icu患者の呼吸苦慈恵icu勉強会 2019/08/06 山口庸子本日の内容 •dyspneaの定義...

ICU患者の呼吸苦

慈恵ICU勉強会2019/08/06山口庸子

本日の内容

• Dyspneaの定義• Dyspneaのメカニズム• ICUでのDyspneaの頻度• Dyspneaの評価• Dyspneaの管理

–薬理学的介入–非薬理学的介入

Dyspneaの定義

• 「強度が異なる定性的に異なる感覚からなる呼吸不快感の主観的経験」

“a subjec8ve experience of breathing discomfort that consists of qualita8vely dis8nct sensa8ons that vary in intensity,”

• 1999年 American Thoracic Society（ATS）で合意声明が出された。その後、ワーキングメンバーによって改訂作業が行われ、2011年ATS国際会議で承認、決定された。

Am J Respir Crit Care Med 1999;159:321–340.Am J Respir Crit Care Med 2012;185:435–452.

Dyspneaのメカニズム

Dyspneaのメカニズム

• 神経生理学的メカニズム– Sources of Sensory Afferent Information– Qualities of Dyspnea– Cerebral Processing of Dyspnea

Am J Respir Crit Care Med Vol 185, Iss. 4, pp 435–452, Feb 15, 2012

Sources of Sensory Afferent Information （感覚求心性情報の源）

• 呼吸感覚に関連する感覚求心性情報の源は多数存在する。

• 呼吸運動システムは、自動（脳幹）および自発的（皮質）運動指令の両方から成る。

• これらの異なる情報源からの刺激は、異なる感覚を引き起こす。


呼吸感覚に関連する求心性情報源

for all dyspnea (e.g., see Reference 45), an idea that has beendisproved (46–48). Nonetheless, an uncomfortable sense of re-spiratory “work” and “effort” is commonly reported by patientswith conditions such as asthma, chronic obstructive pulmonarydisease (COPD), and diseases that impair respiratory muscleperformance (27, 31, 32). Respiratory muscle afferents projectto the cerebral cortex, and subjects report sensations localizedto respiratory muscles when the work of breathing is high (49).Perceptions of work and effort probably arise through somecombination of respiratory muscle afferents and perceived cor-tical motor command or corollary discharge (50). Similar mech-anisms give rise to sensations of work and effort from exercisinglimb muscles (reviewed in References 42 and 51).

During exercise, a variety of physiological adaptations matchalveolar ventilation to metabolic demand in healthy, normal per-sons. As the intensity of exercise increases, individuals generallybecome aware that breathing requires more work or effort, muchas they become aware that exercising limb muscles are workingharder. Until the physiological capacity to match ventilation tometabolic demand is approached, afferent mechanoreceptorfeedback signals that breathing is appropriate to the prevailingrespiratory drive, and breathing distress is minimal (52, 53). Aslong as sensations of increased work or effort in breathing andthe achieved ventilation are consistent with expected responsesto exercise, they are not necessarily unpleasant (52, 53), andmay not even be the primary reason for stopping. However,breathing discomfort is much greater in patients with cardiopul-monary disease and frequently limits exercise (52).

Perception of breathing effort or work can be produced in thelaboratory by external resistive or elastic loads (e.g., 37, 54–57),by volitional hyperpnea (e.g., 54, 58, 59), or by weakening therespiratory muscles via changes in operating length, fatigue, orpartial neuromuscular blockade (60–62). With weakened respi-ratory muscles, the requirement for motor command is in-creased, and the perception of inspiratory force, effort, andwork can be substantially magnified (54, 60, 61, 63), even inthe absence of an increase in ventilation. It is likely that simul-taneous information from muscle afferents focuses and calibra-tes sensation from corollary discharge (42, 64). However, thereis no evidence, as yet, that experimental alterations adequatelyreproduce the sensations experienced by patients with cardio-pulmonary disease.Tightness. “Tightness” is commonly experienced during bron-

choconstriction (27, 65–69). Some studies suggest that chest tight-ness is the dominant experience in the early stages of an asthmaattack, but as airway narrowing worsens, patients also report

work/effort and air hunger/unsatisfied inspiration (65, 68). Bron-choconstriction gives rise to both a sense of tightness and addedphysical work of breathing (65, 70); however, blocking pulmo-nary afferents can diminish tightness (71). In laboratory studies,the relationship between FEV1 and dyspnea intensity in patientswith asthma depends on the particular bronchoprovocationagent used, but descriptions of sensory quality are similar (72).In contrast, patients with asthma are more likely to report tight-ness and less likely to report increased work or effort duringmethacholine-evoked bronchoconstriction than when exposedto large external resistive loads (68) or during cardiopulmonaryexercise testing (73). Mechanical ventilation can eliminate thesense of excessive respiratory work but does not diminish tight-ness (74). There is also evidence in patients with asthma thatperception of increased work/effort does not respond as rapidlyas tightness to treatment with nebulized albuterol (67), suggest-ing that work/effort may be more related to increased respiratorymotor output needed to overcome airflow obstruction (e.g., dueto inflammation), whereas tightness may be more specificallyrelated to stimulation of airway receptors. Together, these find-ings suggest that tightness arises from pulmonary afferents ratherthan being a work-related sensation.Air hunger/unsatisfied inspiration. A perception of not getting

enough (or of needingmore) air, which has been variously labeledas air hunger, unsatisfied inspiration, or an unpleasant urge tobreathe, can be induced experimentally by increasing inspiratorydrive (e.g., with exercise, hypercapnia, or hypoxia), especially ifthe capacity to satisfy the increased ventilatory demand is limited.As the demand for ventilation exceeds the capacity to provide it(which occurs only at very high levels of exercise in healthy indi-viduals but is common in patients with cardiopulmonary or neu-romuscular disease), a state of imbalance develops between themotor drive to breathe, as sensed via corollary discharge, and af-ferent feedback frommechanoreceptors of the respiratory system.This becomes increasingly unpleasant and distressing. Variousterms have been used to describe this imbalance, includinglength–tension inappropriateness (75), efferent–reafferent disso-ciation (76), neuroventilatory dissociation (77), afferent mismatch(78), neuromechanical uncoupling (52, 79, 80), or neuromusculardissociation (53), but none fully captures the interplay amongneurophysiological mechanisms.

Wright and Branscomb (81) proposed the term “air hunger”to describe severe respiratory discomfort evoked by strapping thechest and abdomen with broad adhesive tape during exerciseand subsequently by using a device that limited tidal volumeand respiratory rate during hypoxia. Other investigators have

TABLE 2. POSSIBLE AFFERENT SOURCES FOR RESPIRATORY SENSATION*

Source of Sensation Adequate Stimulus

Medullary respiratory corollary discharge Drives to automatic breathing (hypercapnia, hypoxia, exercise)Primary motor cortex corollary discharge Voluntary respiratory driveLimbic motor corollary discharge EmotionsCarotid and aortic bodies Hypercapnia, hypoxemia, acidosisMedullary chemoreceptors HypercapniaSlowly adapting pulmonary stretch receptors Lung inflationRapidly adapting pulmonary stretch receptors Airway collapse, irritant substances, large fast (sudden) lung inflations/deflationsPulmonary C-fibers (J-receptors) Pulmonary vascular congestionAirway C-fibers Irritant substancesUpper airway “flow” receptors Cooling of airway mucosaMuscle spindles in respiratory pump muscles Muscle length change with breathing motionTendon organs in respiratory pump muscles Muscle active force with breathing motionMetaboreceptors in respiratory pump muscles Metabolic activity of respiratory pumpVascular receptors (heart and lung) Distention of vascular structuresTrigeminal skin receptors Facial skin coolingChest wall joint and skin receptors Tidal breathing motion

* Reviewed, for example, in References 24–26 and 39–41.

438 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 185 2012


呼吸感覚に関連する求心性情報源（訳）

感覚の源適切な刺激

髄質呼吸器の随伴反射“corollary discharge” 自動呼吸を駆り立てる（高炭酸ガス血症、低酸素症、運動）

一次運動野の随伴反射“corollary discharge” 自発呼吸を駆り立てる

辺縁系運動の随伴反射“corollary discharge” 感情

頸動脈および大動脈体高炭酸ガス血症、低酸素血症、アシドーシス

延髄化学受容器高炭酸ガス血症

緩徐に適応する肺伸展受容器肺の膨張

急速に適応する肺伸展受容器気道の虚脱、刺激性物質、大きく急速な（突然の）肺の膨張/収縮

肺C繊維（J受容体）肺血管鬱血

気道C繊維刺激性物質

上気道の”流れ（flow）受容体気道粘膜の冷却

呼吸ポンプ筋肉における筋紡錘呼吸運動による筋肉長の変化

呼吸ポンプ筋肉における腱器官呼吸運動を伴う筋活動力

呼吸ポンプ筋肉における代謝受容体呼吸ポンプの代謝活性

血管受容体（心臓と肺）血管構造の膨張

三叉神経受容体顔の皮膚の冷却

胸壁関節と皮膚受容体一回呼吸運動

Am J Respir Crit Care Med Vol 185, Iss. 4, pp 435–452, Feb 15, 2012より

呼吸調節機構

大脳皮質（呼吸困難）

呼吸中枢（延髄）

肺呼吸筋

化学受容器

血液ガスの変化（PaO2,PaCO2）

機械受容器

肺刺激受容器C繊維受容器肺伸展受容器

日本緩和医療学会「呼吸困難のメカニズム」一部改変

Qualities of Dyspnea（呼吸困難の特質）

1) Work/effort2) Tightness3) Air hunger/unsatisfied inspiration



1) Work/effort– 1960年代から1980年代にかけて、呼吸仕事量/努力の感覚がすべての呼吸困難の原因であると広く信じられていた。

– 喘息、慢性閉塞性肺疾患（COPD）、および呼吸筋機能を損なう疾患などの症状を持つ患者は、一般的に不快な「呼吸努力・呼吸仕事量の増大」の感覚を報告している。

– 呼吸筋求心性神経は大脳皮質に投射され、呼吸の仕事量が多いときに呼吸筋に局在する感覚が生じる。

– 運動負荷など代謝欲求に対して、呼吸仕事や努力の増加と、換気のバランスが保たれている場合は、必ずしも不快ではない。


2) Tightness- 「圧迫感」は気管支収縮時によく経験される。- 気道狭窄が悪化するにつれて、患者は仕事/努力および空気飢餓感/不十分な吸気を報告する。- 機械的換気は、過度の呼吸仕事の感覚を取り除くことができるが、圧迫感を減らすことはない。

- 圧迫感が仕事量関連の感覚ではなく肺求心性神経から生じることを示唆している。


3) Air hunger/unsatisfied inspiration.–十分な空気を得られない感覚は、呼吸飢餓感、不満足な吸気、または呼吸に対する不快な衝動など、さまざまな表現がされる。

–換気の需要がそれを提供する能力を超えているために生じる。（健康な人では非常に過度な運動でのみ起こるが、心肺疾患または神経筋疾患の患者では一般的に生じる）。

–その結果は、随伴反射と呼吸器系の機械受容器からの求心性フィードバックを介して感知される。

Cerebral Processing of Dyspnea（呼吸困難の脳での処理）

• Neuroimaging：神経イメージング–近年、ＰＥＴおよびＭＲＩのような三次元脳マッピング技術が、脳血流の変化から神経活性化を推測するために使用されている。

–呼吸困難時は皮質辺縁系構造が活性化される。–恐怖や不安も呼吸困難を引き起こしたり、悪化させる可能性がある。感情と呼吸感覚の相互作用の脳の相関の研究が行われている。

–長期的には、ニューロイメージング研究の結果は、呼吸困難患者に対するより効果的な治療戦略の開発に貢献するかもしれない。


experience dyspnea at times, for example duringexercise, but in certain situations it becomes unbearable;therefore, it is not surprising that one of the mostpopular rating instruments (ie, Medical ResearchCouncil Dyspnea Scale) defines the highest stage ofdyspnea as “too breathless to leave the house or dress orundress.”3 This is particularly relevant in patients whoare terminally ill, whether afflicted by respiratory-,cardiac-, or cancer-related disorders, because of a finalstage of a chronic process, an acute event, or both.

In some acute or chronic situations, dyspnea resultsfrom a combination of events. Therefore, cliniciansshould evaluate the different onset of manifestation(chronic, acute on chronic, or end-of-life symptoms) tomore effectively manage the dyspnea in these patients.In addition, many patients with advanced diseaseexperience episodic dyspnea, which is poorly understoodand often develops without any identifiable trigger.4 Inthis setting, the assessment of terminal dyspnea is amore complex process because the patient’s ability toperceive and report the symptom may be impaired inthe face of declining cognition and alertness. Althoughinterventions to alleviate breathlessness may or may notwork in a particular patient, clinicians should feelobligated to assure these patients that beyond all thetreatments that have been tried unsuccessfully to curethem, there will always be hope for a peaceful, dignified

death. Unfortunately, palliation of dyspnea has receivedrelatively little attention in clinical practice, in medicalliterature, and (when we consider the emphasis given tothe problem of pain) even among the general public.This holds particularly true when the respiratory distressis associated with acute respiratory failure because mostif not all of the studies on the effects of pharmacologictreatments, such as opioids or oxygen, have excludedthese patients. Therefore, it is not surprising that a studyon the quality of the dying experience for patients in anICU found that only 3% of patients were considered byfamily members to breathe comfortably at the end oflife.5 A study on patients who were terminally ill showedthat breathlessness increased significantly at days 3 and10 before death and remained unchanged thereafter;however, breathlessness was significantly higher forpatients with noncancer diagnoses.6 Despite this finding,evidence shows that palliative care teams are seldomconsulted in the terminal care of patients in the ICU orthose with respiratory disorders.7

The aim of this review is to assess the pharmacologicand nonpharmacologic treatment of breathlessness, withspecific emphasis on patients experiencing respiratorydistress or acute respiratory failure at terminal stages oftheir disease. We also identify knowledge gaps andconsider ways of improving the management of dyspneain patients who are terminally ill.

Materials and MethodsSearch CriteriaThis is a narrative review of the literature based on searches of twomain databases, namely PubMed and The Cochrane Database ofSystematic Reviews, using the key words “dyspnea,” “palliativedyspnea,” “breathlessness,” “palliative breathlessness,” and

“refractory breathlessness.” The search strategy using MedicalSubject Headings was limited to human studies and articles inEnglish or in any other language with an English abstract. Thesearch was done on abstract, title, and key words fields. Becausethis was a narrative review, we conducted a qualitative analysiswithout additional assessments.

TABLE 1 ] Subtypes of Dyspnea

Type of Dyspneic Sensation Origin of the Dyspneic Sensation Example of Disease Possible Treatment

Air hunger Chemoreflex activity Pulmonary hyperinflation(COPD, advanced phase ofasthma attack)

CPAP, NIV, bronchodilators

Increased work ofbreathing

Activity of cerebral motorcortex

Muscle weakness(neuromuscular diseases)

NIV, IMV

Chest tightness Stimulation of slowlyadapting receptors in thelarge airways

Bronchospasm (first phaseof asthma attack)

Bronchodilators

Rapid breathing,tachypnea

Stimulation of pulmonaryC fibers

Interstitial lung disease,pulmonary venouscongestion

Opioids, HFNT

HFNT ¼ high-flow nasal therapy; IMV¼ invasive mechanical ventilation; NIV ¼ noninvasive ventilation.

926 Recent Advances in Chest Medicine [ 1 5 4 # 4 CHES T OC TO B E R 2 0 1 8 ]

CHEST 2018; 154(4):925-934

空気飢感

呼吸仕事量の増加

胸の圧迫感

頻呼吸

タイプ由来疾患例治療

化学反射の活動

大脳皮質運動野の活動

大きな気道において、ゆっくり適応する受容体の刺激

肺のC繊維の刺激

肺の過膨張（COPD,喘息発作の進行期）

筋力低下（神経筋疾患）

気管支痙攣（喘息発作の第一段階）

間質性疾患、肺静脈鬱血

• 呼吸困難の症状は、様々な呼吸器および非呼吸器疾患に由来するサブタイプを有する。

気管支拡張剤

気管支拡張剤

Dyspneaのメカニズムのまとめ

• 呼吸困難を感知する受容器（機械受容器、化学受容器、呼吸中枢）は多数ある。

• 呼吸困難感は主に3つの特徴がある。

–呼吸仕事量の増大、圧迫感、空気飢餓感• 呼吸困難感の表出には、精神的要因、社会的文化的状況も影響する。

ICUでのDyspneaの頻度

• 背景– 香港では緩和ケアの利用のほとんどが癌患者に限られている。非癌患者に対する緩和ケアは始まったばかりである。

• 目的– 癌と非癌による死亡を比較すること

• デザイン– チャートからの後ろ向き研究

• 対象– 2006/1/1〜2006/12/31に、国内の4つの公立病院で、癌および非癌性疾患（CRF、慢性閉塞性肺疾患（COPD）、鬱血性心不全（CHF）で死亡した患者。

– 3つの病院は600、1000、1350床の急性総合病院。1つは425床の回復期病院。

Comparing Noncancer and Cancer Deathsin Hong Kong: A Retrospective Review

Journal of Pain and Symptom Management Vol. 40 No. 5 November 2010

in COPD and CHF patients, the ACP discussionwas first documented within three days beforedeath in 26.9%and35.7%, respectively, whereasthat for CRF and cancer were 9.5% and 11.6%,respectively (P< 0.001).

