The incidence of clinically significant errors in the bedside measurement of pulmonary capillary wedge pressure.

This research determined the incidence of clinically significant errors in the bedside measurement of pulmonary capillary wedge pressure (Pw). Routine Pw measurements were compared to Pw's confirmed by the following more stringent criteria: a) Pw less than the mean pulmonary artery pressure (PAP); b) atrial waveform; c) free-flow through the catheter and blood vessel; and e) aspiration of capillary blood. One hundred and forty-three comparisons were made in 53 ICU ¦ patients being monitored with a Swan-Ganz catheter. Technical problems identified by the careful waveform analysis required to confirm the Pw were present in 41% of the initial measurements. These technical problems included: inadequate waveform (15%), poor dynamic response (20%), overinflation (4%), and suspect Pw (13%). Seventy-five percent of the technical problems were easily corrected by removing air bubbles, tightening loose connections, withdrawing or advancing the catheter 1 to 2 cm, increasing balloon volume, or irrigating the catheter. The remaining problems required catheter repositioning to resolve. The overall incidence of clinically significant errors (? 4 mmHg) was 17%. When a technical problem was present, the incidence of an error ? 4 mmHg was 23%. If the technical problem was not easily corrected and required catheter repositioning to resolve, the incidence of an error ? 4 mmHg was 53%. In the absence of technical problems, an error of ? 4 mmHg occurred in only 4% of the measurements. Aspiration of capillary blood was the only criterion which detected these errors. The criteria used to confirm the Pw requires careful waveform analysis (Criteria 1,2,3) and aspiration of capillary blood (Criteria 4). These criteria identify problems which can lead to errors in the Pw measurements. Identification and correction of technical problems, using waveform analysis improved the accuracy of the Pw measurement and can decrease the incidence of a clinical significant error from M% to only 4%. Waveform analysis, using Criteria 1, 2 and 3 should be done routinely in the critical care setting when measuring the Pw. Capillary blood aspiration (Criterion 4) offers only minimal improvement in the accuracy of the Pw measurement (4%) and is not warranted on a routine basis.

THE INCIDENCE OF CLINICALLY SIGNIFICANT ERRORS IN THE BEDSIDE MEASUREMENT OF PULMONARY CAPILLARY WEDGE PRESSURE by Radene Holt Chapman A thesis submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Master of Science College of Nursing The University of Utah June 1982 ~1982 Radene Holt Chapman All Rights Reserved THE U~IVERSrTY OF UTAH GRADUATE SCHOOL SUPERVISORY CONlivIITTEE i\PPRO\l AL of a thesis submitted bv Radene Holt Chapman This thesis has been read by each member of the following supervisory committee and by majl vote has been found to be satisfactory. Chainnan: Maeona K. Jacobs, R.N., Ph.D. Marjorie 'Cengiz, R.N., ~S. o Reed M. Gardner, Ph.D. THE UNIVERSITY OF UTAH GRADUATE SCHOOL FINAL READING APPROVAL To the Graduate Council of The University of Utah: I have read the thesis of Radene Holt Chaoman in its final form and have found that (I) its format, citations, .and bibliographic style are consistent and acceptable; (2) its illustrative materials including figures, tables. and charts are in place; and (3) the final manuscript is satisfactory to the Supervisory Committee and is ready for submission to the Graduate School. Marjoriej Cengiz, R.N",~ ~.S. ~1emb.:r. Supen isory Committee Linda K. Amos, Ed.D., F.A.A.N. Chairman Dean Approved for the Graduate Council I / James L. Claytp~ Ikan 01 rh", (iradual .... School ABSTRACT This research determined the incidence of clinically signifi­cant errors in the bedside measurement of pulmonary capillary wedge pressure (Pw). Routine Pw measurements were compared to Pw1s con­firmed by the following more stringent criteria: a) Pw less than the mean pulmonary artery pressure (PAP); b) atrial waveform; c) free-flow through the catheter and blood vessel; and e) aspiration of capillary blood. One hundred and forty-three comparisons were made in 53 ICU patients being monitored with a Swan-Ganz catheter. Technical prc­lems identified by the careful waveform analysis required to con­firm the Pw were present in 41% of the initial measurements. Thes~ technical problems included: inadequate waveform (15~b), poor dynan.lc response (20%), overinflation (4%), and suspect Pw (13~~). Seventy­five percent of the technical problems were easily corrected