Description of the surgical technique for condylectomy with minimally invasive surgery to treat interdigital helomas on the lesser toes: a Delphi study

Initially, a coordinating group was established to select a panel of experts to define the items or aspects needed to describe the technique under study. Their work would be included in a questionnaire used to interpret partial results and issue reports about the results after each round. Furthermore, a final report would be prepared describing the technique for condylectomy via MIS on the interdigital helomas of the lesser toes.

A multidisciplinary international panel of ten experts with accredited experience in MIS of the foot was formed. The panel comprised eight Spanish podiatrist professors of MIS of the foot at Spanish universities. One member was president of the Spanish Association of Minimally Invasive Surgery of the Foot (Asociación Española de Cirugía de Mínima Incisión del Pie; AEMIS), two were ex-presidents of the AEMIS, and six were professors accredited by the AAFAS (Academy of Ambulatory Foot & Ankle Surgery). Two panellists were traumatologists (one from Mexico, one from Argentina), one of whom had the title “Distinguished Professor of AAFAS” and was a former vice-president of the AAFAS; the other was a university professor of foot surgery. A lack of agreement on the sample size of Delphi studies exists; however, a minimum of seven and a maximum of 30 is recommended [30].

First, the coordinating group defined the items needed to describe the technique in each question (Table 1). To address these items, a structured questionnaire composed of 43 statements was designed based on the surgical techniques described by the authors in their manuals or publications [15‐19]. Using this questionnaire, participants reported to what extent they agreed or disagreed with each statement using a five-point Likert scale with the following levels: Strongly disagree/Disagree/Unsure/Agree/Strongly agree). In addition, the questionnaire offered the panellists the possibility of including free text in case they disagreed with any of the statements about a certain item. To make the questionnaire more dynamic, images were added to each statement to clarify the text. The questionnaire was distributed to the experts in January 2016 in an electronic format, which guaranteed the anonymity of the responses.

The experts sent a letter of introduction via e-mail that included the information suggested by Landeta [31] inviting them to participate in the study. After providing consent to participate, the experts completed the electronic questionnaire within 2 weeks after receiving the e-mail. This e-mail provided a brief introduction and the link to the questionnaire. The questionnaires were completed anonymously.

Delphi questionnaire responses were instantly recorded in a Microsoft Excel spreadsheet. After each round, the coordinating group analysed the quantitative and qualitative data (free text). The next round was based on the analysis of the responses received in the previous round. The number of rounds was determined by time taken to reach consensus on each item.

After each round, the central tendency measure of the values assigned to each item was calculated for every answer. Thus, the median, mean, and standard deviation of each statement were calculated and reordered based on the average values obtained within each item. The coordinating group considered statements with an average of less than 3 within each item as the experts agreeing that this option was not the most appropriate. The three statements were repeated in the next round of the questionnaire for cases in which none of the three statements reached a score of 3. Therefore, in the second round of the questionnaire, statements with an average score equal to or greater than 3 were included, as were those in which none of the three statements reached the minimum score of 3 within the same item. Statements without a clear statistical difference were also included. At the end of each round, a summary of results was sent to the panel of experts. The data were analysed with the SPSS statistical program (SPSS®, Version 22, Chicago (IL), USA).

First, the Delphi method is not exempt from the possible influence of the researchers because they participated in the development of the questionnaire, the analysis of the partial results, and the editing of the final report. Second, the criteria used to select the experts might not have adequately identified participants with sufficient clinical experience. However, we believe that we selected the professionals with the most experience in this field. In addition, the participation of different types of specialists (podiatrists and traumatologists) enriched this study, offering different points of view. Third, the results of this study might not reflect the opinions of experts from around the world because eight participants were from Spain, and only two additional countries were represented by the remaining two experts. Although the number of panellists is small, a small sample can be considered effective when the experts have similar levels of training and a general understanding of the field of interest [30]. Fourth, the line of progression given when evaluating the proposed statements might have led the experts to give the highest scores when expressing their agreement with the statement, despite it not being what they would perform in clinical practice, because they nevertheless consider it valid. Therefore, this aspect might introduce information bias. Finally, expert opinion is known as level V of evidence-based medicine but this type of study which collects the experience of experts is needed to improve the quality of treatment for patients [32].