CPR was discussed in 95% of all patients. Thediscussants involved in the noncancer and the

cancer patients included patient (11.5% vs.40.1%, P< 0.001); spouse (19.3% vs. 27.9%,P¼ 0.013); children (65.1% vs. 62.2%, nosignificant difference statistically [NS]); parents(0.5% vs. 0.6%, NS); and friends, distantrelatives, and guardians (15.7% vs. 8.7%,P¼ 0.022).

Table 4Symptoms Documented by Physician in the Last Two Weeks of Life

SymptomsCRF

(n¼ 239)COPD

(n¼ 242)CHF

(n¼ 175)Noncancer(n¼ 656)

Cancer(n¼ 183) P-valuea

No. of patients (percent within group)Dyspnea 127 (53.1) 208 (86.0) 144 (82.3) 479 (73.0) 107 (58.5) <0.001Edema 92 (38.5) 55 (22.7) 87 (49.7) 234 (35.7) 59 (32.2) NSPain 82 (34.3) 66 (27.3) 50 (28.6) 198 (30.2) 118 (64.5) <0.001Cough 43 (18.0) 119 (49.2) 32 (18.3) 194 (29.6) 43 (23.5) NSFatigue 68 (28.5) 63 (26.0) 33 (18.9) 164 (25.0) 117 (63.9) <0.001Anorexia 70 (29.3) 39 (16.1) 24 (13.7) 133 (20.3) 88 (48.1) <0.001Bowel problem 51 (21.3) 35 (14.5) 21 (12.0) 107 (16.3) 59 (32.2) <0.001Confusion 35 (14.6) 45 (18.6) 17 (9.7) 97 (14.8) 56 (30.6) <0.001Nausea/vomiting 45 (18.8) 13 (5.4) 11 (6.3) 69 (10.5) 51 (27.9) <0.001Urinary problem 25 (10.5) 19 (7.9) 15 (8.6) 59 (9.0) 31 (16.9) 0.004Sores/wounds 24 (10.0) 10 (4.1) 23 (13.1) 57 (8.7) 24 (13.1) NSRestlessness 16 (6.7) 6 (2.5) 9 (5.1) 31 (4.7) 18 (9.8) 0.013Oral problem 6 (2.5) 20 (8.3) 5 (2.9) 31 (4.7) 23 (12.6) <0.001Distended abdomen 19 (7.9) 7 (2.9) 4 (2.3) 30 (4.6) 37 (20.2) <0.001Cachexia 4 (1.7) 13 (5.4) 3 (1.7) 20 (3.0) 38 (20.8) <0.001Sleep problem 8 (3.3) 5 (2.1) 3 (1.7) 16 (2.4) 30 (16.4) <0.001Hemoptysis 2 (0.8) 10 (4.1) 2 (1.1) 14 (2.1) 8 (4.4) NSDepressed mood 4 (1.7) 6 (2.5) 4 (2.3) 14 (2.1) 15 (8.2) <0.001Pruritus 5 (2.1) 1 (0.4) 0 (0.0) 6 (0.9) 2 (1.1) NSWish to hasten death 0 (0.0) 4 (1.7) 1 (0.6) 5 (0.8) 5 (2.7) 0.045Anxiety 0 (0.0) 3 (1.2) 1 (0.6) 4 (0.6) 8 (4.4) <0.001Suicidal ideation 0 (0.0) 2 (0.8) 0 (0.0) 2 (0.3) 5 (2.7) 0.007Request euthanasia 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 3 (1.6) 0.010Mean no. 3.1 (1.9) 3.2 (1.6) 2.9 (1.5) 3.1 (1.7) 5.2 (2.6) <0.001aP-value refers to level of significance in comparing noncancer patients (n¼ 656) and cancer patients (n¼ 183).

Fig. 1. Analgesics and sedatives prescribed in the last two weeks of life. P-value refers to level of significance in com-paring noncancer patients (n¼ 656) and cancer patients (n¼ 183). NSAID¼ nonsteroidal anti-inflammatory drug.Sedatives include benzodiapzepines and haloperidol.

Vol. 40 No. 5 November 2010 709Comparing Noncancer and Cancer Deaths in Hong Kong

• 癌患者は、「痛み」「疲労」「呼吸困難」の順であったが、非癌患者では、「呼吸困難」がも多い症状であった。

• 基礎疾患に関係なく、呼吸困難、浮腫、疼痛、および疲労がも一般的に報告されている症状であった。

Symptoms experienced by intensive care unit patients at high risk of dying

Crit Care Med. 2010 November ; 38(11): 2155–2160.• 目的

– 死亡のリスクの高いICU患者が経験している症状を詳細にアセスメントすること– せん妄と患者の症状との関係を評価すること

• デザイン

– 前向き観察研究• 対象

– アメリカ西部の三次医療センターにある2つのICU– 18歳以上の死亡リスクの高いICU患者– 以下の患者は死亡リスクが高いとみなされた。

Ø APACHEⅡスコア20以上Ø ICU在室期間3日以上Ø急性心不全および/または急性呼吸不全Ø肝硬変を伴う慢性肝不全

Ø多臓器不全と敗血症

Ø悪性腫瘍の診断に関連する臓器不全

• 方法– 1日おきに14日間まで面接を実施し、10症状の有無と、症状がある場合その強度を聞いた。

– Pain, tired, short of breath, restless, anxious, sad, hungry,scared, thirsty, confused

– Intensity: 1=mild,2=moderate,3=severe– Distress: 1= not very distressing, 2= moderately distressing ,

3=very distressing• 結果

– 1273人がスクリーニングされ、245人が登録された。– 登録患者のうち、171人（69.8%）が少なくとも１つの症状があった。– 登録した245人のうち、32%が病院で死亡。終転帰として22%が死亡。– 34%が人工呼吸管理をされていた– 人工呼吸管理されていた患者は、されていない患者に比べ、症状を報告できた患者が有意に少なかった。（P=0.005）

– 症状を報告できた人とできなかった人の間で、APACHEⅡスコアに有意差はなかった。

Table 1 Prevalence of symptoms across 405 assessments from 171 pa@ents N

IH-P

A A

uthor Manuscript

NIH

-PA

Author M

anuscriptN

IH-P

A A

uthor Manuscript

Puntillo et al. Page 12

Table 1Prevalence of symptoms across 405 assessments from 171 patients

Symptom Percent SE 95% Confidence Interval

Tired 74.7 2.79 69.2–80.2

Thirsty 70.8 3.13 64.6–76.9

Anxious 57.9 3.17 51.7–64.1

Restless 49.0 3.09 42.9–55.0

Hungry 44.8 3.32 38.3–51.3

Short of breath 43.9 3.37 37.3–50.5

Pain 40.4 3.16 34.2–46.6

Sad 33.9 3.00 28.0–39.7

Scared 32.8 3.42 26.1–39.5

Confused 26.6 2.87 21.0–32.3

Note: SEs and 95% confidence intervals used the robust variance estimate from the generalized estimating equation approach.

Crit Care Med. Author manuscript; available in PMC 2012 June 18.

• 約44%の人が、「short of breath」の症状を訴えた。（第６位）

Table 2 Mean intensity ratings (mild = 1,

moderate = 2, severe = 3) for patients who reported symptom to be present N

IH-P

A A

uthor Manuscript

NIH

-PA

Author M

anuscriptN

IH-P

A A

uthor Manuscript


Table 2Mean intensity ratings (mild = 1, moderate = 2, severe = 3) for patients who reportedsymptom to be present

Symptom Intensity Mean SE 95% Confidence Intervals

Tired 1.81 .062 1.69–1.93

Thirsty 2.16 .087 1.99–2.33

Anxious 1.92 .076 1.77–2.06

Restless 1.76 .079 1.61–1.91

Hungry 1.89 .091 1.71–2.07

Short of breath 1.89 .083 1.73–2.05

Pain 1.74 .073 1.60–1.88

Sad 1.85 .107 1.64–2.06

Scared 1.80 .118 1.56–2.03

Confused 1.73 .132 1.47–1.96



NIH

-PA

Author M

anuscriptN

IH-P

A A

uthor Manuscript

NIH

-PA

Author M

anuscript


Table 3Mean distress ratings (mild = 1, moderate = 2, severe = 3) for patients who reportedsymptom to be present

Symptom Distress Mean SE 95% Confidence Intervals

Tired 1.94 .058 1.83–2.06

Thirsty 1.90 .080 1.74–2.06

Anxious 1.93 .076 1.79–1.79

Restless 1.95 .088 1.78–2.12

Hungry 1.62 .087 1.45–1.79

Short of breath 2.34 .088 2.17–2.52

Pain 2.08 .081 1.92–2.24

Sad 1.97 .114 1.74–2.19

Scared 2.15 .102 1.95–2.35

Confused 2.10 .152 1.80–2.40



Table 3 Mean distress ratings (mild = 1,

moderate = 2, severe = 3) for patients who reported symptom to be present

• 「short of breath」の症状は、強度では上位３番目であったが、苦痛の評価では、痛み以上にも苦痛を感じていた。

• 人工呼吸器管理をされている人といない人での有意差はなかった。

Dyspnea in mechanically ventilated critically ill patients Crit Care Med 2011 Vol. 39, No. 9

• 背景– 近のICU管理の傾向は、患者が呼吸困難を抱く危険性をはらんでいる。

ü重度の肺機能障害患者における管理として、浅い鎮静、より少ない一回換気量、自発換気の維持

– 否定的な思い出を通して悪影響を及ぼす可能性がある。• 目的

– 人工呼吸器装着患者における呼吸困難の有病率を評価すること。– その臨床的相関を同定し、そして臨床転帰に対するその影響を調べること。

• デザイン– 前向き観察研究

• 対象– Raymond Poincare University Hospital（360床）のICU（16床）またはla Pi8e -Salpe trie` re hospital （1800床）のICUと呼吸器病棟の内科ICU（10床）

– 24時間以上の挿管管理を受けた成人患者– 5つの簡単な指示に従えた場合に登録された。

ü「目を開く／閉じる」、「私を見る」、「あなたの口を開いて舌を出す」、「うなずいて」、「私が５つまで数えたら眉を上げなさい」

– コミュニケーションが困難である場合、以前に精神障害または認知障害があった場合は除外。

• 方法

– 毎日午前9時から午前12時の間にスクリーニングを実施。– ICU滞在中に一度だけ対象となった。

• 呼吸困難・不安・痛みの評価– 「呼吸困難がありますか」→「はい」の場合はVASで評価。– 呼吸感覚を特徴付けるために、「空気飢餓感“air

hunger”」「過度の呼吸努力 “excessive respiratory effort”」またはその両方、を選択。

– 不安と痛みをVASで評価。

• 人工呼吸器設定について– 呼吸困難があり、医師が人工呼吸器の設定を調整することを選択した場合ü補助換気（ACV）モードを受けている患者は、一回換気量、呼気流量、呼気終末陽圧の増加をテストした。

ü圧補助人工呼吸（PSV）を受けている患者は、初に圧補助を増やすことをテストし、次に吸気トリガーの感度を上げることをテストした。

– 効果の判定ü介入後のVAS評価が介入前のレベルより少なくとも1 cm低い場合、人工呼吸器の設定が呼吸困難の病因に関与しているとした。

METHODS

Patients. The study population includedconsecutive mechanically ventilated adultmedical patients admitted either to the ICU(16 beds) of Raymond Poincare UniversityHospital (360 beds) or to the medical ICU (tenbeds) of the department of respiratory and ICUat la Pitie-Salpetriere hospital (1800 beds)over 6 months. The protocol was approved bythe legal and ethical body recognized byFrench law (Comite de Protection des Per-sonnes, Saint Germain en Laye, France),which waived written informed consent.

Patients were eligible to participate in thestudy if they had been mechanically ventilatedfor !24 hrs—noninvasive ventilation excluded—awake, according to the validated AdaptationTo the Intensive Care Environment (ATICE)scale (17), and able obey to five simple com-mands (“open/close your eyes,” “look at me,”“open your mouth and put out your tongue,”“nod your head,” and “raise your eyebrowswhen I have counted up to five”). They wereincluded as soon as they fulfilled these condi-tions. Patients were not included in the studywhen communication was likely to be difficult(auditory or visual impairment, insufficientcommand of French) or when they wereknown to have prior psychiatric or cognitivedisorders.

Study Procedure. During the study period,the patients at the two ICU sites were system-atically screened for inclusion in the study ona daily basis, between 9 and 12 AM. Potentialcandidates were studied and enrolled as soonas they were sufficiently alert to answer thepreviously described five commands. Theywere only studied once during their ICU stay.

Assessment of Dyspnea, Anxiety, and Pain.Once enrolled, the patients were first asked“do you have trouble breathing”? If the answerwas yes, then they were asked to rate theintensity by placing a cursor on a 10-cm visualanalog scale (VAS) bounded on the left by “norespiratory discomfort” and on the right by“intolerable respiratory discomfort.” The pa-tients were then asked to choose “air hunger,”“excessive respiratory effort,” or both, to char-acterize their respiratory sensations. Finally,they were presented with two additional10-cm VASs to evaluate anxiety (“no anxiety”to “intolerable anxiety”) and pain (“no pain” to“intolerable pain”). When a patient under-stood the principle of assessment but was un-able to move the VAS cursor him- or herself,the observer helped the patient by holding thescale and supporting the patient’s forearm. Ifthe subject was unable to move their arms(like in few cases with severe neuromuscularimpairment), the observers were allowed tomanipulate the VAS cursor following instruc-tions given by the patient. However, this wasnot recommended and avoided whenever pos-sible. The cursor was never adjusted directlyor solely by the investigator.

Descriptive Variables. To describe thestudy population, we recorded age, gender, theseverity of illness—Simplified Acute Physiol-ogy score II—the indication for MV, the na-ture of the patient–ventilator interface (oro-tracheal intubation or tracheotomy), and theduration of MV at the time of dyspnea assess-ment. To describe the patients at the time ofassessment, we recorded heart rate, systolicand diastolic blood pressure, respiratory rate,and temperature. A detailed ”respiratory-oriented“ physical examination was per-formed, systematically noting the presence orabsence of auscultatory abnormalities, thepresence or absence of inspiratory neck mus-cle activity, an inspiratory retraction of thesupraclavicular fossae or intercostal spaces, anabdominal paradox, or a phasic activation ofupper airway dilators (genioglossus and alaenasi). These signs were grouped into a singlevariable, termed ”abnormal breathing dynam-ics“ (coded 0 when none of the signs werepresent and 1 when any of these were ob-served). Blood hemoglobin and arterial bloodgases (PaO2, PaCO2) were recorded as well aschest radiograph abnormalities. Finally, wenoted the type of tracheal tube and ventilatorsettings, including the ventilatory mode, min-ute ventilation, pressure support level, in-spiratory flow, and positive end-expiratorypressure. The duration of MV and the occur-rence of successful extubation within the 3days after the assessment of dyspnea weregathered as well as the mortality rate and theICU length of stay.

Role of Ventilator Settings. When a patientreported being dyspneic, the investigator im-mediately informed the physician in charge. Ifthe physician chose to adjust the ventilatorsettings, he was asked to do so in a standard-ized and stepwise manner, including a re-evaluation of the patient’s perceptions aftereach change. No further setting alterationswere made after a given change had beenassociated with a decrease in dyspnea. In pa-tients receiving assist control ventilation(ACV) mode, increases in tidal volume, in-spiratory flow, and positive end-expiratorypressure were subsequently tested. In patientson pressure support ventilation, increasingpressure support was first tested and then in-creasing the sensitivity of the inspiratory trig-ger. Arbitrarily, we considered that ventilatorsettings were involved in the pathogenesis ofdyspnea when the postintervention VAS ratingwas lower than the preintervention level by atleast 1 cm. The patients who fulfilled thesecriteria/responded to ventilator changes andnondyspneic patients were pooled into a com-mon group that was subsequently comparedwith the group of patients whose dyspneafailed to recede in response to ventilator-setting adjustments.

Statistical Analysis. Nonparametric datawere presented as median and interquartilerange.

