In Reply:
We thank Dr. Lidal, Ms. Norén, and Dr. Mäkelä for their comments on the Practice Guidelines for Central Venous Access.1 We were quite surprised at a request for more transparency, as the American Society of Anesthesiologists evidence-based process for guideline development is remarkably transparent compared with other published guidelines. Nonetheless, we are delighted to have the opportunity to describe our process.
Regarding our literature search, we agree that information pertaining to search terms and databases used was not specified in the guidelines, although an outline of the general search strategy used in the development of all of our guidelines is readily available.2–4 We typically begin with PubMed and expand our search where appropriate (e.g., Cochrane Systematic Review Database, Cochrane Database of Clinical Trials). The search is supplemented by hand searches of reviewed studies and additional citations provided by the task force, consultants, and others. Searches are typically limited to studies reporting original data published in peer-reviewed, English-language journals. Editorials, letters (unless a relevant case is described), and the reporting of meta-analyses conducted by others are excluded. Inclusion and exclusion criteria for literature can be found in the “Focus” section of the guidelines (pages 539–540). By applying these exclusionary criteria, the search protocol easily reduces to the smaller number of appropriate citations indicated in appendix 5 (page 553). Our two methodologists applied these selection criteria, as part of the search and review process. Articles selected were, in turn, verified by means of a reliability assessment involving the entire 12-member task force. Findings from this assessment are described in appendix 5 (page 554).
The process and criteria for appraising the quality of studies for inclusion and methods used for synthesizing evidence are described under the heading “scientific evidence,” pages 540–541, and in appendix 5 (pages 553–554). Meta-analytic results, including tests for heterogeneity, are reported in appendix 5 (page 554), table 1 (page 555), and throughout the text in the guidelines.
A statement regarding conflict of interest was not included in the guidelines, although participation in the task force includes completion of a standard conflict of interest form. Those with major conflicts of interest are excluded from the task force. Financial support was provided entirely by the American Society of Anesthesiologists. All physician members of the task force were volunteers, who in total contributed hundreds of work hours in developing the guidelines.
The literature search was very thorough and was guided by the evidence linkage interventions described in appendix 5 (pages 553–554) and the inclusion/exclusion criteria in the “Focus” section. In the document, we specifically reported the highest level of evidence, as described on pages 540–541. All of the literature reviewed for these guidelines is available as supplemental digital content from the journal’s website. The reporting format for all of the American Society of Anesthesiologists practice parameters first reports literature-based evidence, followed by opinion-based evidence, and then the recommendations. Evidence, including meta-analytic findings pertinent to the use of catheters containing antimicrobial agents, can be found on pages 542–543. Of further note, the article by Ramos et al.5 cited in your letter, while published subsequent to formulation of the guidelines, would not have been included in our evidentiary database because it addressed long-term use of catheters.
As noted in our “Scientific Evidence” section, we rely first on the best scientific evidence available and use opinion-based surveys and other resources to provide guidance for practicality and feasibility purposes. For these guidelines, we conducted formal surveys of 55 expert consultants and 251 practicing physicians to determine appropriateness and practicality of our recommendations, as well as used additional surveys and informal opinion to assess the validity and feasibility of implementing the guidelines. When scientific evidence is unavailable, opinion-based evidence from experts and from our general membership becomes the only formal resource available to provide guidance to the task force.
The American Society of Anesthesiologists system of literature classification has been an exceptional tool for our purposes, and provides consistency, directness, and mitigation of bias not often available from other systems.6