by re­moving air bubbles, tightening loose connections, withdrawing or advancing the catheter 1 to 2 cm, increasing balloon volume, or irrigating the catheter. The remaining problems required catheteT~ repositioning to resolve. The overall incidence of clinically si~­nificant errors (~4 mmHg) was 17%. When a technical problem was present, the incidence of an error >4 mmHg was 23%. If the tecr­nical oroblem was not easily corrected and required catheter re- positioning to resolve, the incidence of an error ~ 4 mmHg was 53%. In the absence of technical problems, an error of ~ 4 mmHg occurred in only 4% of the measurements. Aspiration of capillary blood was the only criterion which detected these errors. The criteria used to confirm the Pw requires careful wave­form analysis (Criteria 1,2,3) and aspiration of capillary blood (Criteria 4). These criteria identify problems which can lead to errors in the Pw measurements. Identification and correction of technical problems, using waveform analysis improved the accuracy of the Pw measurement and can decrease the indicence of a clinical' significant error from 17% to only 4%. Waveform analysis, using Criteria 1, 2 and 3 should be done routinely in the critical care setting when measuring the Pw. Capillary blood aspiration (Criter­ion 4) offers only minimal improvement in the accuracy of the Pw measurement (4%) and is not warranted on a routine basis. v ABSTRACT . . . LIST OF TABLES LIST OF FIGURES. ACKNOWLEDGMENTS. INTRODUCTION Chapter CONTENTS I. CONCEPTUAL FRAMEWORK ...... . Problem Statement and Purpose. Review of Literature ............ . iv .viii ix xi 1 Errors in the Pw Measurement ........ . 11 Causes of Errors . . . . . . 11 Technical Problems . Catheter Position .. Rationale for Study. Research Questions . Role of the Nurse. II. METHODOLOGY ... Design ..... Sample Selection Equ i pment. . . Definitions .. III. RESULTS .... Sample Characteristics . . ... Incidence of Technical Problems. Pw Measurements Following Correction of Technical Measurements ..... . Blood Aspiration ......... . Probability of Accurately Measuring Pw . 21 2f 2E 30 IV. DISCUSSION, RECOMMENDATIONS AND CONCLUSIONS . . . . . . . . . Clinically Significant Errors ........ . Is There a Clinically Significant Difference in the Pw Measurement When Errors are Caused by Easily Correctable Technical Problems? ............... . Does the Aspiration of Capillary Bood Improve the Accuracy of the Pw Measurement? . . . . Is there a Clinica1ly Significant Difference in the Confirmed Pw and the Unconfirmed Pw? . Areas for Further Study .. Recommendations. Conclusions ....... . Appendices 51 51 54 60 63 63 65 66 A. PROTOCOL FOR HEMODYNAMIC MEASUREMENTS . . . 68 B. PW MEASUREMENTS CATEGORICALLY ILLUSTRATED. 74 REFERENCES . 76 vii 1. 2. 3. 4. LIST OF TABLES Pw Measurements Before and After Confirmation of the Pw Illustrating the Incidence of Significant Errors in the Unconfirmed Pw .......... . Diagnostic Classification and Placement of Study Patients: Shock-Trauma lCU (S-TICU) or Inter­mountain Respiratory lCU (lRICU) .... Distribution and Significance of Technical Problems in 143 Pw Measurements ..... . Correlation of Confirmed Pw and Left Atrial Pressure in Five Post Cardiac Surgery Patients. 12 40 62 LIST OF FIGURES 1. Measurement of Pw pressure with Swan Ganz catheter. . 6 2. Direct pressure monitoring system 3. Inadequate waveforms. 4. Poor dynamic response 5. Balloon overinflation ... 6. Suspect or partial wedge catheter . 7. No Pw waveform ...... . 14 15 IS 1 - 2 2 8. Blood flow through the lung in the supine position. . 2 9. Pw measurements in Zone II. . 2 10. Conceptual model .... 11. Flowsheet of steps taken to determine the incidence of cl inically significant errors in the measurement of Pw . . . . . . . . . . . . . . . . . 12. Pw confi rmati on wi th capi 11 a ry blood. . . 13. The outcome of 143 Pw measurements ,during steps taken to confirm the Pw ..... . 14. Pw errors due to technical problems 15. Pw errors due to technical problems related to catheter position ....... . 16. Pw errors identified by capillary blood aspiration. 17. Probability of accurately measuring the Pw within 2 mmHg, 4 mmHg, and ± 6 mmHg. .... 18. Normal variation in Pw. 19. Left-ventricular stroke work index (LVSWI) plotted against Pw ...... . 2 35 37 41 C7 5 20. Partially wedged catheter .......... . 21. Pw measurements categorically illustrated with actual Pw values and differences after corrections were made at Points A,B,C,D . . . . . . . . . . . . . . . . 