The main outcome was dyspnea, catego-rized as “present” or “absent” and therefore

defining two groups of patients. They werecompared using Fisher’s exact test for quali-tative variables and the Wilcoxon rank sumtests for quantitative variables. A multiple lo-gistic regression with a backward stepwisemodel selection was used to identify the fac-tors independently associated with dyspnea.All variables marginally associated with themain outcome at the 0.20 level were consid-ered and sequentially removed from the modelat the 0.05 level. Because dyspnea and anxietyare closely related, the logistic regression wasrepeated after exclusion of anxiety from themodel. Paired Wilcoxon tests were performedto compare dyspnea, anxiety, and pain VASbefore and after adjusting ventilator settingsin the patients reporting dyspnea. Gray’s test(18), with death on MV as a competing event,was used to compare median durations of MVbetween the two groups. All tests were two-sided, and p values " .05 were regarded asindicating statistical significance. The statisti-cal analyses were performed using R 2.6.2 sta-tistical package (The Foundation for Statisti-cal Computing, Vienna, Austria).

RESULTS

Study Population. During the studyperiod, 557 patients were referred to thetwo participating centers. One hundredeighty-eight were mechanically ventilatedfor !24 hrs (Fig. 1). Of those, 21 failed tobe screened for various reasons, and 71were excluded. Among those excluded, 39patients were considered unable to accu-rately grade their dyspnea because of psy-chiatric, acute or chronic cognitive disor-ders or other reasons (insufficientcommand of French, visual or auditive im-pairment). Thirty-two other patients diedbefore regaining consciousness.

Ninety-six patients (age 61 # 18 yrs,60% male) were included for final analy-sis (Table 1). The median duration of MVat the time of the study was 3 days (in-terquartile range, 1–6). The two mainindications for MV were hypoxemic acute

Figure 1. Flow chart of the study. ICU, intensivecare unit.

2060 Crit Care Med 2011 Vol. 39, No. 9

respiratory failure (mainly resulting frompneumonia) and the decompensation ofan underlying neuromuscular disease(Table 1). Table 1 provides the variablesdescribing the patients at the time of thestudy. A tracheotomy had been per-formed before inclusion in 17 patients(18%).

Prevalence and Risk Factors for Dys-pnea. Forty-five patients (47%) reporteddyspnea. The median VAS dyspnea ratingwas 5 (interquartile range, 4–7). Therewere no significant differences betweenpatients with and without dyspnea interms of demographic variables, severityof critical illness, or indication for me-chanical ventilation (Table 1). Body tem-perature, heart rate, and blood pressurewere also similar in both groups. Anxietyand pain were more frequently reportedby dyspneic patients than by nondyspneicones (Table 2).

Dyspneic patients did not exhibit morefrequent abnormalities in breathing dy-namics or more frequent ausculatory ab-normalities than nondyspneic patients(Table 2). This was also the case for chestradiograph abnormalities. There was noassociation between dyspnea and hemo-globin, the PaO2/FIO2 ratio, or PaCO2 (Ta-ble 2). Dyspneic patients were more fre-quently mechanically ventilated with theACV mode than nondyspneic patients(Table 2), but there was no other statis-tically significant difference regardingventilatory settings (Table 2).

Multivariate analysis showed that dys-pnea was independently associated withanxiety (odd ratio [OR], 8.84; 95% confi-dence interval [CI], 3.26 –24.0; p !

.0001), the ACV mode (OR, 4.77; 95% CI,1.60–14.3; p " .005), and heart rate (OR,1.33 per 10 beats/min; 95% CI, 1.02–1.75;p " .038). Removing anxiety from themodel did not notably change the results(assist-control: OR, 3.92, 95% CI, 1.53–10.1, p " .005; heart rate: OR, 1.32, 95%CI, 1.05–1.68, p " .019).

Dyspneic Modalities. Among the 45dyspneic patients, 15 chose only “air hun-ger” to characterize their respiratory dis-comfort, seven chose “excessive respira-tory efforts,” and six chose both thesedescriptors. Seven patients reported hav-ing breathing difficulties but did notchoose between these two descriptors.This was not associated with any cleardifferences between patients, althoughfive of the seven patients who chose “ex-cessive respiratory efforts” only were un-der ACV with an inspiratory flow ratebelow 60 L/min#1.

Role of Ventilator Settings and Poten-tial Causes for Dyspnea. Adjustments inventilator settings were performed by thephysicians in charge of the patients in allthose who reported dyspnea. Thirty-fivepercent of these patients met the “1-cmdecrease in dyspnea rating” criterion thatwas required to consider dyspnea respon-sive to adjustments in ventilator settings.The median reduction in dyspnea ratingwas #4.6 cm (95% CI, #6.1 to #3.2; p ".0005; Table 3). This dyspnea reductionwas accompanied by a significant de-crease in the VAS rating of anxiety (#1.7[#3.3 to #0.2]; p " .041) but not in theVAS rating of pain ($0.3 [#0.7 to $1.2];p " .79). In patients on ACV, increasingtidal volume and/or inspiratory flow was

successful in partially alleviating dyspneain ten cases (22% of dyspneic patients)and increasing positive end-expiratorypressure was successful in one case (2.2%patients). In patients on pressure supportventilation, increasing pressure supportwas successful in three cases (7%) andincreasing the sensitivity of the inspira-tory trigger was successful in one (2.2%).

In the 29 dyspneic patients in whomventilator resetting was not successful,dyspnea was ascribed to a specific causein only nine (31%) patients, includingpartial endotracheal tube obstruction(n " 3), atelectasis (n " 2), major pleuraleffusion (n " 2), severe acute anemia(n " 1), and pneumothorax (n " 1).

Clinical Outcomes. The medianlength of ICU stay was greater in thepatients reporting dyspnea (8 [4–20] vs.14 [8–28] days, p " .017). The medianduration of MV from inclusion to extuba-tion, the rate of successful extubationwithin the 3 days after inclusion, andmortality rate were all comparable be-tween patients with and without dyspnea.

The occurrence of successful extuba-tion within the 3 days after dyspnea as-sessment was significantly less frequentin the patients whose dyspnea failed torecede in response to ventilator settingadjustments than in the other patients(nondyspneic ones and those with a “re-sponsive” dyspnea). The duration of MVwas not statistically different betweenthese two groups, whereas the length ofstay in the ICU tended to be longer whendyspnea did not respond to ventilator ad-justment (Table 4). In this category ofpatients, neuromuscular disease tendedto be more frequent and chronic obstruc-tive pulmonary disease less frequent. Age,Simplified Acute Physiology score II atadmission, the duration of MV before in-clusion, the intensity of dyspnea at base-line, PaO2/FIO2 ratio, and ventilator modedid not differ between the two groups(Table 4).

DISCUSSION

Among the findings of this study, weidentify six salient elements regardingdyspnea in MV: its high prevalence; itsintensity; its association with anxiety; itsrelationships with ventilator settings inone-third of cases; and its associationwith delayed extubation.

Frequency of Dyspnea. Approximatelyhalf of our patients indicated feeling dys-pneic while mechanically ventilated (Ta-bles 1 and 2), a frequency about twice

Table 1. Characteristics of the study population

VariableWhole Cohort

(n " 96)Dyspnea(n " 45)

No Dyspnea(n " 51) p

Male gender, no. (%) 60 (62) 29 (64) 31 (61) .83Median age (IQR), years 64 (48–73) 57 (43–72) 68 (52–73) .24Median Simplified Acute Physiology

Score II at admission (IQR)43 (31–60) 43 (29–60) 43 (33–56) %.99

Indication for mechanical ventilation .57Hypoxemic acute respiratory failure,

no. (%)46 (48) 20 (45) 26 (51)

Decompensation of an underlyingneuromuscular disease, no. (%)

29 (30) 17 (38) 12 (24)

Coma, no. (%) 10 (10) 4 (9) 6 (12)Chronic obstructive pulmonary

disease, no. (%)7 (7) 2 (4) 5 (9)

Others, no. (%) 4 (4) 2 (4) 2 (4)Tracheotomy, no. (%) 17 (18) 7 (16) 10 (20) .79Median time from onset of mechanical

ventilation (IQR), days3 (1–6) 3 (1–6) 3 (1–6) .23

IQR, interquartile range.

2061Crit Care Med 2011 Vol. 39, No. 9

人工呼吸器装着期間の中央値：3日

気管切開施行：17人（18%）

人工呼吸器装着の理由：

（１）低酸素血症（２）神経筋疾患の代償不全

約半数(47%)が呼吸困難を訴えた呼吸困難VASの中央値 5（四分位範囲4〜7）

呼吸困難があった群と無かった群の2群間の有意差はなかった

that of pain. The physical signs collected,chest radiograph abnormalities, bloodgases, and the hemoglobin level were notindependently associated with dyspnea.This indicates that, in this setting, dys-pnea should be actively sought, yet incontrast to pain that is a major determi-nant of ICU-related posttraumatic mani-festations (19), breathing difficulties arenot listed among ICU stressors (20, 21).However, in one post-ICU recollectionstudy, 22% of 75 patients remembered“not getting enough air from their endo-tracheal tube,” something that botheredthem “moderately or extremely” in 92%of cases (19). Recent data suggest thatnegative respiratory-related experiencescan play an important role in the patho-genesis of ICU-related posttraumatic syn-dromes (22, 23). Of note, the prevalenceof dyspnea in our patients might have

been either overestimated because manyof our patients had underlying respira-tory disorders or underestimated becausewe did not perform repeated assessments.

Intensity and ‘Quality’ of Dyspnea.With VAS ratings of !4 in 70% of cases,dyspnea appeared to qualify as “moderateto intense.” Similar pain ratings are anindication for analgesia (24). Only twoverbal descriptors for their respiratorydiscomfort were proposed to our patientsand we did not perform sophisticated psy-chophysical assessments. In particular,we did not attempt to separate the sen-sory and affective components of dyspnea(25). However, we note that they oftenchose “air hunger” to describe their re-spiratory sensation, which has beenshown to be more unpleasant than thesense of excessive inspiratory effort for agiven sensory intensity during experi-mental dyspnea (25). The very strongemotional impact of the ICU environ-ment is a likely source of major dissoci-ation between sensory and emotionalcomponents (26). This could have impor-tant implications in terms of care becausesensory-affective discrepancies could ex-plain either large effects of small inter-ventions or conversely the failure of ap-parently adequate interventions to beefficient.

Dyspnea and Anxiety. Dyspneastrongly correlated with anxiety, whichstresses the clinical relevance of the is-sue. Anxiety and dyspnea have never been

concomitantly assessed in mechanicallyventilated patients, previous studies hav-ing concentrated on the relationships be-tween anxiety and other sources of MV-related discomfort such as extubation orsuctioning (1, 16). The interplay betweenanxiety and dyspnea is complex and caus-ative relationships can exist in both di-rections. Anxiety and fear stimulate ven-tilation and can thus produce dyspnea (7,27). ICU and particularly MV are fearfulexperiences associated with diffuse anxi-ety (8) that can cause posttraumaticstress disorder (28, 29). In this view,some of the respiratory discomfort de-scribed by our patients could represent anonspecific manifestation of anxiety.Anxiety-relieving interventions couldthus have a positive effect on respiratoryrhythm, as reported with music therapy(30). Reciprocally, dyspnea generatesanxiety (1) and relieving dyspnea de-creases anxiety (31). In line with this,anxiety significantly decreased in those ofour patients who reported improvementsafter adjustments of the ventilator set-tings (Table 3). Therefore, relieving dys-pnea is likely to have positive effects onanxiety. Like with anxiety, pain was morefrequently experienced in dyspneic pa-tients than in nondyspneic ones (Table2). Pain also stimulates ventilation (32),and its control could possibly have bene-ficial effects on dyspnea. Because of thisinterrelationship, we believe that anxiety,dyspnea, and pain should be assessed sys-

Table 2. Characteristics of the patients at the time of dyspnea assessment


(n ! 96)Dyspnea(n ! 45)

No Dyspnea(n ! 51) p

Anxiety, no. (%) 44 (46) 32 (71) 12 (24) ".0001If yes, median anxiety visual analog scale (IQR) 5 (4–7) 5 (4–7) 5 (4–6) .64Pain, no. (%) 28 (29) 19 (42) 9 (18) .013If yes, median pain visual analog scale (IQR) 5 (4–6) 5 (4–6) 5 (4–7) .62Median heart rate (IQR), beats/min 93 (81–107) 98 (82–107) 87 (76–106) .15Median respiratory rate (IQR), cycles/min 20 (17–25) 20 (18–28) 19 (17–24) .14Median temperature (IQR), °C 37.3 (36.9–37.7) 37.3 (37.0–37.7) 37.2 (36.9–38.0) .86Median systolic blood pressure (IQR), mm Hg 122 (110–137) 127 (110–140) 119 (108–134) .15Median diastolic blood pressure (IQR), mm Hg 71 (59–83) 73 (60–88) 69 (58–76) .059Abnormalities in breathing dynamics, no. (%) 22 (23) 11 (24) 11 (22) .81Normal auscultation, no. (%) 35 (36) 16 (36) 19 (37) #.99Median hemoglobin (IQR), g/dL 10 (9–12) 10 (9–13) 10 (9–12) .92Median PaO2/FIO2 ratio (IQR) 236 (176–318) 228 (152–318) 241 (191–317) .42Median PaCO2 (IQR), mm Hg 41 (35–55) 42 (38–55) 40 (34–50) .28No radiograph abnormalities, no. (%) 25 (27) 13 (30) 12 (24) .49ACV, no. (%) 54 (56) 31 (69) 23 (45) .024Ventilator settings, no. (%) .094Pressure support ventilation; pressure support "15 cm H2O 19 (20) 6 (13) 13 (25)Pressure support ventilation; pressure support !15 cm H2O 23 (24) 8 (18) 15 (29)ACV; inspiratory flow "60 L/min 23 (24) 15 (33) 8 (16)ACV; inspiratory flow !60 L/min 31 (32) 16 (36) 15 (29)Median minute ventilation, L/min (IQR) 9 (8–12) 10 (8–12) 9 (7–13) .87

IQR, interquartile range; ACV, assist-control ventilation.

Table 3. Effects of adjusting ventilator settings inthe patients reporting dyspnea

ParameterMean Variation, cm

(95% Confidence Interval) p

Dyspnea VAS $4.6 ($6.1 to $3.2) .0005Anxiety VAS $1.7 ($3.3 to $0.2) .041Pain VAS %0.3 ($0.7 to %1.2) .79

VAS, visual analogic scale.Ventilator settings were considered involved

in the pathogenesis of dyspnea if and when thepostintervention VAS rating was inferior by atleast 1 cm to the preintervention one.


• バイタルサイン、血液ガス、ヘモグロビンの値、両群間の差はなかった。• 呼吸困難の患者は、非呼吸困難の患者よりもACVモードで換気されることが多かった。換気設定に関しては、統計的有意差はなかった。

• 不安と疼痛は、呼吸困難「無し」よりも「有り」の患者の方が、有意に多かった。• 多変量解析では、呼吸困難は不安と独立して関連していた。（odd ratio [OR],

8.84; 95% confidence interval [CI], 3.26 –24.0; p <0.0001）

• 呼吸困難の様相（45人の呼吸困難患者）– 「空気飢餓感“air hunger” 」:15人– 「過度の呼吸努力“excessive respiratory efforts”」7人–両方：6人–回答無し：7人

＊「過剰な呼吸努力」を選択した7人中5人はACV下で吸気流速が60 L / min未満であったが、これは患者間の明らかな差には関連していなかった。

that of pain. The physical signs collected,chest radiograph abnormalities, bloodgases, and the hemoglobin level were notindependently associated with dyspnea.This indicates that, in this setting, dys-pnea should be actively sought, yet incontrast to pain that is a major determi-nant of ICU-related posttraumatic mani-festations (19), breathing difficulties arenot listed among ICU stressors (20, 21).However, in one post-ICU recollectionstudy, 22% of 75 patients remembered“not getting enough air from their endo-tracheal tube,” something that botheredthem “moderately or extremely” in 92%of cases (19). Recent data suggest thatnegative respiratory-related experiencescan play an important role in the patho-genesis of ICU-related posttraumatic syn-dromes (22, 23). Of note, the prevalenceof dyspnea in our patients might have

been either overestimated because manyof our patients had underlying respira-tory disorders or underestimated becausewe did not perform repeated assessments.

Intensity and ‘Quality’ of Dyspnea.With VAS ratings of !4 in 70% of cases,dyspnea appeared to qualify as “moderateto intense.” Similar pain ratings are anindication for analgesia (24). Only twoverbal descriptors for their respiratorydiscomfort were proposed to our patientsand we did not perform sophisticated psy-chophysical assessments. In particular,we did not attempt to separate the sen-sory and affective components of dyspnea(25). However, we note that they oftenchose “air hunger” to describe their re-spiratory sensation, which has beenshown to be more unpleasant than thesense of excessive inspiratory effort for agiven sensory intensity during experi-mental dyspnea (25). The very strongemotional impact of the ICU environ-ment is a likely source of major dissoci-ation between sensory and emotionalcomponents (26). This could have impor-tant implications in terms of care becausesensory-affective discrepancies could ex-plain either large effects of small inter-ventions or conversely the failure of ap-parently adequate interventions to beefficient.

