Comparison between continuous thoracic epidural and paravertebral blocks for postoperative analgesia in patients undergoing thoracotomy: meta-analysis of clinical trials Research center: Universidade Federal de Alagoas, Maceió, AL, Brazil.

Fabiano Timbó Barbosa Tatiana Rosa Bezerra Wanderley Barbosa Rafael Martins da Cunha About the authors

Dear Editor,

The article entitled "Comparison between continuous thoracic epidural and paravertebral blocks for postoperative analgesia in patients undergoing thoracotomy: systematic review", recently published in the Journal Revista Brasileira de Anestesiologia, brings out the authors' concern to show the anesthetic therapy effectiveness for treating postoperative pain in chest surgery.11. Júnior A de P, Erdmann TR, Santos TV, et al. Comparação entre bloqueios peridural e paravertebral torácicos contínuos para analgesia pós-operatória em pacientes submetidos a toracotomias: revisão sistemática. Rev Bras Anestesiol. 2013;63:433–42.

Reading the scientific article arouses great interest in readers, however, some points need consideration: the software used for calculations, the sensitivity analysis method by successive meta-analysis, the use of mixed-effect model analysis, and the search to identify statistical heterogeneity.

The software used for the search was reported in the Method and References, but the latter is incorrect, it is impossible to identify the place where it is available and to have access to the software for future searches similar to this.

Successive meta-analysis was used by the authors at some point of this systematic review execution to perform the sensitivity analysis, however, the outcome of this analysis was not reported in the results or discussion, which did not clarify its real contribution in this systematic review. This method allows the identification of the likely source of statistical heterogeneity and the exclusion or not of the included article, in an attempt to consolidate the results.22. Bueno NB. Explorando a heterogeneidade. In: Barbosa FT, editor. Introdução à Revisão Sistemática: a Pesquisa do Futuro. 2013, available at: http://bit.ly/lrs01 [accessed 2.1.14].
http://bit.ly/lrs01...

According to the authors, the models of random and fixed effect were used for meta-analysis calculation; however, the random model was chosen to calculate the meta-analysis whenever I22. Bueno NB. Explorando a heterogeneidade. In: Barbosa FT, editor. Introdução à Revisão Sistemática: a Pesquisa do Futuro. 2013, available at: http://bit.ly/lrs01 [accessed 2.1.14].
http://bit.ly/lrs01...
was greater than 30%. In the analysis of variables "assessment of pain at rest after 24 hours" and "incidence of hypotension", the value of I22. Bueno NB. Explorando a heterogeneidade. In: Barbosa FT, editor. Introdução à Revisão Sistemática: a Pesquisa do Futuro. 2013, available at: http://bit.ly/lrs01 [accessed 2.1.14].
http://bit.ly/lrs01...
was lower than that proposed by the authors, not matching the research method description, and also describing the results by the method of random effect instead of fixed effect. The article report does not allow identifying whether this description of the results was due to the authors' consensual decision or a flaw in the research.

The authors considered the presence of heterogeneity as a research bias when they reported "(…) these results may have been biased by the included studies heterogeneity"; however, the presence of heterogeneity does not indicate bias in a systematic review. Tests for heterogeneity are used to determine whether differences between the included studies are genuine (heterogeneity) or if it occurred randomly during the analysis (homogeneity).33. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in metaanalyses. BMJ. 2003;327:557–60. If the differences occurred randomly, the results found in systematic reviews have more credibility, and if heterogeneity is found, the reasons should be carefully evaluated by the authors to consolidate their results and not only be considered a research bias.

It was noted that the statistical heterogeneity, which is present in most analyzes, was underexploited by the authors, and it is possible to disagree with part of their conclusion: "From this systematic review, it is clear that epidural analgesia is associated with a higher incidence of arterial hypotension and urinary retention when it is used for pain control after thoracotomy in adult patients, with evidence level 1A", as level 1A requires minimal or absent heterogeneity or that heterogeneities are properly explored while conducting a systematic review.

In short, I congratulate the authors for the article, which brings important results for understanding postoperative pain in thoracic surgery. Conclusions of systematic reviews are less incisive regarding the clinical significance of their results when those of the included studies differ from each other.33. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in metaanalyses. BMJ. 2003;327:557–60.

References

  • 1
    Júnior A de P, Erdmann TR, Santos TV, et al. Comparação entre bloqueios peridural e paravertebral torácicos contínuos para analgesia pós-operatória em pacientes submetidos a toracotomias: revisão sistemática. Rev Bras Anestesiol. 2013;63:433–42.
  • 2
    Bueno NB. Explorando a heterogeneidade. In: Barbosa FT, editor. Introdução à Revisão Sistemática: a Pesquisa do Futuro. 2013, available at: http://bit.ly/lrs01 [accessed 2.1.14].
    » http://bit.ly/lrs01
  • 3
    Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in metaanalyses. BMJ. 2003;327:557–60.

Publication Dates

  • Publication in this collection
    Sep-Oct 2014
Sociedade Brasileira de Anestesiologia R. Professor Alfredo Gomes, 36, 22251-080 Botafogo RJ Brasil, Tel: +55 21 2537-8100, Fax: +55 21 2537-8188 - Campinas - SP - Brazil
E-mail: bjan@sbahq.org