Dapari Rahmat, Ismail Halim, Ismail Rosnah, Ismail Noor Hassim
Community Health Department, National University of Malaysia, Malaysia.
Tanaffos. 2017;16(1):46-52.
The Delphi method has been widely applied in many study areas to systematically gather experts' input on particular topic. Recently, it has become increasingly well known in health related research. This paper applied the Fuzzy Delphi method to enhance the validation of a questionnaire pertaining chronic obstructive pulmonary disease (COPD) risk factors among metal industry workers.
A detailed, predefined list of possible risk factors for COPD among metal industry workers was created through a comprehensive and exhaustive review of literature from 1995 to 2015. The COPD questionnaire were distributed among people identified as occupational, environmental, and hygiene experts. Linguistic variable using Likert scale was used by the expert to indicate their expert judgment of each item. Subsequently, the linguistic variable was converted into a triangular fuzzy number. The average score of the fuzzy number will be used to determine whether the item will be removed or retained.
Ten experts were involved in evaluating 26 items. The experts were in agreement with most of the items, with an average fuzzy number range between 0.429 and 0.800. Two items were removed and three items were added, leaving a total 26 items selected for the COPD risk factors questionnaire. The experts were in disagreement with each other for items F10 and F11 where most of the experts claimed that the question is too subjective and based on self-perception only.
The fuzzy Delphi method enhanced the accuracy of the questionnaire pertaining to COPD risk factors, and decreased the length of the established tools.
德尔菲法已在许多研究领域广泛应用，以系统收集专家对特定主题的意见。最近，它在健康相关研究中越来越知名。本文应用模糊德尔菲法来加强对一份关于金属行业工人慢性阻塞性肺疾病（COPD）风险因素问卷的验证。
通过全面详尽地回顾1995年至2015年的文献，创建了一份关于金属行业工人COPD可能风险因素的详细、预定义清单。COPD问卷分发给被认定为职业、环境和卫生专家的人员。专家使用李克特量表的语言变量来表明他们对每个项目的专业判断。随后，将语言变量转换为三角模糊数。模糊数的平均分将用于确定该项目是否被删除或保留。
10名专家参与评估26个项目。专家们对大多数项目意见一致，平均模糊数范围在0.429至0.800之间。删除了2个项目，增加了3个项目，最终选定26个项目用于COPD风险因素问卷。对于项目F10和F11，专家们意见不一，大多数专家认为这些问题过于主观，仅基于自我认知。
模糊德尔菲法提高了与COPD风险因素相关问卷的准确性，并缩短了既定工具的长度。