Dyspnea and Anxiety. Dyspneastrongly correlated with anxiety, whichstresses the clinical relevance of the is-sue. Anxiety and dyspnea have never been

concomitantly assessed in mechanicallyventilated patients, previous studies hav-ing concentrated on the relationships be-tween anxiety and other sources of MV-related discomfort such as extubation orsuctioning (1, 16). The interplay betweenanxiety and dyspnea is complex and caus-ative relationships can exist in both di-rections. Anxiety and fear stimulate ven-tilation and can thus produce dyspnea (7,27). ICU and particularly MV are fearfulexperiences associated with diffuse anxi-ety (8) that can cause posttraumaticstress disorder (28, 29). In this view,some of the respiratory discomfort de-scribed by our patients could represent anonspecific manifestation of anxiety.Anxiety-relieving interventions couldthus have a positive effect on respiratoryrhythm, as reported with music therapy(30). Reciprocally, dyspnea generatesanxiety (1) and relieving dyspnea de-creases anxiety (31). In line with this,anxiety significantly decreased in those ofour patients who reported improvementsafter adjustments of the ventilator set-tings (Table 3). Therefore, relieving dys-pnea is likely to have positive effects onanxiety. Like with anxiety, pain was morefrequently experienced in dyspneic pa-tients than in nondyspneic ones (Table2). Pain also stimulates ventilation (32),and its control could possibly have bene-ficial effects on dyspnea. Because of thisinterrelationship, we believe that anxiety,dyspnea, and pain should be assessed sys-

Table 2. Characteristics of the patients at the time of dyspnea assessment


(n ! 96)Dyspnea(n ! 45)

No Dyspnea(n ! 51) p

Anxiety, no. (%) 44 (46) 32 (71) 12 (24) ".0001If yes, median anxiety visual analog scale (IQR) 5 (4–7) 5 (4–7) 5 (4–6) .64Pain, no. (%) 28 (29) 19 (42) 9 (18) .013If yes, median pain visual analog scale (IQR) 5 (4–6) 5 (4–6) 5 (4–7) .62Median heart rate (IQR), beats/min 93 (81–107) 98 (82–107) 87 (76–106) .15Median respiratory rate (IQR), cycles/min 20 (17–25) 20 (18–28) 19 (17–24) .14Median temperature (IQR), °C 37.3 (36.9–37.7) 37.3 (37.0–37.7) 37.2 (36.9–38.0) .86Median systolic blood pressure (IQR), mm Hg 122 (110–137) 127 (110–140) 119 (108–134) .15Median diastolic blood pressure (IQR), mm Hg 71 (59–83) 73 (60–88) 69 (58–76) .059Abnormalities in breathing dynamics, no. (%) 22 (23) 11 (24) 11 (22) .81Normal auscultation, no. (%) 35 (36) 16 (36) 19 (37) #.99Median hemoglobin (IQR), g/dL 10 (9–12) 10 (9–13) 10 (9–12) .92Median PaO2/FIO2 ratio (IQR) 236 (176–318) 228 (152–318) 241 (191–317) .42Median PaCO2 (IQR), mm Hg 41 (35–55) 42 (38–55) 40 (34–50) .28No radiograph abnormalities, no. (%) 25 (27) 13 (30) 12 (24) .49ACV, no. (%) 54 (56) 31 (69) 23 (45) .024Ventilator settings, no. (%) .094Pressure support ventilation; pressure support "15 cm H2O 19 (20) 6 (13) 13 (25)Pressure support ventilation; pressure support !15 cm H2O 23 (24) 8 (18) 15 (29)ACV; inspiratory flow "60 L/min 23 (24) 15 (33) 8 (16)ACV; inspiratory flow !60 L/min 31 (32) 16 (36) 15 (29)Median minute ventilation, L/min (IQR) 9 (8–12) 10 (8–12) 9 (7–13) .87

IQR, interquartile range; ACV, assist-control ventilation.

Table 3. Effects of adjusting ventilator settings inthe patients reporting dyspnea

ParameterMean Variation, cm

(95% Confidence Interval) p

Dyspnea VAS $4.6 ($6.1 to $3.2) .0005Anxiety VAS $1.7 ($3.3 to $0.2) .041Pain VAS %0.3 ($0.7 to %1.2) .79

VAS, visual analogic scale.Ventilator settings were considered involved

in the pathogenesis of dyspnea if and when thepostintervention VAS rating was inferior by atleast 1 cm to the preintervention one.


• 呼吸困難を訴えた45人全員が人工呼吸器設定調整の介入を受けた。うち16名（35%）が、呼吸困難が改善した。Ø 介入後のVAS評点が介入前よりも少なくとも1 cm減少した場合は、人工呼吸器の設定が呼吸困難に関与していると判断した。

• 呼吸器設定調整により、有意に呼吸困難のVASの評点が低下した。• 不安のVASの評点の有意な減少も伴ったが、痛みの有意な低下は見られなかった。

tematically and together in mechanicallyventilated ICU patients.

Role of Underlying Disease. Neuro-muscular diseases may have been over-represented in our study respective toprevious surveys (33), because one of thecenters (i.e., Raymond Poincare TeachingHospital) is specialized in the care of suchpatients. However, the prevalence of dys-pnea, its intensity, and its characteristicsdid not seem to depend on the cause ofrespiratory failure. This suggests thatdyspnea was not strongly related tomechanisms specific to the underlyingrespiratory disease. Indeed, patients withneuromuscular disease rather report ex-cessive inspiratory effort resulting fromweakness of respiratory muscles and con-sequent increased neuromotor output(34), whereas patients with chronic ob-structive pulmonary disease tend to re-port air hunger that is ascribed to thehyperinflation-induced volume restric-tion and consequent neuromechanicaluncoupling of the respiratory system (35,36). However, we acknowledge that thesmall number of patients with chronicobstructive pulmonary disease patients inour population prevented us from observ-ing such a relationship.

Contribution of Ventilator Settings.Dyspnea responded to changes in venti-lator settings in 35% of our dyspneic pa-tients, suggesting at least their partialresponsibility. There was an independentassociation between dyspnea and the ACVmode, and five of the seven patients whochose “excessive respiratory effort” as the

sole descriptor of their dyspnea receivedthis ventilatory mode with an inspiratoryflow !60 L/min. This is not surprisingbecause this mode leaves almost no roomfor natural respiratory fluctuations andinduces patient–ventilator dyssynchrony(37). Conversely, positive pressure-assisted ventilation (38), proportional-assisted ventilation (38), and negativepressure (39) in healthy subjects induceless dyspnea. However, a study by Knebelet al (12) reported that ACV and pressuresupport ventilation were equivalent interms of dyspnea and anxiety. In our pa-tients, the intervention that was the mostoften associated with dyspnea relief wasan increase in inspiratory flow. This is inline with data from normal volunteers(40) in whom low inspiratory flow is as-sociated with higher inspiratory muscleenergy expenditure (41) and in whomhigh inspiratory flow optimizes respira-tory muscle relaxation (42). Of note,since as early as 1994, clinical guidelineshave mentioned the importance of set-ting high inspiratory flows during ACV(43). Low pressure support levels havealso been associated with a sense of ex-cessive inspiratory effort (44), and lowtidal volumes are associated with air hun-ger in normal subjects (45) and in quad-riplegics (46). We did not observe theseassociations in our patients, possibly as aresult of the small size of the population,but this information must be kept inmind given the generalized use of lowtidal volume ventilatory assistance.

We did not find any significant rela-tionship between dyspnea and PaO2,PaCO2, MV-associated respiratory compli-cations, anemia, etc. This does not meanthat these factors are clinically unimport-ant on an individual basis, and largerstudies will be needed to establish in whatorder and to what extent these elementsmust be taken into consideration in clin-ical practice.

Clinical Outcomes. In addition to alonger ICU stay in dyspneic patents, weobserved that patients without dyspnea orin whom dyspnea responded favorably tothe adjustment of ventilator settings weremore often successfully extubated withinthe 3 subsequent days. This provides im-portant clinical relevance to our settingadjustment procedure. We acknowledgethat this procedure was rudimentary,leading to a possible underestimation ofthe number of patients in whom adjust-ing the ventilator could have been help-ful. We also acknowledge that this findingshould be confirmed in a larger cohort ofpatients in whom a weaning procedurewould be standardized and controlled. Ofnote, both participating centers used theFrench guidelines on weaning (47) anddyspnea and its fluctuations were nottaken into account by the physician incharge of deciding when to extubate.

This finding also suggests that an in-trinsic dyspnogenic factor delayed extu-bation in those patients whose dyspneafailed to recede after the adjustment ofventilator settings. In two-thirds of them,no specific cause of dyspnea was identi-

Table 4. Characteristics and clinical outcomes as a function of the response of dyspnea to the adjustment of ventilator settings

No Dyspnea or PositiveResponse to Ventilator

Setting Adjustments (n " 67)

Absence of Response toVentilator Setting

Adjustments (n " 29) p

Median length of intensive care unit stay (IQR), days 8 (4–23) 12 (9–24) .091Median duration of mechanical ventilation (IQR), days 5 (2–17) 7 (4–18) .68a

No. (%) successful extubations within 3 days 27 (40) 5 (17) .034Indication for mechanical ventilation .26

Hypoxemic acute respiratory failure, no. (%) 32 (48) 14 (48)Decompensation of an underlying neuromuscular disease, no. (%) 17 (25) 12 (41)Coma, no. (%) 8 (12) 2 (7)Chronic obstructive pulmonary disease, no. (%) 7 (10) 0 (0)Other, no. (%) 3 (4) 1 (3)

Median dyspnea visual analog scale at baseline inclusion (if dyspnea) (IQR) 5 (5–7) 5 (2–7) .11Median Simplified Acute Physiology Score II at admission (IQR) 42 (33–58) 44 (26–60) .88Median duration of mechanical ventilation before inclusion (IQR), days 3 (1–6) 3 (1–6) .67Male gender, no. (%) 40 (60) 20 (69) .49Median age (IQR), yrs 67 (52–73) 54 (41–72) .19Median PaO2/FIO2 ratio (IQR) 236 (188–323) 238 (155–304) .63Assist-control ventilation, no. (%) 36 (54) 18 (62) .51

IQR, interquartile range.aGray’s test; p " .15 Wilcoxon test. The absence of response of dyspnea to the adjustment of ventilator settings was defined as the absence of a !1-cm

reduction in the visual analog scale scoring of the dyspneic sensation after this adjustment.

2063Crit Care Med 2011 Vol. 39, No. 9

• 呼吸困難の評価後3日以内の抜管成功は、人工呼吸器設定の調整に反応して呼吸困難が改善しなかった患者では、他の患者（「非呼吸困難者」および「呼吸困難が改善した患者」）より有意に少なかった。

• 背景– ICUにおける死の質と終末期ケアの評価のために、家族の観点から、終末期ケアのどの分野がも重要であるかは不明である。

• 目的– ICUで亡くなった患者の家族による死と死の経験の質を評価すること– 家族による死の評価の質の高さに関連するICUの経験の要素を特定すること。

• デザイン– ICU QODD（the Quality of Dying and Death）を使用した、後ろ向き研究– 患者の死後、4ヶ月〜1年の間に連絡を取った。

• 対象– 大学教育病院または学術退役軍人医療センター（VA）のいずれかの

4つのICUで死亡した患者の家族– 2000年1/1〜12/31に亡くなった患者家族– 死亡した時点でその子孫が少なくとも21歳– 少なくとも3日の滞在の後にICUで死亡した場合

Quality of Dying in the ICU* Ratings by Family MembersCHEST 2005; 128:280–287

quality dying. The Study To Understand Prognosesand Preferences for Outcomes and Risks of Treat-ments investigators16–17 reported significant pain atthe end of life in more than half of patients dying inthe hospital or ICU. Although studies11–17 documenta high prevalence of symptoms in hospital-baseddying, independent associations to the quality ofdying and death in the ICU have yet to be identifiedfrom a family perspective. Outcome measures usedin prior studies,7,16 such as the use of artificialventilation, the timing of do-not-resuscitate (DNR)orders, longer duration of stay in the ICU, and theoccurrence of death in the ICU, have been definedas unfavorable components of dying, but these indi-cators may not be sensitive or appropriate for inter-ventions designed to improve the quality of end-of-life care. Experts have suggested that we may not beassessing the best outcome measures for the qualityof end-of-life care in the ICU.18–21

Despite the prevalence of ICU deaths and docu-mentation of suffering in this large population ofdying patients, no validated measures exist to assessthe overall quality of dying in the ICU. Researchershave identified relevant domains of quality dyingexperiences and are developing tools to assess thequality of dying and death.22–27 The Quality of Dyingand Death (QODD) questionnaire is one tool devel-oped to attempt to measure this construct.22–23 It isunknown what domains of end-of-life care are mostimportant from the family’s perspective for evalua-tion of the quality of dying and end-of-life care in theICU.

This study used a modification of the QODDinstrument (the ICU QODD) to better understandthe family perspective of the experience of dying inan ICU.22–23 Our objectives were to assess ratings ofthe quality of the dying and death experience byfamily members of patients who died in an ICU, andto identify components of the ICU experience asso-ciated with higher quality of dying ratings by familymembers.

Methods and Materials

Study Design and Sampling Strategy

We performed interviews with multiple family members ofpatients who died in an ICU to measure the quality of the dyingexperience. We enrolled family members of patients who died inany of four ICUs of either a university teaching hospital or anacademic Veterans Affairs Medical Center (VA). All deathsduring the calendar year 2000 were reviewed for inclusion.Inclusion criteria required that the decedents were at least 21years old at the time of death and died in the ICU after a stay ofat least 3 days. Family members were contacted between 4months and 1 year after the death. Two to four family memberswho were themselves at least 21 years old and had visited a

minimum of 6 h in the ICU before their loved-one’s death wereidentified for each decedent. This study was conducted with theapproval of the Institutional Review Boards at both institutions,and all family participants gave written informed consent.

Data Collection and Measures

The perceptions by family members of their loved-one’s qualityof dying was measured with the QODD instrument.21–22 Familyperspectives are one source to evaluate the dying experience butcannot be viewed as interchangeable with patient perspectiveswhen available.28–29 The original 31-item QODD instrument (onwhich the current reduced version is based) was developedthrough qualitative studies of patients, family members, andclinicians for completion after death by family members orclinicians.22 The QODD instrument was developed in six con-ceptual domains as shown in Table 1.22 The QODD instrumentwas validated in family respondents in one study based on 204deaths in Missoula County and in another based on 100 patientsreceiving hospice care.23,28 The ICU QODD instrument consistsof 23 of the original 31 questions of QODD instrument thatassess symptoms, experiences, and perceptions about the qualityof the dying process in the last week of life. The 23 items werechosen based on face validity for items having relevance fordeaths occurring in the ICU and also confirming this assessmentby examining nonresponse patterns for ICU deaths (Table 1presents a complete listing of the items in the ICU QODDinstrument). In a prior report from this study,29 we found theICU QODD instrument to have statistically significant, althoughmoderate interrater reliability for patients dying in the ICU whencompleted by two to four family members.

Table 1—Specific Items in the ICU QODD InstrumentOrganized by Conceptual Domains

Symptoms and personal careHaving pain under controlHaving control of eventsBeing able to feed oneselfBeing able to breath comfortably

Preparation for deathFeel at peace with dyingBe unafraid of dyingHave health-care costs providedVisits from religious/spiritual advisorHave a spiritual service or ceremonyHave funeral arrangements in orderSay goodbye to loved onesClear up bad feelings

Family concernsSpending time with family/friendsSpending time alone

Treatment preferencesDiscuss wishes for end-of-life careBeing on ventilatorBeing on dialysis

Whole-person concernsFinding meaning and purposeBeing touched and hugged by loved onesBeing able to laugh and smileKeeping one’s dignity and self-respect

Moment of deathState at moment of deathHaving family present at moment of death

www.chestjournal.org CHEST / 128 / 1 / JULY, 2005 281

症状とパーソナルケア痛みをコントロールできるイベントをコントロールできる自分で食事ができる快適に呼吸できること

死の準備心が安まる死ぬことを恐れない医療費の支給がある宗教的/精神的なアドバイザーの訪問霊的な奉仕やセレモニーを行う順番に葬儀の手配をする愛する人に別れを告げる悪い感情を整理する

家族の懸念家族や友人と過ごす時間一人で過ごす時間

治療の選択終末期ケアの希望について話し合う人工呼吸器の使用透析の使用

全人的な懸念意味と目的を見つける愛する人に触れられ、抱きしめられる笑いと微笑みがある尊厳と自尊心を保つ

死の瞬間死の瞬間の状態死の瞬間に家族がいる

eter strength and minimize the number of predictors whilemaintaining an adjusted R2 near 0.60. We explored for theinfluence of independent variables of historical or potential novelimportance to the end-of-life experience in the ICU by addingback in variables that related to the aggressiveness of care (use ofventilator or dialysis, time in ICU, time in hospital, and others)and different respondent characteristics that might affect thequality rating (witness of upsetting treatments, limitations tovisiting, time the family member knew the decedent, and others).

