COCHRANE ROB 2 · 偏誤風險評讀
糖尿病發生率
← 回到「糖尿病」的圖表
總體判斷 OVERALL+Low(低風險)
Domain 與 overall 判斷由程式依 Cochrane RoB 2 官方演算法計算;AI 只負責回答下方各訊號問題。符號:+ 低風險、? 有些疑慮、- 高風險(不只靠顏色)。
風險偏誤紅綠燈
- +Domain 1 隨機化過程:Low(低風險)
- +Domain 2 偏離預定介入:Low(低風險)
- +Domain 3 缺失結局資料:Low(低風險)
- +Domain 4 結局測量:Low(低風險)
- +Domain 5 選擇性報告:Low(低風險)
- +總體(Overall):Low(低風險)
中文摘要(給一般讀者)
這篇論文的偏誤風險與 limitation
FOCUS 為跨 35 中心、雙盲、安慰劑對照之大型 RCT,採區組隨機化、由獨立統計師以 SAS 產生序列並以 IVRS 中央分配,分配隱匿與隨機化過程穩固,基線兩組均衡(隨機化 domain 低風險)。安慰劑顆粒在各感官面向完整匹配,受試者、研究者與給藥人員皆盲,主要分析以 FAS(885/889)依隨機分組進行 ITT,僅事後排除 4 名不合格者,介入偏差 domain 低風險。結局糖尿病發生率以客觀的連續兩次 OGTT 判定並經獨立終點裁決委員會盲性確認,測量偏誤 domain 低風險。雖共 111 人退出、完成者 778/885,但結局採時間至事件分析、退出在兩組均衡且多因 COVID-19 封控,並有敏感度分析支持穩健,缺失資料偏誤風險低。試驗於 ChiCTR 註冊、SAP 公開於補充材料且分析依事先計畫,選擇性報告風險低。整體各 domain 均為低風險。
詳細評讀(給專業讀者)
Domain 1:隨機化過程 — 判斷:Low
偏誤來源:隨機分組的過程——分配序列是否隨機產生、是否隱蔽到分組為止、基線(第一期開始時)是否平衡。
1.1 Was the allocation sequence random?
分配序列是否隨機產生?
- 回答:是(Yes — 有明確證據支持)
- 原文引述:a block randomized design was used (eMethods in Supplement 1). An interactive web response system was used to centrally manage participant allocation. Randomization numbers were generated independently by a statistician from the Peking University Clinical Institute using SAS, version 9.4 (SAS Institute Inc).採用區組隨機化設計,並以互動式網路回應系統集中管理受試者分配。隨機號碼由北京大學臨床研究所的一位統計師獨立以 SAS 9.4 版產生。
- 理由:由獨立統計師以 SAS 軟體產生隨機號碼、採區組隨機化,明確使用電腦隨機成分,序列為隨機。
1.2 Was the allocation sequence concealed until participants were enrolled and assigned to interventions?
分配是否隱蔽到分組為止?
- 回答:是(Yes — 有明確證據支持)
- 原文引述:An interactive web response system was used to centrally manage participant allocation... A comprehensive blinding protocol was implemented to safeguard treatment allocation concealment prior to randomization among participants, investigators, and other trial personnel.以互動式網路回應系統集中管理受試者分配……並實施完整的盲性方案,在隨機化前對受試者、研究者及其他試驗人員保護治療分配的隱匿。
- 理由:採用互動式網路回應系統(IVRS)中央集中分配,屬遠端集中隨機,分配序列在受試者納入並指派介入前對招募者隱匿,符合分配隱匿要件。
1.3 Did baseline differences between intervention groups suggest a problem with the randomization process?
(第一期開始時)基線差異是否暗示隨機化有問題?
