Delphi studies in social and health sciences—Recommendations for an interdisciplinary standardized reporting (DELPHISTAR). Results of a Delphi study

Abstract

Background:

While different proposals exist for a guideline on reporting Delphi studies, none of them has yet established itself in the health and social sciences and across the range of Delphi variants. This seems critical because empirical studies demonstrate a diversity of modifications in the conduction of Delphi studies and sometimes even errors in the reporting. The aim of the present study is to close this gap and formulate a general reporting guideline. Method:

In an international Delphi procedure, Delphi experts were surveyed online in three rounds to find consensus on a reporting guideline for Delphi studies in the health and social sciences. The respondents were selected via publications of Delphi studies. The preliminary reporting guideline, containing 65 items on five topics and presented for evaluation, had been developed based on a systematic review of the practice of Delphi studies and a systematic review of existing reporting guidelines for Delphi studies. Starting in the second Delphi round, the experts received feedback in the form of mean values, measures of dispersion, a summary of the open-ended responses and their own response in the previous round. The final draft of the reporting guideline contains the items on which at least 75% of the respondents agreed by assigning scale points 6 and 7 on a 7-point Likert scale. Results:

1,072 experts were invited to participate. A total of 91 experts completed the first Delphi round, 69 experts the second round, and 56 experts the third round. Of the 65 items in the first draft of the reporting guideline, consensus was ultimately reached for 38 items addressing the five topics: Title and Abstract (n = 3), Context (n = 7), Method (n = 20), Results (n = 4) and Discussion (n = 4). Items focusing on theoretical research and on dissemination were either rejected or remained subjects of dissent. Discussion:

We assume a high level of acceptance and interdisciplinary suitability regarding the reporting guideline presented here and referred to as the "Delphi studies in social and health sciences–recommendations for an interdisciplinary standardized reporting" (DELPHISTAR). Use of this reporting guideline can substantially improve the ability to compare and evaluate Delphi studies.