第六章 結論與建議
第二節、 建議
一、呼吸照護病房之經營及醫護團隊
1.建立呼吸照護病房的監測指標
在逐步羅吉斯迴歸中，預測模型之六大因素中，住院月數 及使用呼吸器月數等因素，都無法加以調整。除此之外，人 工氣道別中，各醫護團隊要儘量和病患家屬溝通，儘早使用 氣管瘻管作為人工氣道，而非勉強使用插管。其次腎功能中 Creatinine 的監測，儘可能不要使病患因如泌尿道感染、糖尿 病監控不當…等因素而造成 Creatinine 值的上昇。而白血球數 的監控同樣應作為呼吸器依賴患者的常規監測指標，不論是 否因呼吸道的感染，或者是免疫的障礙，對於抗生素即時、
適當的使用，對於呼吸器依賴患者的預後，都有深遠的影響。
2.重新訂定醫療及護理照護流程
呼吸器依賴病患皮膚感染的有無，是預測其死亡率的重要 指標。而皮膚的感染及褥瘡的有無，本即是護理照護品質重 要的參考，而護理品質的良窳又和其他系統感染、餵食、更 換導管、洗浴等有密切關係。皮膚感染的預防、早期診斷、
適當治療，又需呼吸照護團隊以外專科人員的協助，因此它 的品質維持端賴一項周延的醫護計劃，跨科間合作及護理紀 律的落實。
3.建立呼吸器依賴病患之風險評估
各醫院呼吸照護病房之醫療團隊，可依照本研究之死亡率 預測模型對所照護病患之風險及危險性評估，作為健保局協 商以及和家屬溝通的重要依據。
二、衛生政策主管機關
1.訂定呼吸照護病房之品質監控指標
目前全國呼吸器依賴病患已逾 12000 餘人，每位病患均須 長期佔用急性醫院中之病床，耗用醫療資源之鉅，甚為可觀。
但現行制度中之稽核中，並無一致的標準。對於整合試辦計 劃中，呼吸照護病房之呼吸器依賴患者，已經歷多次呼吸器 脫離失敗之患者，仍只在乎其呼吸器脫離率的多寡。至今無 一明確指標系統，作為品質監控的依據。此對病患醫療人權、
醫療品質等，皆有莫大的缺失。
本研究中，曾以表十六回顧整理近二十年相關文獻及報告 中，得到許多在個別研究中具有和呼吸器依賴病患之脫離呼 吸器影響因素、死亡率影響因素及具有重症、慢性和呼吸衰 竭等相關之危險因子等，主管單位可比照 TQIP(Taiwan Quality indicator program)或 THIS(Taiwan Hospital indicator system)模 式，要求各呼吸照護團隊提供完整及全面的資料，依據本研 究之方法，取得更精確及代表性的指標，建立呼吸照護品質 之監測系統。
2.建立疾病嚴重度評估系統及支付標準
透過各危險因子的鑑別，以統計方法確立疾病嚴重度的評 估，作為呼吸器依賴病人的疾病嚴重度分類。再依據各類嚴 重度的治療需求，計算出所需費用及成本，作為各分類等級 給付之依據，如此可矯正現行齊頭式的支付方法，所造成扭 曲之醫療行為(如不願收治較嚴重患者)，亦可節省不必要的醫 療支出。
三、未來研究者
1. 本研究資料收集屬回溯性資料，對於某些資料，因無事前要
求，而有所闕漏。未來研究者，可改採前進式資料收集，即 要求各呼吸照護病房對某些項目之記錄特別留意，如此可蒐 集更完善之資料。
2. 本研究僅對回溯性資料，加以分析研究，未來研究者可將此 預測模型，以現行住院病患之資料，預測其未來一個月之死 亡率，以校正本研究結果之準確度。
3. 本研究以 30 日死亡率作為探討的依變項，未來研究者可依 需求，將其死亡率預測時間改為 7 日、10 日、15 日等不同 時間，則其將分別代表不同之意義。
4. 本研究僅以鑑別其危險因子為目的，未來研究者可依此再建 立疾病嚴重評估系統，現行呼吸器依賴病患嚴重等級的分類 依據。
5. 本研究未以成本及財務為研究目的，未來研究者可依成本及 各等級嚴重度病患之基本醫療需求，計算出所需費用，提供 健保局及相關業者間財務協商之重要依據。
6. 未來研究者可對整體呼吸照護體系對於總額給付下合理的
7. 本研究醫院，均為加入整合照護計劃(IDS)之醫院，無法區 別是否加入整合照護計劃，對死亡率或醫療品質是否有所差 異，未來研究者可探討整合照護計劃，對呼吸器依賴患者的 影響(分層研究)。
表一 Outline of Ventilator Classification System
Input Electric Pneumatic Control scheme
Control variables Pressure
Volume Flow Time