- 回答:否(No — 有明確證據支持)
- 原文引述:The JLD and placebo groups had similar distributions for age, sex, waist circumference, fasting and 2-hour postprandial glucose levels, HbA1c, fasting insulin level, blood pressure, lipid levels, HOMA-IR, ABI, CIMT, and the prevalence of comorbidities, use of antihypertensive agents, and use of lipid-lowering agents.津力達組與安慰劑組在年齡、性別、腰圍、空腹及餐後 2 小時血糖、糖化血色素、空腹胰島素、血壓、血脂、HOMA-IR、ABI、CIMT 及共病盛行率、降壓藥與降脂藥使用上分布相近。
- 理由:基線特徵表(n=442 對 443,幾近 1:1)兩組各項指標分布相近,無顯示隨機化問題的失衡,組別人數亦符合預定 1:1 比例。
Domain 2:偏離預定介入 — 判斷:Low
偏誤來源:實際執行時偏離預定的介入——盲性如何、是否依「隨機分到的組別」分析(ITT)。
2.1 Were participants aware of their assigned intervention during the trial?
參與者是否(在每一期)知道自己的分組?
- 回答:否(No — 有明確證據支持)
- 原文引述:To maintain blinding integrity, placebo granules were meticulously matched with JLD in terms of color, odor, taste, shape, texture, specifications, appearance, packaging, labeling, and identification, making it virtually impossible for participants to distinguish them.為維持盲性完整,安慰劑顆粒在顏色、氣味、味道、形狀、質地、規格、外觀、包裝、標籤與辨識上均與津力達精心匹配,使受試者幾乎無法分辨。
- 理由:雙盲、安慰劑顆粒於各感官面向與試驗藥完整匹配,受試者無法辨識所分配介入;且不良事件兩組相近、無特異於某組之毒性可破盲。
2.2 Were carers and people delivering the interventions aware of participants' assigned intervention during the trial?
照護/執行者是否(在每一期)知道分組?
- 回答:否(No — 有明確證據支持)
- 原文引述:A comprehensive blinding protocol was implemented to safeguard treatment allocation concealment prior to randomization among participants, investigators, and other trial personnel.實施完整盲性方案,於隨機化前對受試者、研究者及其他試驗人員保護治療分配的隱匿。
- 理由:研究者與其他試驗人員同受盲;安慰劑與試驗藥外觀無從分辨、採中央 IVRS 分配,照護者與給藥人員無從得知分組。
2.3 Were there deviations from the intended intervention that arose because of the trial context?
是否因試驗情境造成偏離?
- 回答:不需評估(NA)
- 原因:條件題:依 2.1/2.2 的答案,本題依 RoB 2 官方路由不需評估(NA)。
2.4 Were these deviations likely to have affected the outcome?
這些偏離是否可能影響結局?
- 回答:不需評估(NA)
- 原因:條件題:依 2.3 的答案,本題依 RoB 2 官方路由不需評估(NA)。
2.5 Were these deviations from intended intervention balanced between groups?
這些偏離在兩組間是否平衡?
- 回答:不需評估(NA)
- 原因:條件題:依 2.4 的答案,本題依 RoB 2 官方路由不需評估(NA)。
2.6 Was an appropriate analysis used to estimate the effect of assignment to intervention?
是否用了適當分析(ITT)?
- 回答:是(Yes — 有明確證據支持)
- 原文引述:The full analysis set (FAS) for primary and secondary end point analyses comprised all randomized participants meeting inclusion criteria and receiving at least 1 study drug dose... A total of 889 participants were randomized, of whom 885 were in the full analysis set.主要與次要終點分析的全分析集(FAS)涵蓋所有符合納入條件且至少接受 1 劑研究藥物的隨機受試者……共 889 人隨機分組,其中 885 人納入全分析集。
- 理由:以 FAS(885/889)進行 ITT 主要分析,依隨機分組分析;僅 4 名(不符納入條件之事後排除)未納入,屬可接受的事後排除不合格者,分析適當以估計分配效應。
2.7 Was there potential for a substantial impact (on the result) of the failure to analyse participants in the group to which they were randomized?
未依隨機分組分析,是否可能對結果造成重大影響?
- 回答:不需評估(NA)
- 原因:條件題:依 2.6 的答案,本題依 RoB 2 官方路由不需評估(NA)。
Domain 3:缺失結局資料 — 判斷:Low
偏誤來源:結局資料缺失——是否幾乎所有人都有資料、缺失是否可能與病情(結局真實值)相關。
3.1 Were data for this outcome available for all, or nearly all, participants randomized?
是否幾乎所有受試者都有此結局的資料?