74 x ACKNOWLEDGMENTS There are several people I wish to thank who have given me help and encouragement in the research and preparation of this thesis. My husband Hal, for being critical, yet supportive; my committee members Maeona Jacobs, R.N., Ph.D. and Reed Gardner, Ph.D., for their careful review and help in writing; Marge Cengiz. R.N., M.S., for her help not only as a committee member, but as a teacher, co-worker and especially friend. I would also like to thank Dr. Alan Morris for his influence on my clinical abilities as a nurse and in recognizing the importance of careful data col­lection whether for research or patient care. This research is part of a larger study done under his direction. Lastly lowe special thanks to the nurses in the Intermountain Respiratory and Shock-Trauma Intensive Care Units at LDS Hospital. INTRODUCTION The introduction of the Swan-Ganz catheter in 1970 has made routine measurement of pulmonary capillary wedge pressure (Pw) com­mon practice in the intensive care setting (Swan, Ganz, Forrester, Marcus, Diamond & Brown, 1970; Pace, 1977; Walinsky, 1977). When accurate, the pulmonary capillary wedge pressure (Pw) provides val ,­able information for the assessment of the adequacy of intravascul~ fluid volume and left ventricular function (Swan et a1., 1970; Pac" 1977; Hurst, Logue, Schlant & Wenger, 1978, pp. 479-514; Lalli, l~ Vij, Babcock, Magilligan, 1981; Walinsky, 1977). Errors in meaSUl­ing pulmonary capillary wedge pressure (Pw) could lead to mismana~:~­ment of patients already critically ill. Technical problems with monitoring equipment and/or the catheter itself, catheter position, improper zeroing, and respiratory variation can result in these errors. If transducers and recorders are properly zeroed and Pw is measured at end-expiration (Cengiz, 1980), errors due to poor zeroing techniques a~d respiration can be eliminated. This re­search focused on the incidence and clinical significance of errors in the Pw resulting from technical problems and catheter position. Identification of these problems, the physiological and mechanical mechanisms which may cause errors, and their correction were also addressed. CHAPTER I CONCEPTUAL FRAMEWORK Problem Statement and Purpose The rapid advances in medical technology during the last few years requires nurses and physicians to know more than physiology pathophysiology, physical assessment, and therapies, in order to care for critically ill patients. They also must have a knowledgf of sophisticated instrumentation. Many parameters, including ar­terial blood pressure, heart rate, respiratory rate, temperature, pulmonary artery pressure, and pulmonary capillary wedge pressure, which were measured with difficulty a few years ago are now auto­matically displayed and need only to be observed by medical perso~­nel. Often nurses and physicians do not knriw how the value is de­rived, or how to identify and troubleshoot problems that occur. Patients may be seriously mismanaged if errors are made in the auto­mated measurement of these important clinical parameters. The pUl­monary capillary wedge pressure (Pw) measurement is an example of automation's impact on clinical parameters. The measurement ap­pears to be simple, straightforward and exact; this may not be the case. When the Swan-Ganz catheter was introduced, it was reported thE Pw could be measured successfully only 72% of the time (Swan et al .. 3 1970). Similar levels of accuracy have been observed in the car-diac catheterization laboratory where Pw has been measured and used since the 1940 l s (Hurst et al., 1978; Rappaport & Dexter, 1958). Even though automated systems are used in the cardiac catheteri-zation laboratory, critical analysis of the Pw waveform is made and specific criteria are used to measure and confirm accurate Pw mea-surement. Critical care units in large university hospitals and small community hospitals now commonly measure Pw with automated devices alone. The balloon on the catheter tip is inflated and, as long as the value of the oscilloscope or digital display decreases, it is recorded and assumed to be correct. This technique is simple and takes little time, however, the waveform is usually not care-fully analyzed (Woods, 1976; Shin, Ayella & McAslan, 1977; Brown, 1977). The lack of careful waveform analysis and confirmation of the Pw measurement in the intensive care (lCU) setting may lead to errors in clinical decisions. This research addressed whether there is a clinically signif~- cant difference between the Pw as measured routinely in the leU setting and the Pw measured and confirmed using cardiac catheteri-zationlaboratorycriteria, including waveform