Phase variables Trigger Limit Cycle Baseline
Conditional variables Modes of ventilation Control subsystems
Control circuit Drive mechanism Output control vale Output
表二 試辦計畫中各階段適用之病患、定義、照護病房及支付方式
呼 吸 困 難 的 感 覺。
（陳金淵：呼吸器依賴患者試辦管理式照護成效評估,p4,2002）
表三 89 年 7-12 月參與試辦計畫個案每人次之實際費用與支付費 用
比較(RCW)
醫院層級 費用狀況 人次 平均住 院天數
平均住 院費用
平均申 請金額
平均部 分負擔 醫學中心 定額費用<實際費用 20 16.5 101,868 74,144 1,171
定額費用>實際費用 16 23.1 71,877 77,285 - 區域醫院
定額費用<實際費用 90 27.6 113,729 95,583 77 定額費用>實際費用 26 55.9 186,819 196,724 93 地區醫院
定額費用<實際費用 180 50.4 206,822 182,065 236 定額費用>實際費用 42 43.4 143,031 151,223 58 合計 定額費用<實際費用 290 41.0 170,693 147,783 251
總計 332 41.3 167,194 148,218 227
（陳金淵：呼吸器依賴患者試辦管理式照護成效評估,p64,2002）
表四 各照護階段適用支付標準及規定
護日數≦
第四階段
(1)呼吸治療專業人員訪視
七、申報 P1013C 或 P1014C 者，若經查明各陽人員訪視
房，回轉加護病房者須符合 全民健康保險醫療費用支 付標準加護病床使用適應 症，不符合適應症而回轉 者，以 P1010C 或 P1012C 支付。
十、回轉加護病房照護者，
病情好轉，應回第三階段或 第四階段，回第三階段者，
以 P1010C 或 P1012C 支付。
十一、其他未盡事宜，依全 民健康保險居家照護作業 要點辦理。
（全民健康保險呼吸器依賴患者整合性照護前瞻性支付方式試辦計劃第三版,2002）
表五 Case Mix
Case Mix
Case-mix can be used to:
˙allocate or justify resources used
˙ as part of quality improvement through evaluating the outcome observed.
For resources, case-mix needs to consider separately:
˙ resources needed to provide care, maintaining the patient’s safety and life
˙resources needed to provide treatment, interventions that modify disability.
For evaluating outcome, case-mix needs to consider:
˙natural history/prognosis in absence of any treatment
˙prognosis in terms of probability of responding to treatment
˙prognosis/natural history as a result of the intervention in those who respond
˙influence of contextual factors.
It is recommended that rehabilitation services:
˙use dependency measures to record resources devoted to care
˙use process measures to improve quality
˙avoid using case-mix in view of the multiple factors that need to be taken into account.