- 回答:可能否(Probably no — 綜合判斷後傾向「否」)
- 原文引述:Among initial participants, 111 withdrew (55 in the JLD group and 56 in the placebo group)... Ultimately, 885 participants (442 in the JLD group and 443 in the placebo group) comprised the FAS... 388 Completed the trial (JLD); 390 Completed the trial (placebo).起初受試者中 111 人退出(津力達組 55 人、安慰劑組 56 人)……最終 885 人(津力達 442、安慰劑 443)納入 FAS;完成試驗者津力達組 388 人、安慰劑組 390 人。
- 理由:共 111 人退出(含因 COVID-19 失訪 31 人、不願續留 62 人等),完成者僅 778/885。結局為時間至事件(每 3 個月 OGTT),退出者後續結局狀態不完整,並非幾近全部受試者皆有結局資料。
3.2 Is there evidence that the result was not biased by missing outcome data?
是否有證據顯示結果未受缺失資料偏誤?
- 回答:可能是(Probably yes — 綜合判斷後傾向「是」)
- 原文引述:The sensitivity analysis results suggest that the results of the FAS analysis were robust (eTable 1 in Supplement 1)... Annual incidence rates per person-year were 13.70% and 22.05% in the JLD and placebo groups.敏感度分析結果顯示 FAS 分析結果穩健(補充材料 1 eTable 1)……以人年計算之年發生率,津力達組與安慰劑組分別為 13.70% 與 22.05%。
- 理由:採時間至事件之 Cox/log-rank 分析並以人年發生率處理追蹤時間差異,退出在兩組均衡(55 對 56;COVID 失訪 15 對 16),且有敏感度分析支持結果穩健,缺失資料造成偏誤的可能性低。
3.3 Could missingness in the outcome depend on its true value?
缺失是否可能取決於結局真實值?
- 回答:不需評估(NA)
- 原因:條件題:依 3.2 的答案,本題依 RoB 2 官方路由不需評估(NA)。
3.4 Is it likely that missingness in the outcome depended on its true value?
缺失是否很可能取決於結局真實值?
- 回答:不需評估(NA)
- 原因:條件題:依 3.3 的答案,本題依 RoB 2 官方路由不需評估(NA)。
Domain 4:結局測量 — 判斷:Low
偏誤來源:結局的測量——測量工具是否恰當、評估者是否盲化、評估是否可能受分組知情影響。
4.1 Was the method of measuring the outcome inappropriate?
結局測量方法是否不適當?
- 回答:否(No — 有明確證據支持)
- 原文引述:The primary outcome was the occurrence of diabetes, which was determined by 2 consecutive oral glucose tolerance tests... After a center's principal investigator determined that a participant had reached the end point, the participant's data were submitted to the end point adjudication committee for final confirmation.主要結局為糖尿病的發生,以連續兩次口服葡萄糖耐量試驗判定……由中心主持人判定受試者達到終點後,其資料送交終點裁決委員會作最終確認。
- 理由:以客觀實驗室標準(連續兩次 OGTT、依 2017 中國指南)定義糖尿病,並經獨立終點裁決委員會確認,測量方法適當,兩組一致。
4.2 Could measurement or ascertainment of the outcome have differed between intervention groups?
(每一序列內)結局測量是否可能因介入而異?
- 回答:否(No — 有明確證據支持)
- 原文引述:Participants underwent an oral glucose tolerance test (OGTT) every 3 months... Throughout the study, participants underwent capillary blood glucose monitoring... every month.受試者每 3 個月接受一次 OGTT……整個研究期間每月接受指尖血糖監測。
- 理由:兩組以相同方案(每 3 個月 OGTT、每月血糖監測)等頻率測量與確認結局,結局之測量或確認方式在兩組間無差異。
4.3 Were outcome assessors aware of the intervention received by study participants?
結局評估者是否知道受試者接受的介入?
