Disclaimer. Improving MET-based patient care using treatment algorithms. Michael R. Pinsky, MD, Dr hc. Different Environments Demand Different Rules

Transcription

1 Michael R. Pinsky, MD - June 29, 26 Improving MET-based patient care using Michael R. Pinsky, MD, Dr hc Department of Critical Care Medicine University of Pittsburgh Disclaimer Michael R. Pinsky, MD is the inventor of a University of Pittsburgh-held patent Functional Hemodynamic Monitoring Michael R. Pinsky, MD is a consultant regarding hemodynamic monitoring systems for: LiDCO Edwards Life Sciences Different Environments Demand Different Rules Hemodynamic Monitoring Emergency Department Rapid, minimally invasive, high sensitivity Triage, trends, protocols Operating Room Accurate, invasive, high specificity Close titration, zero tolerance for complications ICU Somewhere in between ER and OR Non-ICU God only knows special invasive minimally invasive Pre-op OR ICU ER floor Intensity Pinsky & Payen. Functional Hemodynamic Monitoring. Springer 24 MET-based Non-ICU care Problem: Now what? How to deal with complex and often uncommon presentations of acute complications and natural disease progression in an environment with limited diagnostic resources? Solution: Hierarchy of care: life threatening issues first Protocolized management Pittway algorithms Standard operating procedures Protocolized Support of Active Medical Care Protocol for managing hypertension (n=29, 166, 2231)» Brownbridge et al. J R Coll Gen Pract 36:198-22, 1986 Feedback comparing care with diabetes protocol (n=226)» McAlister et al. Brit Med J 293:67-4, 1986 Reminder of prior tests when ordering new ones (n=9496) reduction in test ordering» McDonald et al. N Engl J Med 295:1351-5, 1976 Alerts for in-patients taking digoxin n=396) action taken» White et al. J Am Coll Cardiol 4:571-6, 1984 Recommendations for management of 79 acute medical problems (n=155) # of errors in ordering tests» Young. Medical Informatics 6:13-7,

2 Michael R. Pinsky, MD - June 29, 26 Protocolized Care: Theory Protocolized minimizes practice variance Allows individualization of care only as exception Improves process in the real world Where error are a major course of impairment Decreases process in the ideal world Where excellence 24/7 is the standard of care Reduced Variance in Practice as a Source of Improved Outcome from Critical Illness Assuming that more good is not beneficial but a little bad is terrible Eliminating the few transgressions in care (prolonged F I O 2, high Paw) will minimize complications Decreased noise will also allow for discrimination among treatments ARDSNet NHLBI Study Group P =.45 at 28 days 6 ml/kg 12 ml/kg ARDS Network N Engl J Med 342: , 2 Protocolized Care Reduces Outcome Variance PaO 2 Distribution Target PaO 2 = > 6 mm Hg 5 4 Minimum 28 Control Group Maximum Points Mean 96 Median 86 1 Variance Minimum 11 Protocol Group Maximum 45 3 Points Mean 77 Median 7 1 Variance PaO 2 (mm Hg) East TD, et al. Protocolized Care Reduces Outcome Variance Tidal Volume Distribution Control Group Protocol Group Vt (ml) Minimum Maximum Points Mean Median Variance Minimum Maximum Points Mean Median Variance East TD, et al. Problems with Protocolized Care Requires defined input and feedback SpO 2 and FiO 2 MAP and vasoactive drugs Blood glucose and insulin Limited protocols available Major clinical issues not addressed Circulatory shock Acute respiratory failure 2