(Clinical Rehabilitation pp183-185,1999)
表六 How to establish clinical score value Symptoms History
Operation Sex Oxygenation Diagnostics
Age Laboratory Values Cardio-vascular
parameters
Drugs Preexisting
Disease(s) Malignancy
Patient A Patient F Patient K Patient C Patient H
Patient B Patient E Patient L
Reduction
B
Different clinical situations
Score value X
Reduction
A
Complex clinical parameters
Score value X
(Langenback’s Arch Sung p56,2002)
表七 Steps in the development of a score system Selection of
Parameters
Point weights and composition
Data
base Score
Observation and Documentation
Depends on the entity to be described:
• mortality
• morbidity
• severity
• organ function And on method used:
• experts
• statistical procedures
Application and Validation
(Langenback’s Arch Sung p57,2002)
表八 Meaning and Computation of Admission Scores
Score Meaning of Score Computation of Score 0 No significant findings All other, such as cases with
multiple group 0 KCFs 1 Minimal finding, indicating a low
potential for organ failure
One or two group 1 KCFs 2 Either acute findings connoting a
short time course with an unclear potential for organ failure, or severe findings with high potential for future organ failure
One or two group 2 KCF or
Three or more group 1 KCFs
3 Both acute and severe findings indicating a high potential for imminent organ failure
One group 3 KCF or
Two or more group 2 KCFs 4 Critical findings indicating the
presence of organ failure
Two or more group 3 KCFs
(American journal of public Health Vol.81,No.1,p78,1991)
表九 Meaning and Computation of Mid-Stay Score
Mid-Stay Score Computation of Mid-Stay Score No morbidity All other
Morbid One group 2 acute KCF
or
One or more Group 2
Persistent KCFs Major morbidity One or more group
3 acute KCF or
More than one Group 2 acute KCF
or
One or more Group 2 acute KCF and one or More group 2 Persistent KCF
(American journal of public Health, Vol.81,No.1,p78, 1991)
表十 Number of Categories in Hospital Patient Classification System Classification System Number of
Primary Categories
Number of Partitions or Subcategories
Total Categories Diagnosis Related Groups
(Federal Register, 1985) 468 - 468
Severity of Illness Index With
DRGs(Horn and Horn, 1986) 468 4 1872
Disease Staging
(Conklin et al., 1984) 420 4 1680
Patient Management
Categories(Young, 1984) 800 - 800
Prospextive Individualized
Reimbursement(Johansen, 1986) 337 - 337
(Journal of medical system,Vol.12,No.4,p271,1989)
表十一 Summary Decription of Severity Measures Severity Measure Data Requirements Definition of
Severity
Diagnosis Used
Classification System
DRG Dischaye abstant Resource of needs
Yes Cost-based relative weight
Patient Management Categories
Discharge abstract Resource needs
Yes Cost-based relative weight compared with an average of 1.0: also classifies based on major surgery.
APACHEⅡ Chart review Risk of imminent death
No Score from 0 to 71 indicating overall patient severity.
Computerized Severity Index
Chart review Treatment difficulty presented to physicians
Yes Score from 1 to 4 for patient overall and for each /CD-9 -CM diagnostic code.
Disease Staging Q-Scale
Discharge abstract Resource needs
No Percentage compared with an average of 100
indicating overall patient severity.
MedisGroups Chart review Risk of imminent organ failure
No Score from 0 to 4 on admission and morbidity score at mid-stay review indicating overall severity.
表十二 Findings from Severity Reviews Severity
Measure
Severity Scores Quality Implications and Other Comments
Patient Management Categories
Relative cost weight =3.846 The patient is assigned to six PMCs, including AMI/
congestive heart failure without operation, acute renal failure without dialysis, and
pneumonia.
Normative studies found high death rates in several of the patient’s PMCs. Therefore, death would not be an unexpected outcome, and the case would not be flagged for a quality review.
If it were reviewed, the patient’s hospital course would be compared with the components of care designated for the PMCs assigned.
APACHEⅡ Admission APACHEⅡscore =9 Day 9 APACHEⅡ score =15
Normative data would be used to determine if this type of patient with an admission score of 9 has a high probability of death. If so , a quality review may not be indicated.
Computerized Severity Index
Admission overall severity =2 Maximum Overall Severity =4 Discharge overall severity =4 The /CD-9-CM diagnostic codes for acute myocardial infarction receive a score of 2.