- 回答:否(No — 有明確證據支持)
- 原文引述:Participant safety and end point adjudication were conducted independently by the DSMB and end point adjudication committee, respectively... A comprehensive blinding protocol was implemented to safeguard treatment allocation concealment... among participants, investigators, and other trial personnel.受試者安全與終點裁決分別由資料安全監測委員會(DSMB)與終點裁決委員會獨立執行……並對受試者、研究者及其他試驗人員實施完整盲性。
- 理由:研究為雙盲、結局由獨立終點裁決委員會在盲性下確認,結局評估者不知所接受之介入。
4.4 Could assessment of the outcome have been influenced by knowledge of intervention received?
結局評估是否可能受介入知情影響?
- 回答:不需評估(NA)
- 原因:條件題:依 4.3 的答案,本題依 RoB 2 官方路由不需評估(NA)。
4.5 Is it likely that assessment of the outcome was influenced by knowledge of intervention received?
結局評估是否很可能受介入知情影響?
- 回答:不需評估(NA)
- 原因:條件題:依 4.4 的答案,本題依 RoB 2 官方路由不需評估(NA)。
Domain 5:選擇性報告 — 判斷:Low
偏誤來源:選擇性報告結果——是否依事前訂定的分析計畫分析、有無從多個結果中挑選有利者報告(交叉另含:是否因偵測到殘留而只報第一期)。
5.1 Were the data that produced this result analysed in accordance with a pre-specified analysis plan that was finalized before unblinded outcome data were available for analysis?
是否依事前訂定(取得未盲化資料前定版)的分析計畫分析?
- 回答:是(Yes — 有明確證據支持)
- 原文引述:Statistical analyses adhered to the prespecified plan (statistical analysis plan in Supplement 2)... The trial protocol and statistical analysis plan for this multicenter, double-blind, placebo-controlled RCT (FOCUS; ChiCTR1900023241) are given in Supplement 2.統計分析依循事先制定之計畫(補充材料 2 之統計分析計畫)……本試驗的研究計畫書與統計分析計畫(FOCUS;ChiCTR1900023241)載於補充材料 2。
- 理由:研究於 ChiCTR1900023241 註冊,研究計畫書與統計分析計畫公開於補充材料 2,且明載分析依事先制定計畫進行,可判定主要分析依事先指定且於解盲前定稿之計畫執行。
5.2 Is the numerical result being assessed likely to have been selected, on the basis of the results, from multiple eligible outcome measurements (e.g. scales, definitions, time points) within the outcome domain?
此數值結果是否可能依結果,從多個量表/定義/時間點中選出?
- 回答:可能否(Probably no — 綜合判斷後傾向「否」)
- 原文引述:The primary outcome was the occurrence of diabetes, which was determined by 2 consecutive oral glucose tolerance tests... The primary end point, incidence of diabetes, was compared between groups using a log-rank test.主要結局為糖尿病的發生,以連續兩次 OGTT 判定……主要終點糖尿病發生率以 log-rank 檢定進行組間比較。
- 理由:主要結局糖尿病發生率為單一、事先明確定義之終點(連續兩次 OGTT),與註冊及計畫書一致,無多種量表/定義/時間點間擇優報告之跡象。
5.3 Is the numerical result being assessed likely to have been selected, on the basis of the results, from multiple eligible analyses of the data?
此數值結果是否可能依結果,從多種分析中選出?
- 回答:可能否(Probably no — 綜合判斷後傾向「否」)
- 原文引述:a Cox proportional hazards regression model incorporating age, sex, and study center as covariates was used to estimate the hazard ratio (HR) and 95% CI between treatment groups... The full analysis set (FAS) for primary and secondary end point analyses comprised all randomized participants.以納入年齡、性別與研究中心為共變項之 Cox 比例風險迴歸模型估計組間風險比(HR)與 95% 信賴區間……主要與次要終點分析以全分析集(FAS)為對象。
- 理由:主要分析方法(log-rank 與事先指定共變項之 Cox 模型、FAS)與計畫一致,並另有敏感度分析佐證穩健,無依結果於多種分析間擇優報告之跡象。