Because there were only 38 independent deaths, models werealso built using repeated-measure analysis and generalized esti-mating equations to account for correlations within families.Multilevel linear modeling for repeated measures was done usingintraclass correlation coefficient method.33–34 Regression for thisconfirmatory process was performed in SAS with Proc Genmodand Proc GLM (SAS Institute).

ResultsRecruitment and Patient/Family Characteristics

We identified 365 ICU deaths at the study sitesduring the study year. One hundred eight decedentsmet the inclusion criteria, and families of 38% of theeligible decedents were enrolled; details on enroll-ment were reported in a prior publication.29 Themost common reasons cited for nonparticipation offamily members were family members feeling tooemotional (32%) or only one family member willingto participate (24%). Ninety-four family memberscompleted the interview. Sixty-eight percent of thedecedents were male, 39% died at the VA, and 61%died at the university (Table 2). Ninety-seven per-cent were white, and the mean age was 59 years.Decedents had a median length of stay in the ICU of6 days (range, 3 to 43 days). Families reported that88% of patients had received either mechanicalventilation or hemodialysis. The mean APACHE IIscore at time of admission to the ICU was 21. Theattributable cause of death was distributed amongthe following organ systems: cardiac (32%), respira-tory (29%), neurologic (18%), GI (16%), and meta-bolic/renal (5%).

Of the 70 nonparticipating eligible decedents,71% were male, 35% died at the VA, 93% werewhite, and mean age was 65 years (SD, 15 years).There were no significant differences in demo-graphic characteristics or characteristics of diseasebetween decedents whose family members enrolledand decedents whose family members declined par-ticipation. Similarly, no significant demographic dif-ferences existed between the family members whoparticipated and those who were contacted butrefused to participate (n ! 59).

Experience of Dying in the ICU

Family members reported that 24% of the patientsdying in the ICU were never aware they were dying

and 34% were aware of dying only in the last week oflife. They reported that 76% of dying ICU patientswere able to talk with them in a way they couldunderstand in the last week of life. All but fourpatients had a DNR order written prior to death(89%) at an average of 1 day before death (range, 0to 30 days).

The mean ICU QODD score for the decedents inthis study was 60 points on the 0- to 100-point scale(SD, 14) [Table 3]. The ICU QODD score had anormal distribution and was strongly correlated(r ! 0.47 to 0.98) with the other single-item out-come measures of quality of the dying experience(quality of moment of death, ICU as place of deathrating, and the quality of life of last 7 days of life).The quality of moment of death was rated at 67 of100 points, and the ICU as the place of death wasrated at 61 of 100 points. The rating for the quality ofthe last 7 days of life was low (32 points on a 0- to100-point scale), and families reported that thepatients experienced substantial physical symptomsin the last week of life (Table 3). Patients werereported by families to have satisfactory pain controlless than half the time. The medical record wasremarkable in the near universal absence of dataregarding assessment of pain and other symptoms orin how sedatives and analgesics were titrated tosymptom relief.

Table 2—Characteristics of the Patients Who Died inthe ICU (n ! 38)*

Characteristics Data

Mean age (SD), yr 59 (15)Male gender 68White race 97Family conference 45ECOG score 1 mo prior (% " 2) 58Use of mechanical ventilator or kidney dialysis 88Family asked to participate in withdrawal of care

decisions87

Family asked to participate in withholding of caredecisions

71

Families reporting enough time to visit withloved ones

87

Days in hospital 9 (3–60)Days in ICU 6 (3–43)Days from writing DNR to death 1 (0–30)Mean (SD) APACHE II score 21 (6)Primary organ system failure at time of death

Cardiovascular 32Respiratory 29Neurologic 18GI 16Renal/metabolic 5

*Data are presented as % or median (range) unless otherwiseindicated.

www.chestjournal.org CHEST / 128 / 1 / JULY, 2005 283

Associations to Higher Quality-of-Dying Ratings

The strongest independent correlations with theICU QODD score were feeling at peace with dying,having control of events, being unafraid of dying,keeping dignity and self-respect, and finding mean-ing and purpose (Table 4). Aspects that did notappear to have important associations with qualityICU dying include the following: days in hospital orICU; having DNR order written within 2 days ofdeath; use of mechanical ventilator or kidney dialysis;frequency of visiting with any family members orspending time alone; decreased functions such asinability to feed self or maintain bowel/bladder con-trol; state of consciousness at the moment of death;family participation in withholding or withdrawaldiscussions; whether or not someone was present atthe time of death; and whether or not a familyconference was held. There was no association ofICU QODD score detected for demographic char-acteristics of the decedent such as system of primaryorgan system derangement, APACHE II score, race,or gender. There were no differences in ICU QODDscores or other quality ratings across the four ICUsor two hospitals of the study.

Multivariate analysis produced best models with 9to 10 independent variables that achieved a maxi-mum adjusted R2 value near 0.65 or mean deviance1.1 by hierarchical modeling. A four-variable model

achieved independently significant t tests for allparameters and maintained an adjusted R2 value at0.60 or mean deviance 1.06 by hierarchical model-ing. Results of hierarchical linear modeling with thegeneralized estimating equations, repeated-mea-sures technique, and multivariate linear regressionacross all 91 valid respondents resulted in identicallycomposed models. We present the data from multi-variate linear regression across the 91 respondents.

Four individual items best account for the overallquality-of-dying ratings as measured by the ICUQODD score: how often the patient appeared tohave his or her pain was under control, how often thepatient appeared to have control over what was goingon around him or her, how often the patient ap-peared to feel at peace with dying, and how often thepatient appeared to keep his or her dignity andself-respect (Table 5). After controlling for thesefour explanatory variables, none of the followingwere associated with the rating of the quality of ICUdying: APACHE score (p ! 0.58); number of days in

Table 3—Family Members’ Perceptions of Patient’sICU Dying and Death Experience

Variables Data

Dependent outcome variables, mean (SD)ICU QODD score (0–100) 60 (14)Quality of moment of death (0–100) 67 (25)ICU as place of death (0–100) 61 (28)Quality of life last 7 d (0–100) 32 (26)

Patient’s experiences in the dying process(independent explanatory variables), %

YesUnconscious just before moment of death 87Enough time for family to visit 87Visit with a religious or spiritual advisor 78Able to communicate during the last 7 d 76Aware of dying by last 7 d of life 76Discuss wishes regarding care with physician 47Saying goodbye to loved ones 47Finding meaning and purpose 32

Most/all of the timePain under control 47Unafraid of dying 39Keeping dignity and self-respect 32Feeling at peace with dying 30Control of events 8Control of bowel and bladder 5Breathing comfortably 3Able to feed themselves 3Energy to do things 0

Table 4—Bivariate Associations to the ICU QODDInstrument Rating Score

FactorsCorrelationCoefficient p Value

Factors with significant correlations " 0.3Feeling at peace with dying 0.69 # 0.001Control of events 0.62 # 0.001Unafraid of dying 0.59 # 0.001Keeping dignity and self-respect 0.50 0.001Finding meaning and purpose 0.50 0.002Saying goodbye to loved ones 0.44 0.006Aware of dying " 7 d 0.42 0.005Pain under control 0.42 0.009Enough time to visit in the ICU 0.33 0.004ECOG score ! 2 1 mo prior to dying 0.33 0.043

Selected factors that did not demonstrateassociation with higher ICU QODDscores by bivariate analysis

Age 0.01 0.61Days in hospital 0.01 0.86Days in the ICU 0.01 0.58Having DNR written within 2 d of

death0.02 0.85

Having visit by spiritual advisor 0.03 0.93Having a family member present at

moment of death0.04 0.89

Having family conference 0.07 0.69Family asked to participate in

withdrawal decisions0.11 0.53

Family asked to participate inwithholding decisions

0.11 0.51

Use of mechanical ventilator or kidneydialysis

0.14 0.28

APACHE II score 0.18 0.28Discussing care preferences with

physician0.20 0.19

Able to communicate last 7 d 0.23 0.53

284 Clinical Investigations

• Crit Care Med. 2010; 38:808–818. • 家族が、患者が終末期に快適に呼吸ができていたと認識していたのは、3％だけだった。

ICUでのDyspneaの頻度のまとめ

• 基礎疾患にかかわらず、一般的にも多く見られる症状で

ある。

• 苦痛の程度は、痛み以上にも苦痛を感じていた。

• 不安と強く相関していた。• 人工呼吸器管理患者の約半数で、呼吸困難を感じていた。• 呼吸困難の有病率、その強度、その特徴は、呼吸不全の原因に左右されなかった。

• 人工呼吸器設定は、呼吸困難の少なくとも部分的な要因となることが示唆されている。

Dyspneaの評価

• 従来から使用されている評価ツール– NRS（Numerical Rating Scale）– VAS（Visual Analogue Scale）＜特徴＞

Ø呼吸困難は多次元感覚であるが、自己申告の一次元（一つの局面からの程度）で評価されている

Øコミュニケーションの取れる患者が対象

• 自己申告能力が低下している非コミュニケーション患者の呼吸苦を客観的に評価したツール– respiratory and behavioral signs (respiratory distress

observation scales [RDOS])– intensive care RDOS [IC-RDOS]

Diagnostic Accuracy of Respiratory Distress Observation Scales as Surrogates of Dyspnea Self-report in Intensive Care Unit Patients

Anesthesiology 2015; 123:830-7• 背景

– 呼吸に伴う苦痛は急性疾患患者において主要症状であるが、正当に評価されていない。

– コミュニケーションが取れない重症患者の呼吸困難を評価するための検証された客観的ツールはない。

– 緩和ケアにおいて、呼吸困難を自己申告する代わりのツールとして、Respiratory Distress observation scale（RDOS：呼吸困難観察尺度）が検証されている。

• 目的

– ICU患者の呼吸困難の推定ツールとしてRDOCおよびIC-RDOS（intensive care RDOS）と、D-VAS（dyspnea visual analog scale）との関係を評価すること。

• デザイン：前向き観察研究

• 対象：1500床の大学病院の16床のICU• 期間：2011年12月2日〜2012年4月13日（derivation data）,2013年7月22日〜

2013年10月29日（validation data）

• 方法– 入院の初の24時間（平日のみ）の午前8時から午前10時の間

– 担当医師と担当看護師それぞれによりデータ収集– 呼吸困難感の有無の査定

Ø 「よく呼吸できますか」、「呼吸が快適ですか」、「呼吸に悩まされますか」、「呼吸が困難です」：少なくとも2つの異なる質問をして、答えは一致していなければならない。

– 呼吸困難の強度Ø 10 cm D-VAS：「呼吸困難を伴わない」から「耐えられない呼吸器障害」まで

ALN201412073_SUPPLEMENTAL DIGITAL CONTENT

ALN201412073 Supplemental Digital Content 1 Table 1 Original respiratory Distress Observation Scale (RDOS) as described by Campbell et al. [adapted from Campbell ML, J Palliat Med 2010]

Variable 0 Point 1 Point 2 Points Total

Heart rate (beats per minute) < 90 90 - 109 ≥ 110

Respiratory rate (breaths per minute) ≤ 18 19 - 30 > 30

Restlessness: non purposeful movements None Occasional slight movements Frequent movements

Paradoxical breathing pattern: abdomen moves in on inspiration None - Present

Accessory muscle use: rise in clavicle during inspiration None Slight rise Pronounced rise

Grunting at end expiration: guttural sound None - Present

Nasal flaring: involuntary movement of nares None - Present Look of fear Eyes wide open Facial muscles tense Brow furrowed Mouth open Teeth together Total

心拍数

呼吸数

落ち着きのなさ：意図しない動き

奇異呼吸：吸気時の腹部の動き

呼吸補助筋の使用：吸気時の鎖骨の挙上

呼気の終わりの呻吟：荒くゼイゼイする喉音

鼻翼呼吸：鼻腔の不本意な動き

恐れの表情

オリジナルのRDOS（呼吸困難観察尺度）

Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

Anesthesiology 2015; 123:830-7� �� Persichini et al.

Respiratory Distress Observational Scales

The patients were evaluated only once (no indication about responsiveness). They were evaluated during the first day of their stay (hence a low proportion of communicating patients

under mechanical ventilation; table 1). Patients were included in the derivation study on a consecutive basis, but this was done on week days only and with a “real-life” concern: we insisted on having the physician in charge of the patient col-lecting the data rather than a dedicated investigator. This resulted in a number of patients being missed because each physician was in charge of overseeing more than one patient. This could have biased the results toward a less severe popu-lation. We do not think that this occurred because the 193 communicant and noncommunicant patients in whom RDOS data were gathered (fig. 1) had severity scores identi-cal to those of the admitted population (e.g., median simpli-fied acute physiology score II of 42 and 43, respectively). For all the above reasons, it is not possible to claim generalized validity for RDOS or IC-RDOS: other study designs in other contexts would probably lead to different numerical results. This should, however, not invalidate the notion that the observation scales can be useful to assess dyspnea in the ICU.

D-VAS was chosen as the reference index in this study because it is, to the best of our knowledge, the only psycho-metric tool that has been used to evaluate dyspnea in ICU patients in the literature. Visual analog scales are widely used for the measurement of symptoms, and their validity to assess dyspnea has been established by several studies in various con-texts.26 VAS is considered most suited as a “within-subject” tool and less so to compare patients, even though the com-parison between a dyspnea VAS and a verbal rating scale has shown quasiredundancy in cancer patients.27 In the ICU, VAS to measure dyspnea has not been the object of a specific valida-tion, but it has been shown that it was feasible in this context, well correlated with anxiety (which is clinically and physiolog-ically expected), and responsive to therapeutic intervention.6 Of interest, the mere fact that we did find a relation between D-VAS (purely psychometric) and RDOS/IC-RDOS (purely physical) not only support the putative interest of observation scales to evaluate respiratory suffering but also, reciprocally, lends support to some “between-patient” validity of D-VAS.

Finally on this, dyspnea is multidimensional1,2 and imper-fectly apprehended by unidimensional tools such as D-VAS. RDOS and IC-RDOS, because they incorporate “facial expres-sion of fear,” could correlate better a multidimensional score, like the recently validated “multidimensional dyspnea profile.”28 This hypothesis would be interesting to test in future studies.

Comparison with Available DataTo our knowledge, and even though the use of RDOS has recently been advocated in palliative ICU situations,7 there is no previous study of RDOS in ICU patients. Yet, the relation between fear display and an asphyxial threat has first been noted during failed mechanical ventilation weaning trials.10 Fear display was present across a wide range of cognitive states,10 which probably should make it sufficient to bring dyspnea to the mind of ICU caregivers. The interest of a com-posite clinical score such as RDOS or IC-RDOS is under-lined by the fact that, in a study of dyspnea in mechanically

Fig. 2. Spearman rank correlation plot between dyspnea vi-sual analog scale (D-VAS) and the respiratory distress obser-vation scale (RDOS, A) or the modified intensive care RDOS (IC-RDOS, B) in the 120 patients from the derivation cohort who were able to communicate verbally with their physician.

Table 2. Calculation of the Modified Intensive Care Respiratory Distress Observation Scale

Parameter Score

0 3.31—Heart rate (beats/min) + (heart rate)/652—Use of neck muscle during inspiration If present +1 If absent −13—Abdominal paradox during inspiration If present +1 If absent −14—Facial expression of fear If present +1 If absent −15—Supplemental oxygen If present +0.7 If absent −0.7

吸気時の頚部の筋肉の使用

奇異呼吸：腹部が吸気時に陥没、呼気時に膨隆

恐怖の表情

酸素補給

心拍数

IC-RDOS（重症患者用に修正された呼吸困難観察尺度）

Copyright © 2015, the American Society of Anesthesiologists, Inc. Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.

Anesthesiology 2015; 123:830-7� �� Persichini et al.

Respiratory Distress Observational Scales

The patients were evaluated only once (no indication about responsiveness). They were evaluated during the first day of their stay (hence a low proportion of communicating patients

under mechanical ventilation; table 1). Patients were included in the derivation study on a consecutive basis, but this was done on week days only and with a “real-life” concern: we insisted on having the physician in charge of the patient col-lecting the data rather than a dedicated investigator. This resulted in a number of patients being missed because each physician was in charge of overseeing more than one patient. This could have biased the results toward a less severe popu-lation. We do not think that this occurred because the 193 communicant and noncommunicant patients in whom RDOS data were gathered (fig. 1) had severity scores identi-cal to those of the admitted population (e.g., median simpli-fied acute physiology score II of 42 and 43, respectively). For all the above reasons, it is not possible to claim generalized validity for RDOS or IC-RDOS: other study designs in other contexts would probably lead to different numerical results. This should, however, not invalidate the notion that the observation scales can be useful to assess dyspnea in the ICU.