Whenever a patient dies in-hospital, both maximum and discharge severity are assigned a score of 4. Because maximum severity is greater than admission severity, this case is flagged for quality review. Because of the absence of findings relating to bursitis, the accuracy of this ICD-9-CM code is questioned.
Disease Staging
Computerized Q-Scale Scores Score on day 1=329
Score on days 4 to 11 =370 Clinical criteria version:
Stage of coronary artery disease Stage on day 1 =3.2
Stage on day 4 =3.3 Stage on day 11 =3.9
Stage 3.2 is acute myocardial infarction with heart block, while 3.3 includes congestive heart failure and 3.9 indicates cardiac arrest. Both the Q-Scale score and clinical criteria stage increased over the hospital stay. Because of the relatively long stay over which severity
worsened, the case would receive a quality review. If the patient had died earlier, the case might not have been as suspect.
MedisGroups Admission severity =2 Mid-stay review =”morbid”
The decision about whether to flag this case for quality review would depend upon normative data. If acute myocardial infarction patients with admission scores of 2 have low morbidity and death rates, the presence of these events in this case would indicate the need for a review.
(QRB P379,Dec 1989)
表十三 Characteristic of Indices Included in Study
Index Input Data Version Used Scale
Produced
Scoring Available Computerized Disease
Staging(CDS)
ICD-9 CM codes from discharge abstracts.
No comorbidity.
No death adjustment.
Ordinal No
Medisgroup(MDGRP) Physiology findings obtained from chart review.
Admission
information only.
No death adjustment.
Ordinal No
Computerized
Security Index(CSI)
ICD-9-CM codes from discharge abstract and physiologic findings from chart review.
Admission
information only.
No death adjustment.
Ordinal No
Patient Management Category(PMC)
ICD-9-CM codes from discharge abstracts.
No procedur
No death adjustment.
Interval No
Acute Physiology and Chronic Health Evaluation (APACHE)
Physiologic findings from chart reviews.
Version II Interval Yes
Ischemic Heart Disease Index (IHDI)
Physiologic findings from chart reviews.
─ Interval Yes
Predictive Index for Myocardial Infarction(PIMI)
Physiologic findings from chart reviews.
─ Interval Yes
(Medical Care Vol.28,No.9,p764,1990)
表十四 Common Problems in the Difficult-to-Wean Patient
1. Anemia
2. Increased work of breathing 3. Secretions,atelectasis,plugging 4. Dyspnea
5. Malposition
6. Respiratory muscle fatigue 7. Hemodynamic and fluid
problems 8. Infection
9. Metabolic problems 10. Nutrition
11. Bowel problems 12. Exercise
13. Psychologic problems 14. Sleep disturbances 15. Pain
(Critical Care Nurse,Vol.9,No.1,p43 ,2000)
表十五 Management of Common Problems in the Difficult-to-Wean Patient
Problem Management Strategy
Anemia 1.Transfuse when Hb 10 and Hct
be a factor in decreased tissue oxygenation
1.Tube-related
a. Change size of small ET tube
b. Cut length of ET tube if 2 inches past mouth c. Deflate cuff if all breathing is spontaneous and risk of aspiration is minimal
2.Secretion-related(see secretions/stelectasis/
Plugging)
3.Bronchospasm-related a.Administer bronchodilators
Methylxanthines(Keep theophylline level between 10 and 20 pg/kg) c.Treatment of cause(see text) 1.Systemically hydrate
2.Extra humidily,especially at night
3.Maximally bronchodilate when necessary 4.Coughing exercises followed by deep inhalation 5.Chest physiotherapy and postural drainage 6.Suctioning
7.Other breathing exercises(segmental) 1.Proper positioning(OOB,dangling,leaming forward)
2.Decrease tidal volume on ventilator slowly 3.Periodic insufflation with ambu bag during longer times off ventilator
4.Increase endurance
a.Altemate weaning with rest to promote endurance
b.Inspiratory resistive training 5.Codeine to block dyspnea sensation
b.Inspiratory resistive training 5.Codeine to block dyspnea sensation