regarding Staphylococcus aureus colonization and postoperative infection occurrence, respectively. Because our article was accepted as a Viewpoint article, we did not have enough space to appropriately describe the statistical method. The variables in our trial were not normally distributed, and the more appropriate way would be to report them as median, instead of mean and SD, as well observed by Dr. Mangano and colleagues. However, because we found many median values that were 0 (Tables 1 and 2), and again considering the lack of space, we believed that mean and SD values would be more elucidating to the reader. We thank Dr. Mangano et al. very much for their interest in our trial. Daniela F. Veiga, M.D., Ph.D. Joel Veiga-Filho, M.D. Division of Plastic Surgery Department of Surgery Universidade do Vale do Sapucaí Pouso Alegre, Brazil Division of Plastic Surgery Department of Surgery Universidade Federal de São Paulo São Paulo, Brazil Carlos Américo Veiga Damasceno, Ph.D. Department of Microbiology Universidade do Vale do Sapucaí Pouso Alegre, Brazil Yara Juliano, Ph.D. Department of Bioestatistics Universidade do Vale do Sapucaí Pouso Alegre, Brazil Department of Bioestatistics Universidade Federal de São Paulo São Paulo, Brazil Lydia M. Ferreira, M.D., Ph.D. Division of Plastic Surgery Department of Surgery Universidade Federal de São Paulo São Paulo, Brazil DISCLOSURE None of the authors has a financial interest in the content of this reply or of the article being discussed.