D-VAS was chosen as the reference index in this study because it is, to the best of our knowledge, the only psycho-metric tool that has been used to evaluate dyspnea in ICU patients in the literature. Visual analog scales are widely used for the measurement of symptoms, and their validity to assess dyspnea has been established by several studies in various con-texts.26 VAS is considered most suited as a “within-subject” tool and less so to compare patients, even though the com-parison between a dyspnea VAS and a verbal rating scale has shown quasiredundancy in cancer patients.27 In the ICU, VAS to measure dyspnea has not been the object of a specific valida-tion, but it has been shown that it was feasible in this context, well correlated with anxiety (which is clinically and physiolog-ically expected), and responsive to therapeutic intervention.6 Of interest, the mere fact that we did find a relation between D-VAS (purely psychometric) and RDOS/IC-RDOS (purely physical) not only support the putative interest of observation scales to evaluate respiratory suffering but also, reciprocally, lends support to some “between-patient” validity of D-VAS.

Finally on this, dyspnea is multidimensional1,2 and imper-fectly apprehended by unidimensional tools such as D-VAS. RDOS and IC-RDOS, because they incorporate “facial expres-sion of fear,” could correlate better a multidimensional score, like the recently validated “multidimensional dyspnea profile.”28 This hypothesis would be interesting to test in future studies.

Comparison with Available DataTo our knowledge, and even though the use of RDOS has recently been advocated in palliative ICU situations,7 there is no previous study of RDOS in ICU patients. Yet, the relation between fear display and an asphyxial threat has first been noted during failed mechanical ventilation weaning trials.10 Fear display was present across a wide range of cognitive states,10 which probably should make it sufficient to bring dyspnea to the mind of ICU caregivers. The interest of a com-posite clinical score such as RDOS or IC-RDOS is under-lined by the fact that, in a study of dyspnea in mechanically

Fig. 2. Spearman rank correlation plot between dyspnea vi-sual analog scale (D-VAS) and the respiratory distress obser-vation scale (RDOS, A) or the modified intensive care RDOS (IC-RDOS, B) in the 120 patients from the derivation cohort who were able to communicate verbally with their physician.

Table 2. Calculation of the Modified Intensive Care Respiratory Distress Observation Scale

Parameter Score

0 3.31—Heart rate (beats/min) + (heart rate)/652—Use of neck muscle during inspiration If present +1 If absent −13—Abdominal paradox during inspiration If present +1 If absent −14—Facial expression of fear If present +1 If absent −15—Supplemental oxygen If present +0.7 If absent −0.7

- D-VASとRDOSは有意に相関していた。

- D-VASとIC-RDOSは有意に相関していた

- D-VAS対IC-RDOS相関はD-VAS対RDOS相関より有意に優れていた（P = 0.04）

Dyspneaの管理

Management of Dyspnea in theTerminally IllLara Pisani, MD, PhD; Nicholas S. Hill, MD, FCCP; Angela Maria Grazia Pacilli, MD; Massimiliano Polastri, MSc;and Stefano Nava, MD

The genesis of dyspnea involves the activation of several mechanisms that are mediated and

perceived depending on previous experiences, values, emotions, and beliefs. Breathlessness

may become unbearable, especially in patients who are terminally ill, whether afflicted by

respiratory-, cardiac-, or cancer-related disorders, because of a final stage of a chronic process,

an acute event, or both. Compared with pain, palliation of dyspnea has received relatively little

attention in clinical practice and the medical literature. This is particularly true when the

breathlessness is associated with acute respiratory failure because most of the studies on

pharmacologic and nonpharmacologic treatments of respiratory distress have excluded such

patients. Assessments of the quality of dying for patients in an ICU consistently show that few

patients are considered by family members to breathe comfortably at the end of their life. This

review focuses on the management of dyspnea in patients with advanced terminal illness,

summarizing clinical trial evidence on pharmacologic and nonpharmacologic interventions

available for these patients. CHEST 2018; 154(4):925-934

KEY WORDS: cancer; COPD; dyspnea; noninvasive ventilation; opioids

“How people die lives on in the memoriesof those who live on.”

Cicely Saunders

As summarized in an American ThoracicSociety position paper on mechanisms ofdyspnea in COPD, dyspnea is defined as “thesubjective experience of breathingdiscomfort” and the physiology is complex.1

It involves the activation of severalmechanisms that lead to increased work ofbreathing; stimulation of receptors inairways, lung parenchyma, or chest wall; andexcessive stimulation of the respiratory

center by peripheral or centralchemoreceptors. The sensation of dyspnea isalso complex and is dynamically mediatedand perceived depending on our previousexperiences, values, emotions, and beliefs.Therefore, the symptom of dyspnea hassubtypes that are described in different waysand derived from different respiratory andnonrespiratory disorders, as summarized inTable 1. Furthermore, as determined by PETscan, relief of dyspnea involves acharacteristic brain activation different fromthat subserving dyspnea perception.2 We all

ABBREVIATIONS: HFNT = high-flow nasal therapy; NIV = noninva-sive ventilation; RCT = randomized controlled trialAFFILIATIONS: From the Department of Cardiac-Thoracic andVascular Diseases (Dr Pisani), Respiratory and Critical Care Unit,University Hospital St. Orsola-Malpighi, Bologna, Italy; the Division ofPulmonary, Critical Care, and Sleep Medicine (Dr Hill), Tufts MedicalCenter, Boston, MA; the Department of Clinical, Integrated andExperimental Medicine (Drs Pacilli and Nava), Alma Mater Studio-rum, University of Bologna, Bologna, Italy; and the Medical Depart-ment of Continuity of Care and Disability, Physical Medicine and

Rehabilitation (Mr Polastri), University Hospital St. Orsola-Malpighi,Bologna, Italy.CORRESPONDENCE TO: Stefano Nava, MD, Department of Clinical,Integrated and Experimental Medicine (DIMES) and University Hos-pital St. Orsola-Malpighi Department of Cardiac-Thoracic andVascular Diseases, Respiratory and Critical Care Unit, Via G. Mas-sarenti 9, Bologna 40138, Italy; e-mail: [email protected] ! 2018 American College of Chest Physicians. Published byElsevier Inc. All rights reserved.DOI: https://doi.org/10.1016/j.chest.2018.04.003

[ Recent Advances in Chest Medicine ]

chestjournal.org 925

Management of Dyspnea in theTerminally IllLara Pisani, MD, PhD; Nicholas S. Hill, MD, FCCP; Angela Maria Grazia Pacilli, MD; Massimiliano Polastri, MSc;and Stefano Nava, MD

The genesis of dyspnea involves the activation of several mechanisms that are mediated and

perceived depending on previous experiences, values, emotions, and beliefs. Breathlessness

may become unbearable, especially in patients who are terminally ill, whether afflicted by

respiratory-, cardiac-, or cancer-related disorders, because of a final stage of a chronic process,

an acute event, or both. Compared with pain, palliation of dyspnea has received relatively little

attention in clinical practice and the medical literature. This is particularly true when the

breathlessness is associated with acute respiratory failure because most of the studies on

pharmacologic and nonpharmacologic treatments of respiratory distress have excluded such

patients. Assessments of the quality of dying for patients in an ICU consistently show that few

patients are considered by family members to breathe comfortably at the end of their life. This

review focuses on the management of dyspnea in patients with advanced terminal illness,

summarizing clinical trial evidence on pharmacologic and nonpharmacologic interventions

available for these patients. CHEST 2018; 154(4):925-934

KEY WORDS: cancer; COPD; dyspnea; noninvasive ventilation; opioids

“How people die lives on in the memoriesof those who live on.”

Cicely Saunders

As summarized in an American ThoracicSociety position paper on mechanisms ofdyspnea in COPD, dyspnea is defined as “thesubjective experience of breathingdiscomfort” and the physiology is complex.1

It involves the activation of severalmechanisms that lead to increased work ofbreathing; stimulation of receptors inairways, lung parenchyma, or chest wall; andexcessive stimulation of the respiratory

center by peripheral or centralchemoreceptors. The sensation of dyspnea isalso complex and is dynamically mediatedand perceived depending on our previousexperiences, values, emotions, and beliefs.Therefore, the symptom of dyspnea hassubtypes that are described in different waysand derived from different respiratory andnonrespiratory disorders, as summarized inTable 1. Furthermore, as determined by PETscan, relief of dyspnea involves acharacteristic brain activation different fromthat subserving dyspnea perception.2 We all

ABBREVIATIONS: HFNT = high-flow nasal therapy; NIV = noninva-sive ventilation; RCT = randomized controlled trialAFFILIATIONS: From the Department of Cardiac-Thoracic andVascular Diseases (Dr Pisani), Respiratory and Critical Care Unit,University Hospital St. Orsola-Malpighi, Bologna, Italy; the Division ofPulmonary, Critical Care, and Sleep Medicine (Dr Hill), Tufts MedicalCenter, Boston, MA; the Department of Clinical, Integrated andExperimental Medicine (Drs Pacilli and Nava), Alma Mater Studio-rum, University of Bologna, Bologna, Italy; and the Medical Depart-ment of Continuity of Care and Disability, Physical Medicine and

Rehabilitation (Mr Polastri), University Hospital St. Orsola-Malpighi,Bologna, Italy.CORRESPONDENCE TO: Stefano Nava, MD, Department of Clinical,Integrated and Experimental Medicine (DIMES) and University Hos-pital St. Orsola-Malpighi Department of Cardiac-Thoracic andVascular Diseases, Respiratory and Critical Care Unit, Via G. Mas-sarenti 9, Bologna 40138, Italy; e-mail: [email protected] ! 2018 American College of Chest Physicians. Published byElsevier Inc. All rights reserved.DOI: https://doi.org/10.1016/j.chest.2018.04.003

[ Recent Advances in Chest Medicine ]


• 認知と覚醒の低下のために、症状を認識し訴えることができない患者もいるため、末期呼吸困難の評価はより複雑なプロセスである。

• 呼吸困難の緩和に関して、疼痛に比べると比較的注目が少ない。（臨床、医学文献、一般の人の感覚でも）

• オピオイドなどの薬理学的治療の効果や酸素療法に関する研究のほとんどが呼吸困難の患者を除外している。呼吸困難が急性呼吸不全に関連している場合には特に当てはまる。

• 目的–末期段階で呼吸困難または急性呼吸不全を経験している患者に焦点を当て、息切れの薬理学的および非薬理学的治療を評価する。

–知識のギャップを特定し、末期疾患患者の呼吸困難の管理を改善する方法を検討する。

• 方法– PubMedとThe Cochrane Database を使用。– key words “dyspnea,” “palliative dyspnea,”

“breathlessness,” “palliative breathlessness,”“refractory breathlessness.”

–ナラティブレビューのため、追加の評価なしで定性分析を行なった。

Opioidについて

• オピオイドによる呼吸困難の緩和のメカニズムは、右後帯状回の中枢受容体に対する作用がある。

• オピオイドは、疼痛と同様に、「自発呼吸の抑制」と「皮質活動の調節」の両方によって作用する可能性がある。

• 低用量では必ずしも呼吸器の駆動力を変化させないが、高用量では呼吸抑制および意識レベルの低下が起こる可能

性がある。

• 臨床医の間での懸念は、オピオイドが呼吸不全患者の呼吸抑制、混乱、そして早死を含む有害事象を引き起こす可能

性があるということ。

Opioidについて（２）

• 近のコクランレビューに含まれる研究（26件）– 半分は「進行疾患および末期疾患を有する成人の難治性呼吸困難患者が対象

– 残りの半分は、急性呼吸不全または呼吸困難を除外。末期の癌患者または非癌患者が対象

• さまざまな薬物療法（ジアモルヒネ、ジヒドロコデイン、モルヒネ、およびフェンタニル）、製剤（霧状、皮下、および経口）、投与スケジュール（単回投与VS 複数投与）が使用されたが、これらの実施に対する推奨や結論には制限がある。

Cochrane Database Syst Rev. 2016;(3): CD011008.

Figure 5. Forest plot of comparison: 13 Opioids versus placebo - sensitivity analysis, outcome: 13.1Breathlessness.

In order to interpret this in a more meaningful context, we con-verted this standard mean difference to a mean difference on acommonly used dyspnoea scale, that being the VAS 100 mm scale.Using the post treatment standard deviation from a large study(Abernethy 2003), we calculated an effect size of 9.6 mm (95% CI−13.44 to −5.52) on a 100 mm VAS scale. The point estimateappears to meet the clinically important difference threshold butthe confidence intervals still include values which would not beconsidered clinically significant and therefore there is still someuncertainty about the effectiveness of the intervention.

Type of opioid

There was a strong treatment effect for morphine (post-treatmentscores: six studies, 188 participants; SMD −0.33, 95% CI −0.62to −0.04; P = 0.02 (Analysis 2.1); change from baseline: fourstudies, 134 participants, SMD −0.12, 95% CI −0.48 to 0.24;P = 0.50 (Analysis 2.1)), and for dihydrocodeine (post-treatmentscore: four studies, 107 participants; SMD −0.41, 95% CI −0.80to −0.03; P = 0.04 (Analysis 3.1)).There was no effect for hydromorphone (change from baseline:one study, 20 participants, MD −0.26; 95% CI −1.17 to 0.65;P = 0.57), oral diamorphine (post treatment: one study, 10 par-

ticipants; MD 0.50, 95% CI −1.44 to 2.44; P = 0.61), oxy-codone (change from baseline: one study, 35 participants; MD0.08, 95% CI −0.87 to 1.03; P = 0.16 (Analysis 3.1)), or fentanyl(change from baseline: one study, 10 participants, MD 0.20, 95%CI −2.50 to 2.90; P = 0.88 (Analysis 4.1); post-treatment score:one study, 12 participants; MD −0.40, 95% CI −2.76 to 1.96;P = 0.74 (Analysis 4.1)).

Condition

There were insufficient data to suggest opioids would be morebeneficial in any specific condition. The effect for COPD was asfollows: change from baseline: two studies, 23 participants, SMD−0.49, 95% CI −1.08 to 0.10; P = 0.1; post-treatment scores:eight studies, 131 participants; SMD −0.24; 95% CI −0.48 to0.01; P = 0.1, (Analysis 5.1). For cancer-related dyspnoea it was:change from baseline: three studies, 39 participants, SMD −0.41;95% CI −0.89 to 0.06; P = 0.21; post-treatment score: one study,10 participants, SMD −0.75; 95% CI −1.67 to 0.16; P = 0.11(Analysis 5.1). There was no significant difference overall for heartfailure (change from baseline: one study, 35 participants; SMD

21Opioids for the palliation of refractory breathlessness in adults with advanced disease and terminal illness (Review)

Copyright © 2019 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.


• オピオイドは「Breathlessness」に対し、効果がある。

Analysis 2.1. Comparison 2 Opioids versus placebo, Outcome 1 Breathlessness.

Review: Opioids for the palliation of refractory breathlessness in adults with advanced disease and terminal illness

Comparison: 2 Opioids versus placebo

Outcome: 1 Breathlessness

Study or subgroup Opioids Placebo

Std.Mean

Difference Weight

Std.Mean

Difference

N Mean(SD) N Mean(SD) IV,Fixed,95% CI IV,Fixed,95% CI

1 Morphine: change from baseline

Mazzocato 1999 9 -25 (10) 9 0.6 (7.7) 6.8 % -2.73 [ -4.10, -1.36 ]

Light 1996 7 -0.14 (0.09) 7 0.29 (0.9) 10.8 % -0.63 [ -1.71, 0.45 ]

Poole 1998 16 -2.5 (5.68) 16 -0.44 (3.4) 25.8 % -0.43 [ -1.13, 0.27 ]

Oxberry 2011 35 -0.41 (2.51) 35 -1.37 (1.86) 56.6 % 0.43 [ -0.04, 0.90 ]

Subtotal (95% CI) 67 67 100.0 % -0.12 [ -0.48, 0.24 ]Heterogeneity: Chi2 = 20.75, df = 3 (P = 0.00012); I2 =86%

Test for overall effect: Z = 0.67 (P = 0.50)

2 Morphine: post-treatment score

Bruera 1993 10 16 (18) 10 35 (29) 10.0 % -0.75 [ -1.67, 0.16 ]

Leung 1996 10 8.7 (2.2) 10 9.7 (1) 10.4 % -0.56 [ -1.46, 0.34 ]

Abernethy 2003 38 40.3 (23) 38 49.9 (24) 40.5 % -0.40 [ -0.86, 0.05 ]

Jankleson 1997 16 4 (2) 16 4.5 (2.6) 17.3 % -0.21 [ -0.91, 0.48 ]

Harris-Eze 1995 6 6.3 (2.7) 6 6.5 (3.5) 6.5 % -0.06 [ -1.19, 1.07 ]

Noseda 1997 14 -42 (27) 14 -43 (27) 15.2 % 0.04 [ -0.70, 0.78 ]

Subtotal (95% CI) 94 94 100.0 % -0.33 [ -0.62, -0.04 ]Heterogeneity: Chi2 = 2.45, df = 5 (P = 0.78); I2 =0.0%


3 Dihydrocodeine: post-treatment score

Chua 1997 12 2.91 (0.86) 12 3.6 (0.76) 21.1 % -0.82 [ -1.66, 0.02 ]

Johnson 1983 18 4.6 (2.1) 18 5.6 (2.3) 33.9 % -0.44 [ -1.11, 0.22 ]

Woodcock 1981 12 5.54 (1.91) 12 6.33 (2) 22.7 % -0.39 [ -1.20, 0.42 ]

Bar-Or 1982 11 -42.6 (12) 12 -42.5 (8.6) 22.2 % -0.01 [ -0.83, 0.81 ]

Subtotal (95% CI) 53 54 100.0 % -0.41 [ -0.80, -0.03 ]Heterogeneity: Chi2 = 1.85, df = 3 (P = 0.60); I2 =0.0%


Test for subgroup differences: Chi2 = 1.34, df = 2 (P = 0.51), I2 =0.0%

-4 -2 0 2 4

Favours opioids Favours placebo

78Opioids for the palliation of refractory breathlessness in adults with advanced disease and terminal illness (Review)


• モルヒネの「Breathlessness」に対する効果は、暫定的である。• ベースラインからの変化で評価したものでは、有効性が示されなかったが、治療後のスコアで評価したものは、有効性が示された。


Opioidについて（3）

• 明確な科学的証拠は乏しいが、いくつかの国際ガイドラインでは、呼吸困難の管理におけるオピオイドの使用が推奨されている。

• そのいくつかは、使用すべきプロトコルおよび投与量についてのガイダンスも提供している。

ResultsA total of 151 citations were retrieved using the MedicalSubject Headings previously listed, including 134citations on PubMed and 17 on The Cochrane Databaseof Systematic Reviews.