analysis and confirm-ation with capillary blood aspiration. Review of Literature Physiologic Basis of Pw as a Guide in Assessing Cardiac Function Starling's law of the heart states that the more the heart is 4 filled during diastole, the greater the quantity of blood it will pump during systole. The degree of stretch imposed by the filling of the ventricles during diastole is termed preload and is reflected by the left ventricular end diastolic pressure (LVEDP). Increased LVEDP (preload) during diastole is associated with increased stroke volume and thus, cardiac work during systole. When the heart be­gins to fail, the LVEDP increases to maintain cardiac output. At high pressures the heart may not be able to compensate and cardiac output falls. Conversely, a decrease in LVEDP indicates a lower blood volume available to the heart. The Pw has been shown to equi., the LVEDP as long as the mitral valve is competent (Hurst et al., 1978), and is used clinically as an indication of the preload stat'~ of the left ventricle. When correlated with right atrial pressure, mean blood pressure, and cardiac output, Pw is also used in the assessment of left ventricular function and fluid status. In many patients, the Pw is elevated before there is clinical evidence of left heart failure or increased fluid volume. Often patient symp­toms and bedside examination are not sufficiently sensitive to docu­ment a problem until serious progression of the primary process has occurred. Gallops, rales, pulmonary edema, hypotension and other clinical signs are often late findings which are preceded by a changE in the Pw (Hurst et al., 1978; Forrester, Diamond, Chatterjee, & Swar 1976). A Pw of greater than 18 mmHg (normal = 6 - 12 mmHg) is consistent with assessment of left ventricular failure or fluid overload. Values of less than 11 mmHg associated with hypotension or other symptoms of shock, indicate hypovolemia (Pace, 1977; Shoe- 5 maker & Thompson, 1980, pp. 10-23; Forrester et al., 1976). Techniques of measurinq Pw. The Pw is measured by occlusion ,jT blood flow from the pulmonary artery by a catheter. This separates the pulmonary capillary bed and left heart from the right heart and pulmonary artery_ Since there are no valves from the pulmo­nary capillary bed to the left atrium, the tip of the catheter senses the equilibrated pressure of one system which, during dias­tole, includes: the pulmonary capillary bed, left atrium, mitral valve, and left ventricle. The Pw equals the LVEDP (See Figure 1). Both the Swan-Ganz catheter, used in the ICU setting and the Cournand catheter, used in the cardiac catheterization laboratory, utilize the occlusion principle in measuring the Pw. The Swan­Ganz catheter is a flow-directed, balloon tipped catheter which oc cludes the vessel (wedges) when the balloon is inflated. It is positioned in the main pulmonary artery and, when the balloon is inflated, floats out into a smaller vessel until it wedges. When the balloon is deflated, the tip returns to the main pulmonary artery (Swan et al., 1970; Lalli, 1978). The catheter can be in­serted and used in the ICU without fluoroscopy. The Cournand catheter is a teflon catheter which must be in­serted and manipulated under fluoroscopy. The catheter itself oc­cludes the vessel, and so, must be advanced distally and manually manipulated under flouroscopy for each Pw measurement (Pace, 1977). Methods used to verify the accuracy of the Pw measurement. Whet the p\'1 is measured in the catheterization laboratory five criteria are used to confirm the accuracy of the measurement: alveolus pulmonary vein Ll closed pul monary closed aortic valve ! left open \I tricuspid val ve va 1 ve systemic ~ right circulation ventricle atrium Fiqure 1. Measurement of Pw pressure with Swan Ganz catheter. When the balloon is infla , occlusion of the pulmonary artery blood flow occurs, and measurement of the equili­brated pressure of one system, which during diastole, includes: the pulmona capillary bed, left atrium, mitral valve, and left ventricle, is made. Labora tory, Understandi n~!l1o~YDami c me~surements made "Ii th Swan-Ganz catheter. Santa Ana, California: Author, 1979. (J) 1. The Pw must be less than the mean pulmonary artery pressure (PAP) (Pace, 1977; Hurst et al.~ 1978; Mendel, 1968, p. 234). Since blood flows from a higher to a lower pressure, a Pw higher than the PAP would indicate a reversal in the normal blood flow, backwards through the lungs from the left heart. Since this does not oc­cur, the Pw greater than the PAP indicates an incorrect sensing of the Pw pressure. 