運用德菲法發展台灣婦產科低效益照護清單

Use of the Delphi method to develop a list of low-value care practices in obstetrics and gynecology in Taiwan

目標：低效益照護／過度使用漸受關注，如何衡量低效益照護是重要的問題，過去研究發展低效益照護清單多以文獻或臨床醫師凝聚共識，針對台灣婦產科低效益照護清單研究闕如，本研究目的，將以德菲法發展臺灣婦產科低效益照護清單。方法：參考國際明智選擇建議和專家意見，發展婦產科低效益照護問卷。納入醫師為邀請台灣婦產科學會從北中南東醫療院所推薦醫師名單，最終再以德菲法產生低效益照護代表性清單，同意尺度為1分（非常不同意）至9分（非常同意），專家共識定義為中位數≧7且評≧7之比例至少75％，再依評≧7之比例排序以決定前5名清單。結果：本研究共有12名專家參與，兩回合問卷從38項項目中，最終推薦清單共有6項，因排序同名次，7分以上人數比例為91.7％，包括3項檢查（臍動脈都普勒超音波、D-Dimer、紅血球沉降率）、2項處置（催生或剖腹產、子宮切除術）及1項藥品（抗生素）。結論：本研究以德菲法選出6項適合我國國情的清單，政府單位可參考本研究成果發展婦產科臨床指引，能將清單納入且訂出規範；建立資訊整合平台提供低效益照護資訊，不僅有助於醫療人員做出正確決策，也能夠提升民眾對低效益照護之概念。

Objectives: The problem of low-value care/overuse has become a focal point among the medical community. Determining how to identify and measure low-value care is a central concern. Previous studies have formulated lists of low-value care practices based on earlier literature or the consensus of clinicians, but no study has yet to focus on low-value care in obstetrics and gynecology in Taiwan. We used the Delphi method to compile a list of low-value care practices in obstetrics and gynecology in Taiwan. Methods: By referring to recommendations from international Choosing Wisely lists and expert opinion, we developed a questionnaire about low-value care practices in obstetrics and gynecology. The Taiwan Association of Obstetrics and Gynecology recommended a list of suitable doctors for inclusion from the medical institutions in the northern, central, southern and eastern regions of Taiwan, and the Delphi method was then used to produce a representative list of low-value care practices. The agreement scale ranged from 1 (strongly disagree) to 9 (strongly agree). Expert consensus was defined as a median ≥ 7 with at least 75% scoring ≥ 7. All recommendations were ranked by percentage of score ≥ 7 to determine a list of the top five low-care practices. Results: A total of 12 experts completed in two rounds of questionnaires from which six lists of options were selected from a total of 38 recommendations because the rankings were identical. The lists with 91% scoring ≥ 7 included three examinations (umbilical artery Doppler, D-Dimer, and erythrocyte sedimentation rate), two procedures (elective, non-medically indicated inductions of labor or Cesarean deliveries, hysterectomy), and one drug (antibiotics). Conclusions: In this study, the Delphi method yielded six lists suitable for conditions in Taiwan. Health officials can refer to these research results when developing clinical practice guidelines for obstetrics and gynecology that can incorporate indicators and set standards. Establishing an information integration platform to provide low-value care information will strengthen the decision-making of medical practitioners and enhance public understanding of low-value care.