Opioids

Mechanisms of dyspnea relief by opioids include theaction on central receptors in the right posteriorcingulate gyrus. At lower doses this mechanism does notnecessarily alter respiratory drive, but respiratorydepression and reduction of consciousness level canoccur with larger doses.8

Thirteen out of 26 studies included in a recent Cochranereview9 on “opioids for the palliation of refractorydyspnea in adults with advanced disease and terminalillness,” assessed dyspnea during an exercise test,whereas the remaining studies, most of which had acuterespiratory failure or distress as exclusion criteria,examined patients with end-stage cancer or noncancerpatients. These studies used various medications(diamorphine, dihydrocodeine, morphine, andfentanyl), formulations (nebulized, subcutaneous, andoral), and schedules of administration (single dosevs multiple doses), limiting the conclusions orrecommendations that can be made. Nevertheless, theCochrane review concluded that some low-qualityevidence shows benefit for the use of oral or parenteralopioids to palliate dyspnea, but no evidence to supportthe use of nebulized opioids. This is in keeping with theCanadian guidelines10 on managing dyspnea in patientswith advanced COPD (grade of evidence 2C).

Only one study has dealt with the management ofdyspnea in patients with acute respiratory distress, withmost patients also presenting with acute respiratoryfailure (mean SaO2 under oxygen, 92%).11 Thisrandomized, placebo-controlled, and double-blind studyaimed to assess the effect of nebulized morphine (10 and20 mg) on dyspnea perceived at rest. Subjects benefitedequally from nebulized saline or morphine via a placeboor other nonspecific effect, and nebulized morphine hadno specific effect on dyspnea.

However, despite the lack of clear scientific evidence,several international guidelines recommend the use ofopioids in the management of dyspnea, and some ofthem also provide guidance for the protocol and dosagethat should be used.1,12-16 For example, an advisoryboard of intensivists has suggested that “low and slow”intravenous titration of an immediate-release opioid,

repeated every 15 min, should be provided until thepatient reaches an acceptable level of symptomcontrol.17 A continuous infusion may be substituted inthe case of persistent dyspnea (Table 2).

Why do we lack studies on opioids for acute respiratoryfailure? Opioids likely act both by depressingspontaneous respiratory drive and by modulatingcortical activity, as they do in pain. A justifiable concernamong clinicians is that opioids could thereby causeadverse events, including respiratory depression,confusion, and premature death in patients withrespiratory failure. However, in a Swedish nationalprospective study, treatment with lower dose opioidswas not associated with an increased risk of admission tohospital or death in patients who were oxygen-dependent, regardless of whether the patient was naiveto opioids or had hypercapnia.18

Nevertheless, half of Dutch chest physicians or residentsin respiratory medicine rarely or never prescribe opioidsfor refractory dyspnea because of the fear of respiratorydepression and adverse effects such as nausea andconstipation and resistance on the part of the patient.19

This concern on the part of some physicians about theblunting effect of morphine on ventilatory drive may beexcessive based on the greater relief of dyspnea

TABLE 2 ] Suggested Morphine Protocol for theManagement of Dyspnea in Opioid-NaivePatients Who Are Terminally Ill

! Start therapy with morphine at low doses (1-2 mg IV),titrate up every 15 min until desired effect is achieved.In patients with severe dyspnea, may titrate morerapidly.

! Monitor dyspnea, comfort, respiratory rate, andpatient’s response with sedation scale (RASS).

! In case of persistent dyspnea, switch to a continuousinfusion.

! It should be kept in mind that the correct dose andinterval for opioid administration in all patients arethose that relieve dyspnea without intolerable adverseeffects.

! Once the patient’s dyspnea has been controlled,maintain the effective basal infusion rate.

! If patients experience significant opioid-related sideeffects, such as nausea and/or vomiting, considerreducing the morphine dosage or changing to anequianalgesic dose of hydromorphone (5 mg morphineto 1 mg hydromorphone).

! Prevent opioid-associated constipation (laxatives ormu-opioid receptor inhibitors should be routinelyprescribed).

RASS ¼ Richmond Agitation-Sedation Scale. (Data from Marciniuk et al.10)


• 低用量（1〜2 mg、1〜4 mg）のモルヒネを開始し、効果が得られるまで15分ごとに投与。重度の呼吸困難患者ではより急速に調整してよい。

• 呼吸困難、快適さ、呼吸数、および患者の反応を鎮静スケール（RASS）で監視。

• 持続性呼吸困難の場合は、持続注入に切り替える。

• 全患者におけるオピオイド投与の正しい用量および間隔は、耐え難い副作用を伴わずに呼吸困難を軽減するものである。患者の呼吸困難が抑制されたら、有効基礎注入量を維持する。

• 吐き気や嘔吐などのオピオイド関連の重大な副作用がある場合は、モルヒネの投与量を減らすか、等モル濃度のヒドロモルフォン（5 mgモルヒネから1 mgヒドロモルフォン）に変更することを検討。

• オピオイド関連便秘を予防する（下剤またはmu-opioid受容体阻害薬は日常的に処方されるべきである）。

Benzodiazepinesについて

• ベンゾジアゼピンは、知覚の感度を変えることによって、呼吸困難の刺激に対する患者の反応を減らすことができる。

• 呼吸困難の抗不安薬療法に関する近のコクランレビューでは、8つの研究が含まれた。– いずれも対象に呼吸不全が含まれなかった。– 分析は、「Breathlessness」軽減のためのベンゾジアゼピンの有益性を示さなかった。

– モルヒネと比較して、癌患者の偶発的な息切れ（突発性呼吸困難）の予防において有意な効果は示されなかった。

– ベンゾジアゼピンはプラセボと比較し、有意に多くの有害事象（眠気と傾眠）が生じた。しかし、ベンゾジアゼピンとモルヒネを比較した場合には、有害事象はモルヒネより少なかった。

Eur Respir J. 2015;45(6):1681-1691. Thorax. 2010;65(1):21-26.

Cochrane Database Syst Rev. 2016;(10):CD007354.

Analysis 1.1. Comparison 1 Overall, Outcome 1 Placebo-controlled/cross-over design.

Review: Benzodiazepines for the relief of breathlessness in advanced malignant and non-malignant diseases in adults

Comparison: 1 Overall

Outcome: 1 Placebo-controlled/cross-over design

Study or subgroup Benzodiazepine Placebo

Std.Mean

Difference Weight

Std.Mean

Difference

N Mean(SD) N Mean(SD) IV,Random,95% CI IV,Random,95% CI

Harrison (unpublished) 17 44.49 (25.06) 17 45.94 (25.74) 22.1 % -0.06 [ -0.73, 0.62 ]

Man 1986 24 3 (0.8) 24 3 (0.8) 31.2 % 0.0 [ -0.57, 0.57 ]

Shivaram 1989 8 3.63 (1.92) 8 3 (1.6) 10.2 % 0.34 [ -0.65, 1.33 ]

Stege 2010 14 4.2 (2.9) 14 4.1 (2.5) 18.2 % 0.04 [ -0.71, 0.78 ]

Woodcock 1981 15 3.33 (1.11) 15 4.07 (0.88) 18.2 % -0.72 [ -1.46, 0.02 ]

Total (95% CI) 78 78 100.0 % -0.10 [ -0.42, 0.21 ]Heterogeneity: Tau2 = 0.0; Chi2 = 3.69, df = 4 (P = 0.45); I2 =0.0%


Test for subgroup differences: Not applicable

-2 -1 0 1 2

Favours benzodiazepine Favours placebo

51Benzodiazepines for the relief of breathlessness in advanced malignant and non-malignant diseases in adults (Review)


Cochrane Database Syst Rev. 2016;(10):CD007354.

• ベンソジアゼピンの「Breathlessness」に対する有効性は、示されなかった。

Benzodiazepinesについて（２）• オピオイドと鎮静剤の併用は、終末期に呼吸困難や不安を治療するために一般的に使用されている。

• 併用療法（オピオイドとロラゼパム）を検討した研究– 重度の呼吸困難および関連不安を伴う低酸素症の患者（22/26例、SaO2 <90％）において、呼吸困難および呼吸数が有意に減少した。

– ガス交換に影響を及ぼさなかった。

• この組み合わせに関して、寿命を縮めるかもしれないという懸念がしばしば提起される。しかし、治療の開始あるいはオピオイドや鎮静薬の投与量の増加が、実際に死を早めるという確固たる証拠はない。

Support Care Cancer. 2011;19(12):2027-2033.

Lancet Oncol. 2003;4(5):312-318.

Benzodiazepinesについて（3）

• ベンゾジアゼピンおよび他の抗不安薬は、特に非常に苦痛を感じ不安を抱いている患者において、オピオイド

および他の非薬理学的治療が、「Breathlessness息切れ」に対して有効でない場合、第二選択または第三選

択治療と考えられる。

Corticosteroidsについて

• コルチコステロイドの使用の理論的根拠は、肺機能障害の原因となる気道の炎症および浮腫を減少させることである。

• 観察研究では、食欲不振、悪心、疼痛、気分の低さ、嘔吐、脱力のほとんどの患者でステロイドで症状スコアが改善されたが、呼吸困難は改善しなかったことが示された。

Palliat Med. 2001;15(1):3-8.

Cor8costeroidsについて（２）

• 重症の呼吸困難患者におけるコルチコステロイドの効果に関する唯一のランダム化比較試験

– 癌患者を対象– 呼吸困難の数値評価尺度スコアはデキサメタゾン群では有意に低下したがプラセボ群では低下しなかった。

• 多くの医師がコルチコステロイドは、終末期患者の呼吸困難を治療するのに有用であると認識しているが、呼吸困難の緩和において使用を支持する証拠は非常に弱い。

J Pain Symptom Manage. 2016;52(1):8-16.e1.

Diuretics（利尿薬）について

• ループ利尿薬の静注使用は、末期心不全およびリンパ管癌腫症を伴う呼吸困難の患者の肺鬱血を減らすのに役立つかもしれない。

• 噴霧フロセミドが呼吸困難を改善するための新規アプローチとして研究されてきた。

– 正確な作用機序はまだ知られていない。– 動物モデルおよび試験管内モデルでは、平滑筋のコリン作動性、非コリン作動性、ノルアドレナリン作動性収縮に対する保護効果、肺伸張受容体の活性化および迷走神経刺激受容体の阻害を含む複数の機構の関与を示唆する。

• 噴霧フロセミドによる有益性の証拠はまだ示されていない。• 末期患者の呼吸困難の管理のための噴霧利尿薬の使用は、まだ物議をかもしているが推奨できない

J Pain Symptom Manage. 2008;36(4): 424-441.

Am J Respir Crit Care Med. 2000;161(6):1963-1967.

Oxygen Therapyについて1) 鼻カニューレ、ベンチュリ、または非再呼吸マスクによる標準酸素の使用

• 癌関連呼吸困難の緩和における標準酸素の効果– 合計179人の患者を含む6つの研究を含むメタアナリシス– すべての研究は二重盲検無作為化試験で行われた。（酸素対空気ボンベ、ある試験では酸素はヘリウム濃縮空気または医療用空気と比較）。

– 酸素投与の期間は通常非常に短く（約1時間）、呼吸困難を改善しなかった。

• 末期がんおよび呼吸困難の患者における酸素使用を支持する十分な根拠はない。

Acta Oncol. 2012;51(8):996-1008.

Oxygen Therapyについて（２）2) ①HFNT（high-flow nasal therapy）• 呼吸困難および主に進行悪性腫瘍に関連した急性呼吸不全を有する183人の患者を無作為に選択– 患者は血液を含むさまざまな悪性腫瘍（29％）、肺（17％）、消化管（15％）、肉腫（6％）、頭頸部CNS（5％）、乳房（4％）、その他の腫瘍（24％）

– 対象者の78％が入院中に死亡。– HFNTの使用中、快適性および呼吸困難を改善し（41％）、あるいは安定したまま（44％）であった。悪化したのは15％。

– HFNTの中央時間 3日。

• 標準的な酸素療法と比較して、HFNTは快適さおよび呼吸困難の軽減を提供することにおいてより機能し得るが、HFNTの可能性を確立するためにはさらに対照試験が必要である。

J Palliat Med. 2011;14(7):835-839.

Oxygen Therapyについて（3）

2) ②NIV（non-invasive ventilation）• 呼吸困難の感覚は吸気負荷と強く相関している。そのためＮＩＶは、末期疾患患者における呼吸困難の感覚を軽減するためにも有用である。

• 分泌物除去および気道を保護することができない点は注意が必要である。

• NIVヨーロッパ呼吸器学会/アメリカ胸部学会特別調査委員会は、NIVの使用が通常の慣習ではない場合には、適切な患者の選択とスタッフの訓練の必要性を強調している。

Non-pharmacologic Treatments（非薬物療法）について

• 2008年に、コクランレビュー72文献は、呼吸困難を治療するために非薬理学的介入を評価した。

– 歩行補助具や太極拳：これらのアプローチは呼吸困難のある患者に適用するのは難しい。

– 呼吸法：通常数週間または数ヶ月にわたって行われ、呼吸困難に対する一貫した好ましい効果はなく、呼吸の仕事を増加させることさえあるため末期症状の患者に役立つ可能性は低い。

– 鍼治療：近のレビューでは、臨床的に重要な改善を見出した。

– シータミュージック：シータミュージックを用いたガイド画像が呼吸困難患者の緩和に役立つ介入であることが示された。深いリラクゼーション効果。

J Pain Manage. 2019 Apr 17• 目的

– 呼吸困難の治療のためのファン療法の有効性を評価すること• デザイン

– システマティックレビュー• 対象

– 1946年1月1日〜2018年9月31日までに発表されたファン療法研究。• 方法

– Medline, EMBASE, Web of Science, Scopus, CINAHL, PsycInfo, and Cochrane Library の電子データベースで検索。

– 検索用語は、‘‘dyspnea,’’ ‘‘dysponea,’’ ‘‘dyspneic,’’ ‘‘short of breath,’’ ‘‘shortness of breath,’’ ‘‘breathless,’’ ‘‘breathlessness,’’ ‘‘breathing difficulty,’’ ‘‘labored breathing,’’ and ‘‘fan.’’

– “fan”は、handheld or electric fan と定義された。• 結果

– 92論文がリクルートされ10論文が選択基準を満たした。

Fan Therapy for the Treatment of Dyspnea in Adults: A Systematic Review

dyspnea assessment varied; five studies used a numericrating scale, one used a visual analog scale, and fourused the Borg scale or modified Borg scale.