3 Michael R. Pinsky, MD - June 29, 26 Monitoring Truth No monitoring device, no matter how accurate or insightful its data will improve outcome, Unless coupled to a treatment, which itself improves outcome Pinsky & Payen. Functional Hemodynamic Monitoring, Springer Verlag, 24 Pinsky & Payen. Crit Care 9: Why Not Give Volume to All Hemodynamically Unstable Patients? Signs of cardiovascular insufficiency are impressive but not specific Hypotension must decrease blood flow to the heart and brain Most forms of circulatory shock have a pathological component of decreased effective circulatory blood volume Predicting Fluid Responsiveness in ICU Patients Responders / Non-Responders % Responders Calvin (Surgery 81) 2 / 8 71 % Schneider (Am Heart J 88) 13 / 5 72 % Reuse (Chest 9) 26 / % Magder (J Crit Care 92) 17 / % Diebel (Arch Surgery 92) 13 / 9 59 % Diebel (J Trauma 94) 26 / 39 4 % Wagner (Chest 98) 2 / % Tavernier (Anesthesio 98) 21 / 14 6 % Magder (J Crit Care 99) 13 / % Tousignant (A Analg ) 16 / 24 4 % Michard (AJRCCM ) 16 / 24 4 % Feissel (Chest 1) 1 / 9 53 % Mean 211 / % Neither ΔCVP or ΔPpao Mirror ΔSV Lichtwarck-Aschoff et al. Intensive Care Med 18: 142-7, 1992 Michard & Teboul. Chest 121:2-8, 22 Neither CVP or Ppao reflect Ventricular Volumes or Tract Preload-Responsiveness Preload Preload Responsiveness Dr. Pinsky s Hemodynamic Truths Tachycardia is never a good thing Hypotension is always pathological There is no normal cardiac output CVP is only elevated in disease Peripheral edema is of cosmetic concern Kumar et al. Crit Care Med 32:691-9,

4 Michael R. Pinsky, MD - June 29, 26 Functional Hemodynamic Questions Will cardiac output increase with fluid resuscitation, and if so, by how much? Is arterial tone increased, normal or decreased? Is the heart able to maintain an adequate output under pressure without high filling pressures? Predicting Preload Responsiveness from CVP waveform analysis Inspiratory fall in CVP No inspiratory fall in CVP Respiratory variations in CVP predict response to fluid challenge Initial Central Venous Pressure 2 Change in Cardiac Output (L/min) CO increased in 1/14 pts Negative CVP Response CO increased in 16/19 pts Positive CVP Response Magder et al. J Crit Care 7:76-85, 1992 CVP mm Hg No Insp Fall +ve Insp Fall Magder et al. J Crit Care 7:76-85, 1992 Predicting Preload-Responsiveness Passive Leg Raising Protocol Base 1 Legs up Base 2 NaCl 5 ml Post VE Monnet et al. Intensive Care Med 31: , 25 Δ Mean aortic flow (%) Change in Mean Aortic Flow during PLR base 1 legs up base 2 post VE Non responders Responders P <.5 Monnet et al. Intensive Care Med 31: ,

5 Michael R. Pinsky, MD - June 29, 26 Mean aortic blood flow is better than PP in Predicting Preload- Responsiveness during PLR during Spontaneous Breathing and with Arrhythmias % change from Base PLR-induced changes in Aortic flow NR R PLR-induced changes in PP NR R p <.5 Monnet et al. Crit Care Med 34:142-7, 26 Rule: CVP and Flow Variation ΔCVP with spontaneous breathing & Δflow with passive leg raising accurately identify subjects whose cardiac output will increase during a fluid challenge and by how much If either ΔΧVP or Δflow are not present, then fluid loading will not increase cardiac output Arterial Tone Assessed by Ventriculo-Arterial Coupling Stroke Volume Defines Arterial Pressure 2 Mean Arterial Pressure 5 5 Stroke Volume E a Arterial Elastance (E a ) Ventriculo-Arterial Coupling Vasopressor Therapy Pain - Hemorrhage Acute Heart Failure Tamponade + - Pulse Pressure Volume Loading Increased catecholamines Exercise + Stroke Volume Vasodilator Therapy Sepsis Continuous Monitoring of Preload Responsiveness CVP Arterial Pressure Non-invasive Fenapres Pulse oximetry pulse density contour Invasive Arterial catheterization Arterial flow USCOM Esophageal Pulsed Doppler Arrow HemoSonic Deltex CardiaQ Echocardiogram Combined Pressure and Flow Pulse Contour Technology PiCCO, LiDCO Hemodynamic Monitoring Protocol Is the patient hemodynamically stable? Do Nothing Yes No Volume bolus Add Vasopressor Is the patient preload-responsive? Yes Is the patient hypotensive and have reduced vasomotor tone? Yes No Yes No No Volume bolus Add Vasopressor Add Inotrope Reassess the patient 5

6 Michael R. Pinsky, MD - June 29, 26 Thank You 6