2. The Pw waveform must be an atrial pressure wave­form (P ace, 1977; H u r s t eta 1 ., 1978 ; Me n de 1 ~ 1968; Langerlof & Werko, 1947). The atrial pres­sure waveform indicates that the catheter is re­flecting the pressure in the left atrium and that the pressure or flow from the pulmonary artery is totally occluded. 3. Free-flow must be present with the catheter in the wedged position (Rappaport & Dexter, 1958). Free-flow or open communication through the catheter and the pulmonary capillary bed is es sential if the Pw is to be correctly measured. This indicates that there is no obstruction in the catheter, at the tip, or in the vessel. 4. A palpable "give ll should be present \1Jhen the catheter is pulled back from the wedged posi­on (Mendel, 1968; Langerlof & Werko, 1947). 7 This "give" indicates that the catheter has been IIwedged" in a small vessel. 5. Highly oxygenated blood should be aspirated from the wedged position (Swan et al., 1970; Rappaport & Dexter, 1958; Mendel, 1968; Langerlof & Werko, 1947; Dexter, Hayes, Burwell, Eppinger, Sagerson & Evans, 1947; Brewster & McIlroy 1973). Avail­able explanations of highly oxygenated specimens aspirated through a "wedged" catheter all lead to the conclusion that the catheter tip has been separated from the blood pool (venous) in the pulmonary artery and has probably produced a local lung region of high ventilation to perfusion ratio due to the obstruction of the blood flow (Rappa­port & Dexter, 1958; Brewster & McIlroy, 1973; Archer & Cobb, 1974). Rappaport and Dexter (1958) first described this observation and felt it to be the most important of the five criteria. Some authors state that, depending on the location of the catheter tip, in the presence of a high right to left pulmonary shunt, highly oxygenated blood cannot be aspirated even though an accurate Pw may be measured (Swan et al., 1970; Shapiro, Smith, Pribble, Murray & Cheney, 1974). 8 A widely used critique of the Swan-Ganz catheters discussed n an extensive review article by Pace (1977) dismissed the utility of the capillary blood criterion in Swan-Ganz catheters. He re­ferredtothree authors to justify the exclusion. The findings pre­sented in these articles, however, do not indicate that the capil­lary blood criteria is inappropriate in the Pw measurement with the Swan-Ganz catheter. Swan and Ganz (1970) found that when blood was aspirated from the wedged position it varied in oxygen content between that of pulmonary capillary and arterial blood. They suggest downstream venous blood contaminated the capillary blood, due to the balloon wedging in a larger vessel than the Cournand catheter does. This problem can be remedied as recommended by withdrawing between 5 a 15 mlls of blood before taking the capillary sample (Suter, Lind­auer, Fairley, & Schlobohm, 1975). Brewster and McIlroy (1975) suggested that in patients breath ing supplemental oxygen, the oxygen content in the capillary bloo: may not be as sensitive an indicator as the pH and PC02" They dij not suggest capillary blood aspiration is not useful in the bedside confirmation of Pw with a Swan-Ganz catheter. A third author cit20 by Pace in his review discussed the ability to accurately measure the Pw in dogs being treated with positive end-expiratory pressur~ (PEEP). Capillary blood aspiration was attempted to confirm the Pw but was unobtainable at high levels of PEEP, when the Pw was greater than the left atrial pressure. When the left atrial pres­sure rose in parallel with thePw, capillary blood was obtained. This finding validates the usefulness of capillary blood in de­termining the accuracy of the Pw as a reflection of left atrial pressure (Scharf, ealdini & Ingram, 1977). The criteria commonly used to identify the Pw when measured in the IeU with a Swan-Ganz catheter are: 1. The Pw must be less than the mean pulmonary artery pressure (PAP) (Pace, 1977). This ob­servation is often made by simply observing a change in the mean pressure reading on a digi­tal readout or oscilloscope to a lower value and not by actually evaluating the waveform. 