The risk of bias for all included studies is summa-rized in Table 2. Seven of the nine included random-ized controlled trials had sufficient information todetermine the risk of allocation bias as low risk, theother two randomized controlled trials did not havesufficient information.11,12 Although there are eightstudies stated single blinding, the blinding processwas not complete or verified from the methods pro-vided. There was no blinding in other two studies, acohort13 and a randomized controlled trial.14 All 10studies were assessed as high risk of blinding bias.All the studies addressed withdrawals and incomplete

data, and reported the selective outcomes, and weretherefore judged as low risk of bias. All the studies,apart from one, a cohort design, were free from otherbias and were judged as low risk, whereas the cohortstudy as high risk.Six of the 10 included studies determined that the

use of a fan was associated with a clinical benefit(Table 3). Galbraith et al.6 and Kako et al.15 bothshowed that a handheld fan directed at the facereduced the sensation of breathless compared with afan directed at the leg (P < 0.05). In another study,Kako et al.16 reported relief of dyspnea in patients us-ing the fan to their faces (P ¼ 0.02), and also foundthat one hour was not a sufficient washout periodfor fan therapy.Puspawati et al.17 compared the use of diaphragmatic

breathing exercise with airflow from a fan and the use ofdiaphragmatic breathing exercises alone in 21 hospital-ized lung cancer patients. They concluded that theairflow stimulation from a handheld fan decreased thedyspnea sensation and breathing frequency in nonhy-poxemic lung cancer patients (P ¼ 0.001).Marchetti et al.18 confirmed the efficacy of fan

therapy in 10 outpatients with chronic obstructivepulmonary disease by comparing the fan-to-face andfan-to-leg approaches. The fan-to-face technique re-sulted in improved exercise performance (P ¼ 0.03).Wong et al.12 evaluated 30 Chinese inpatients in

whom advanced cancer was a principal cause of dys-pnea. They applied the symptom scale with fan andaccompanied caregivers. They found that fan therapywas effective in alleviating dyspnea in this population(P ¼ 0.01).Four studies15e18 failed to demonstrate significant

alleviation of dyspnea with the use of fan therapy. Bau-sewein et al.19 compared the use of fan therapy with a

Records identified through database searching (n=95)

Records screened (n=92)

Duplicates removed (n=3)

Full-text articles assessed for eligibility (n=18)

Studies included in qualitative synthesis (n=10)

Full-text articles excluded (n=8)• No fan intervention (n=2)• No dyspnea assessment (n=6)

Records excluded (n=74)• Review & guideline (n=39)• No fan intervention (n=24)• Case report (n=8)• Pediatric (n=3)

Fig. 1. Study flow chart.

Table 1Characteristics of Included Studies

Study (Reference) N SettingStudyDesign Blinding

Cancer-RelatedDyspnea

COPD-RelatedDyspnea Age, yrs Fan Therapy

Duration,minutes

KPS

n (%) n (%) Mean (SD) Mean (SD)

Bausewein 201019 70 Hospital RCT Single blind 25 (36) 45 (64) 66 (9) NR 62 (11)Booth 201613 31 Hospital/hospice Single

cohortOpen label 12 (39) 7 (23) 75 (11) 5 NR

Galbraith 20106 50 Hospital RCT Single blind 11 (22) 26 (52) 71 (33e90) 5 NRJohnson 201614 49 Clinic/hospice RCT Open label 13 (27) 23 (47) 68 (12)a NR 70 (7)Kako 2018a15 9 Hospital RCT Single blind 7 (78) 2 (22) 77 (9) 5 3b

Kako 2018b16 40 Hospital RCT Single blind 40 (100) 0 (0) 71 (8)c 5 42 (11)d

Marchetti 201518 10 Clinic RCT Single blind 0 (0) 10 (100) 62 (6) NR NRO’Driscoll 201111 34 Hospital/clinic RCT Single blind 0 (0) 34 (100) 67 (9) NR NRPuspawati 201717 21 Hospital RCT Single blind 21 (100) 0 (0) 54 (9) 5 NRWong 201712 30 Hospital RCT Single blind 30 (100) 0 (0) NR 5 NRTotal 344 d d d 159 (46) 147 (43) d d d

COPD ¼ chronic obstructive pulmonary disease; KPS ¼ Karnofsky performance status; RCT ¼ randomized controlled trial; NR ¼ not reported; ECOG ¼ EasternCooperative Oncology Group.aThe mean age of the control group was 68 years (SD ¼ nine years).b(ECOG) performance status was used.cThe mean age of the control group was 67 years (SD ¼ 12 years).dThe mean KPS of the control group was 43 (SD ¼ 7).

Vol. - No. - - 2019 3Fan Therapy for the Treatment of Dyspnea

• 10件のうち９件がランダム化比較試験、1件が単一コホート試験• 約80%は入院患者• 対象344名中、46%が、呼吸困難の主な原因が癌だった。• ファン療法の介入期間は5分（期間が報告されている研究のみ）• 呼吸困難の評価方法：5件はNRS、1件はVAS、4件はBorg スケールまたは修正

Borgスケール。

This review has intrinsic limitations, notably thelimited number of studies and the limited quality ofthose studies (e.g., they did not use pharmacologic stra-tegies or assess the etiology of dyspnea).Moreover, thereare some limitations in the design of these randomizedcontrolled trials. Moreover, the potential limitation ofvote counting on the number of studies could also bea limitation of a systematic review with small studiesbias.23 There are two clinical trials ongoing on similartopic and hopefully will provide more evidence on therole of fan therapy in the management of dyspnea.

Two important areas of future research includemulti-center studies with larger numbers of patients, whichmay lead to a better assessment of the effectiveness offan therapy; a better characterization of the cause of dys-pnea (e.g., cancer, heart failure, or restrictive or obstruc-tive pulmonary disease); and an appropriate blinding ofthe intervention. Because of the subjective nature of dys-pnea, the lack of existing double-blinded studies is of

concern, and such studies would be an important contri-bution to this research area.

ConclusionLimited direct evidence from randomized clinical tri-

als indicates that fan therapy is effective at alleviatingdys-pnea. Additional trials are warranted to confirm thisfinding and explore the use of fan therapy for the treat-ment of dyspnea in more diverse populations and set-tings. More research is needed to better characterizethe role of fan therapy in the management of dyspnea.

Disclosures and AcknowledgmentsThe authors thank the Department of Scientific

Publication at the University of Texas MD AndersonCancer Center for editorial assistance.

Table 3Summary of Study Findings

Study (Reference) andIntervention

Absolute Change in Dyspnea

ConclusionMean (SD/Range) P-value

Bausewein 201019

Fan !0.6 (2.1)a 0.9 The effectiveness of a fan at relieving dyspnea could notbe establishedA placebo wristband !0.8 (2.67)a

Booth 201613

Fan !1.04b N/A 50% of patients with dyspnea at rest experienced relief using ahandheld fan

Galbraith 20106

Fan to face !7 (1.5 to 14.5)c 0.003 A handheld fan to the face reduced dyspneaFan to leg !1.5 (!2 to 7)c

Johnson 201614

Fan !6 (2)b 0.853 Confirmed the feasibility of a definitive multisite trial to study theuse of a handheld fan as part of the self-management of chronicdyspnea

No intervention !5 (4)b

Kako 2018a15

No fan 0b 0.02 A fan to the face resulted in relief of dyspneaFan to leg 0b

Fan to face !0.7b

Kako 2018b16

Fan to face !1.35 (!1.85 to !0.84)b 0.001 A fan to the face was effective at alleviating dyspnea in terminally illcancer patientsFan to leg !0.1 (!0.53 to 0.33)b

Marchetti 201518

Fan to face 5 (0e10)a 0.03 A fan to the face improved exercise performance in patients with COPDFan to leg 6.5 (0e10)a

O’Driscoll 201111

Room air 5.1 (1.7)a N/A A fan had no apparent physical or placebo effectElectric fan 5.1 (1.7)a

Air mask 5.3 (1.6)a

Oxygen mask 5.1 (1.7)a

Puspawati 201717

Fan with diaphragmaticbreath exercise

!1.21 (0.56)a 0.003 A handheld fan decreased dyspnea in non-hypoxemic, dyspneic lungcancer patients

Diaphragmatic breathingexercise

!0.69 (0.46)a

Wong 201712

Fan !1.53b 0.01 Fan therapy was effective at alleviating dyspnea in Chinese patients withadvanced cancerAccompanied by

caregivers with no fan!0.13b

COPD ¼ chronic obstructive pulmonary disease; N/A ¼ not applicable.aModified Borg Scale/Borg Scale.bNumeric rating scale.cVisual Analog Scale.

Vol. - No. - - 2019 5Fan Therapy for the Treatment of Dyspnea

• 9件のRCTのうち6件で、有意にファン療法は効果的であるとの結果。• Fan-to-faceの方法がも効果的だった。（fan-to-legに比べて）• ファン療法前後の皮膚温の変化を報告した研究は1件のみ。• ファン療法の期間の影響を調査した研究はない。

結論：限られた研究結果からは、ファン療法が呼吸困難を効果的に軽減する可能性があることを示している。

Palliative care in the ICU: relief of pain, dyspnea, and thirst—A report from the IPAL-ICU Advisory Board

• The Improving Pallia8ve Care in the ICU (IPAL-ICU) Project,（ICUにおける緩和ケアの改善プロジェクト）

– 国立衛生研究所および緩和ケア推進センターにより支援受ける。– 重症発症から緩和ケアと集中治療をうまく統合するための技術支援、証拠、およびツールを共有する。

– 緩和だけでなく積極的治療受けている、すべてのICU患者を含む。– 学際的なIPAL-ICU諮問委員会が、緩和ケアと集中治療における各メンバーの専門知識を組み合わせて、重症疾患時の疼痛、呼吸困難、渇きの評価と管理における課題に取り組んだ。

• 以下の質問に焦点を当てた。1) ICUにおけるこれら3つの一般的な症状を評価するために必要な重

要な要素は何ですか？

2) 重篤な病気の間にこれらの症状を管理するための適な戦略は何ですか？

3) 症状のケアを改善のためにどのように体系化できるか

Palliative care in the ICU: relief of pain, dyspnea, and thirst—A report from the IPAL-ICU Advisory Board

Kathleen Puntillo,University of California at San Francisco, San Francisco, CA, USA

Judith Eve Nelson,Icahn School of Medicine at Mount Sinai, New York, NY, USA

David Weissman,Medical College of Wisconsin, Milwaukee, WI, USA

Randall Curtis,University of Washington, Seattle, WA, USA

Stefanie Weiss,Icahn School of Medicine at Mount Sinai, New York, NY, USA

Jennifer Frontera,Cleveland Clinic, Cleveland, OH, USA

Michelle Gabriel,Veteran’s Administration Palo Alto, Palo Alto, CA, USA

Ross Hays,University of Washington, Seattle, WA, USA

Dana Lustbader,North Shore-Long Island Jewish Health System, Hyde Park, NY, USA

Anne Mosenthal,University Medical and Dental of New Jersey, Newark, NJ, USA

Colleen Mulkerin,Hartford Hospital, Hartford, CT, USA

Daniel Ray,Lehigh Valley Health Network, Allentown, PA, USA

Rick Bassett,St. Luke’s Hospital, Boise, ID, USA

Renee Boss,Johns Hopkins University School of Medicine, Baltimore, MD, USA

Karen Brasel, and

Correspondence to: Judith Eve Nelson.Conflicts of interest On behalf of all authors, the corresponding author states that there is no conflict of interest.

HHS Public AccessAuthor manuscriptIntensive Care Med. Author manuscript; available in PMC 2017 May 12.

Published in final edited form as:Intensive Care Med. 2014 February ; 40(2): 235–248. doi:10.1007/s00134-013-3153-z.

Author Manuscript

Author Manuscript

Author Manuscript

Author Manuscript

• 多くの患者は症状をリアルタイムに自分で訴えられないことが多い。

• 重症疾患の診断および治療の介入は、それ自体が症状の悪化の原因ともなり、体位変換など日常の非侵襲的

な患者ケアも同様である。

• 症状を軽減するための治療は、複数の臓器不全に伴う複雑な薬理学的および生理学的問題を考慮に入れなけ

ればならない。

• 同時に、症状の苦痛の軽減は、緩和ケアの基本的な側面であり、予後にかかわらず入院時から、すべてのICU患者にとって包括的なクリティカルケアの不可欠な要素

である。Crit Care Med. 2010; 38:808–818.

Author Manuscript

Author Manuscript

Author Manuscript

Author Manuscript


Table 1

Symptom assessment scales and their feasibility for self-reporting ICU patients

Type of instrument Description Comments

Self-report—unidimensionalpain, thirst, or dyspnea scale

0–10 numeric rating scale (NRS) Higher numbers indicate greater pain (or dyspnea, orthirst) intensity. Example: 0 = no pain and10 = worst possible pain. NRS has both construct[98, 99] and concurrent validity [98, 100]A 0–10 horizontal, visually enlarged laminated NRS forpain is most feasible and valid for ICU patients [22]

Self-report—unidimensionalpain or dyspnea scale

0–10 cm visual analog scale (VAS) A 10-cm line, with anchors at either end. Example: oneend is marked “no pain” and the other end is marked“pain as bad as it could be” or “the worst imaginablepain”. The patient makes a vertical line through thehorizontal 10-cm line to indicate his or her painintensity. The clinician then measures from the left ofthe horizontal line to the patient’s vertical line with aruler and assigns a score. Difficult for many patientsto understand and use

Self-report body outlinediagram pain scale

Identifies pain location. Diagram of frontand back of a “neutral” body dividedinto body sections

Patient is able to point to, or mark, the place or places onthe diagram where pain is felt. Has been used withICU patients [19]

Self-report—multidimensionalpain scale

McGill Pain Questionnaire-Short Form(MPQ-SF)

Scale with 11 sensory (such as sharp or throbbing) andfour affective (such as fearful or sickening) words.MPQ-SF also has a VAS and a verbal scale. The15-word scale has been used in ICU patients [19]

Self-report—multiplesymptoms scale

Condensed Form of the MemorialSymptom Assessment Scale—MSAS[24]

Measures a diverse group of symptoms. Has been usedwith critically ill patients [101]


Edmonton Symptom Assessment Scale[25]

Measures a diverse group of symptoms. Has been usedwith critically ill patients [12]


10-item symptom assessment scale Modification of MSAS and Edmonton. Validated in ICUpatients [13]

Intensive Care Med. Author manuscript; available in PMC 2017 May 12.

• NRSは呼吸困難を評価するために頻繁に使用されてきた。• VASも使用されるが、ICU患者には理解して使用するのは難しいかもしれない。

Dyspnea緩和のための介入

Author Manuscript

Author Manuscript

Author Manuscript

Author Manuscript


Table 3

Dyspnea interventions, mode of action, and rationale

Intervention Dose Mode of action Rationale

Optimal positioning, usuallyupright with arms elevated andsupported [62, 63]

Whenever patient reports dyspneaor displays respiratory distress

Increases pulmonary volumecapacity

Increases air exchange whichmay improve oxygenationand carbon dioxideclearance and reduceinspiratory effort

Balance rest with activitySpace nursing care

Guided by dyspnea/respiratorydistress

Decreases excessive oxygenconsumption

Prevents hypoxemia

Oxygen as indicated by goals oftherapy; not useful innormoxemia or when the patientis near death and in no distress[8, 68, 69]

Variable, guided by goals oftherapy and patientcharacteristics

Improves the partial pressure ofoxygen; reduces lacticacidemia

Treats hypoxemia

Cold cloth on face [108] As needed Trigeminal nerve stimulation;action on dyspnea unknown

Anecdotal reports of patientrelief; inexpensive; easy toperform

Opioids, such as morphine orfentanyl [54]

Low doses titrated to the patient’sreport of dyspnea or display ofdyspnea behaviors is effective;oral or parenteral; no evidence tosupport inhaled administration;no evidence on dosing regimens

Uncertain direct effect;reduced brainstem sensitivityto oxygen and carbondioxide; altered centralnervous perception

Strong evidence-base supportseffectiveness

Benzodiazepines, such aslorazepam or midazolam [73]

Low doses titrated to the patient’sreport of dyspnea or display ofdyspnea behaviors; no evidencefor benzodiazepine regimens

Anxiolysis Fear or anxiety oftenaccompanies dyspnea

Intensive Care Med. Author manuscript; available in PMC 2017 May 12.

三叉神経刺激容易に実行できる冷たいタオルで顔を拭く患者の希望時

適な姿勢の調整：腕を広げ支える

肺容量の増加

休息と活動のバランス

過剰な酸素消費を減らす

治療目標に合わせた酸素投与

オピオイド投与（モルヒネ、フェンタニル）

ベンゾジアゼピン投与（ロラぜパム、ミダゾラム）

酸素分圧の改善、乳酸血症の減少

直接的効果は不確か：酸素、二酸化炭素に対する脳幹の感受性の低下：

中枢神経の知覚の変化

不安の軽減

私見

• 痛みは術後患者も多いことから注目しがちだが、痛み以上に「呼吸困難」「呼吸苦」が発生していることを改めて学んだ。

• ICUでは呼吸困難を適切に訴えられない患者も多く、客観的評価であるIC-RDOSは有効そうである。

• 「呼吸苦」に対する症状マネージメントとしての薬物治療はエビデンスとして確立されていない部分もあるが、まずは患者の「呼吸困難」に気づき評価することが重要であると感じた。

• Fan療法、冷たいタオルで顔の清拭などの非薬物療法はすぐに臨床に取り入れられそうである。

icu患者の呼吸苦慈恵icu勉強会 2019/08/06 山口庸子 本日の内容 •dyspneaの定義...

Documents

icu患者の呼吸苦慈恵icu勉強会 2019/08/06 山口庸子本日の内容 •dyspneaの定義...