自覺肌力問卷的發展與重症系統性乏力的觀察研究

Abstract

研究背景與目的：重症系統性乏力的診斷必須排除入加護病房前的系統性乏力，本研究旨於發展病人自覺肌力問卷用以排除入住加護病房前的系統性乏力並初探內科加護病房患者轉出加護病房時的重症系統性乏力盛行率。 研究方法：病人自覺肌力問卷的發展分兩階段進行。第一階段，依據重症系統性乏力黃金診斷指標Medical Research Council scale, MRC擬定3-5分肌力強度的問卷題目，並以李克特量表完成專家內容效度。第二階段，採前瞻性世代研究，收取入加護病房超過24小時之成年病患，排除入院前長期臥床者或是入院前長期意識不清或昏迷者，於出加護病房時完成MRC之評估，並於出加護病房後一個月再完成自覺肌力問卷。分析比較病人自覺肌力問卷和黃金指標工具MRC的結果差異，包括皮爾森相關係數、Cohen’s Kappa係數、敏感性、特異度、陽性預測值及陰性預測值等篩檢指標。 研究結果：第一階段發展之病人自覺肌力問卷共18題，題目是對應至MRC六組左右對稱受測部位3-5分肌力強度的肌力動作，總分60分，具良好的專家內容效度。第二階段研究自2018年8月至2019年2月共納入31名內科加護病房患者，評估其轉出加護病房時重症系統性乏力的有無並追蹤其至轉出後一個月。轉出加護病房時，MRC評估為＜48分共有13名，重症系統性乏力盛行率為42%。比較轉出加護病房一個月後所填的自覺轉出時肌力得分和一個月前黃金診斷指標MRC，結果顯示二者得分的皮爾森相關係數r=0.64，結果判定的Kappa一致性係數κ=0.33；以自覺肌力問卷篩檢重症系統性乏力的敏感性為100%，但特異度僅有38%，換算陽性預測值為50%，陰性預測值100%。進一步檢視偽陽性病人之特性，若調整加入Mini-mental Status Examination, MMSE得分<20分做為搭配自覺肌力問卷<48分來篩檢重症系統性乏力的新標準，將獲得敏感性91%，特異度89%，陽性預測值83%，陰性預測值94%。 結論：初步發展的自覺肌力問卷在篩檢重症性統性乏力具有高敏感性，可排除系統性乏力的病人。惟特異度過低，顯示個案或許低估自己在轉出加護病房的肌力表現，又或是問卷肌力表現設定過高，未來研究可進一步提升問卷題目鑑別度亦或可考慮結合MMSE<20分來排除入住加護病房前的重症系統性乏力，以降低偽陽性發生。

Background & Objectives: Manual Medical Research Council (MRC) Scale for Muscle Strength is well-established to screen for intensive care unit - acquired weakness (ICUAW). It is fundamental to rule out pre-ICU weakness in order to diagnose ICUAW in the ICU settings. The aims of this study were to develop a self-reported muscle strength Questionnaire-MRC (Q-MRC) to rule out patients’ pre-ICU weakness and to explore the prevalence of ICUAW using the MRC scale. Method: The development of Q-MRC consists of two phases. At phase one, items were drafted to correspond with the manual MRC scale in the muscle strength levels of 3 to 5. A 7-expert panel rated each item using a 4-point Likert scale. At phase two, a prospective cohort study was conducted at a university hospital’s 6 medical ICUs. Excluding bedridden and patients who were unable to follow command, 31 participants aged≧20 years and stayed ICU over 24 hours were enrolled. The manual MRC was completed at ICU discharge by trained nurses and Q-MRC was completed one-month after ICU discharge using patient interviews. Results of manual MRC and Q-MRC were compared using Pearson correlation, Cohen’s Kappa statistics, sensitivity, specificity, and positive/negative predictive values. Results: The 18-item Q-MRC was developed with good content validity. Items were properly matched to muscle strength level 3-5 among 6 pairs of MRC testing groups. From August 2018 to February 2019, with 31 participants completed manual MRC at ICU discharge. 13 (42%) scored <48 and found to have ICUAW. Comparing Q-MRC with MRC, the Pearson correlation of the scores was 0.64 and Cohen’s kappa of two tools was 0.33 indicating fair agreement. The sensitivity of Q-MRC was 100% and 38% in specificity, 50% in positive predictive value (PPV), and 100% in negative predictive value (NPV). Further analyses revealed that adding the Mini-mental status examination (MMSE) <20 as a must-have criterion enhances the Q-MRC performance as the MMSE<20 plus Q-MRC<48 resulting in 91% in sensitivity, 89% in specificity, 83% in PPV, and 94% in NPV. Conclusion: The self-report Q-MRC is very sensitive to rule out ICUAW, but the false-positive requiring further work. Whether ICU patients tended to under-report their muscle strengths or item scoring of Q-MRC was too strict requiring further study. Nevertheless, adding low MMSE status as an additional criterion enhances the performance of Q-MRC by reducing false-positive of cases.