2. A chanqe in the pulmonary artery waveform when the balloon is inflated (Pace, 1977). This ob­servation is usually made by observing the waveform on the oscilloscope for a change from the pulsa­tile waveform of the pulmonary artery pressure to a lower more atrial waveform. Ie Utilizing the above criteria, Pw measurements in the ICU set­ting are not commonly based on critical analysis of the waveform, freeflow of the catheter system, or aspiration of capillary blood as they are in the cardiac catheterization laboratory_ All of the criteria used in the catheterization laboratory, except the "palp­able give" (Criterion 4) can be assessed at the bedside with a Swan-Ganz catheter. In this study, an unconfirmed Pw is defined using the two criteria outlined above. A Pw verified at the bed­side by all of the criteria used in the catheterization laboratory, except a "palpable give," is defined as a confirmed Pw. 11 Errors in the Pw Measurement Identification of Errors Confirmation of the Pw leads to identification and correction of problems not appreciated in the unconfirmed Pw measurement. A preliminary survey of 12 patients in an lCU demonstrated that un­confirmed Pw was different than the confirmed value (See Table 1) (Morris & Holt, 1980). One of these patients (*) was treated for 12 hours on the basis of an unconfirmed Pw of 24 mmHg. Diuretics were administered and fluids restricted, because an assessment was made of left ventricular dysfunction and increased intravascular volume. When, after 12 hours, the Pw remained between 22 and 24 mmHg and the patient was deteriorating, an attempt vias made to confirm the Pw. After the S\'Jan-Ganz catheter was repositioned to meet the confirmed Pw criteria, the confirmed Pw was 8 mmHg. The patient was actually hypovolemic with normal left ventricular funt­tion. Using the confirmed Pw as a guide, blood and fluids were administered and the patient showed marked improvement over the next two hours. Causes of There are four major sources of error in the Pw measurement: a) technical problems; b) catheter position; c) respiratory vari­ation; and d) improper zeroing and calibration. Of these, problems due to improper zeroing, calibration, and respiratory variation CJn be controlled and were not part of this study (Cengiz, 1980). Table 1 Pw Measurements Before and After Confirmation of the Pw Illustrating the Incidence of Significant Errors in the Unconfirmed Pw Pw unconfirmed Pw confirmed 30 22 8 15 8 24* 23 12 36 21 7 14 8 8 12 6 12 8 13 8 18 4 20 20 lL. Note: Mean difference = 11 mmHg + 6 (SO); Range (4-22) (Morri~; & Ho 1 t, 1980) 13 Technical Problems Technical problems have been reported in 30% of Pw measuremen:s (Morris & Holt, 1980). These are related to the operation of the entire measurement system: catheter, transducer, monitor and con­nections (See Figure 2). These problems are identified and cor­rected by the practitioner, if the criteria for a confirmed Pw are met. Only a few are identified and corrected when the routine or unconfirmed criteria are applied. Identified technical problems include: 1. Inadequate tracing (Archer & Cobb, 1974; Shapiro et al., 1974) ( Figure 3). This is a waveform which contains little high frequency information by visual observation. The catheter may be oc­cluded with blood, the tip pressed against the vessel wall, an air bubble may be in the trans­ducer or flush device, or the pressure in the fluid source of the flush system may be too low. Correction may require: removal of bubbles, ir­rigation of the catheter, tightening loose con­nections, withdrawal of the catheter 1-2 cm, de­flation and reinflation of the balloon or replace-ment of pressure in flush system. 2. Poor dynamic response or "flush" (Gardner, 1981; Gardner, Bond & Clark, 1977) (See Figure 4). If a sudden p change in pressure fails to produce rapid oscillations, there is an inadequate flush. pressurized flush solution I • V • Bedside pressure monitor electric cable transducer Figure 2. Direct pressure monitoring system. Commonly used system to directly measure Pw and other pressures in critical care areas. System includes: catheter, stopcocks, intra­o (continuous flush device), pressurized solution, transducer, pressure monitor, and recorder. I--' +:> 100 90 70 60 -~-=i:··j t Figure 3. Inadequate waveforms. Above tracings are examples of identified problems, corrected by: (A l ) rC~0'/,1' 0~ air huhhlps from ~rilnsducer: (13~) withdrawinq cat,heter 3 crn; and (C,) inc,reasing ~ preSSU(i.:lllflu~)h sy t(:111 Ll) J(H) illl:!l,.! 1" IlI·i, ii, :. i' Ii,: Ihlvefonns belov'_ Paper speed 25 (!Hlll ,( A B c o Intraflo \r~ 100 - 90 80 ' 70 60 50 - 40 10 OmmHg rl B' 1 I i Figure 4. Poor dynamic response. Above problems are corrected by: (Al) irrigating catheter with a syringe; (Bl) removing air bubbles from Intraflo; (C1) withdrawi catheter 3 cm.; (01) t i q fl t C~ n i loose connect ion n; s u 1 t i III ~ t, "; ~ ~..J _-l. •• ~ - ,-,.: ~ r ':"' ::; P 0 n s e s 1 0 w . Paper S 0 e e dis ') r; film I <:; F! C • I-' m The IIflush method ll has been described by Gardner et al. (1977) as a clinical check for· free-flow in the catheter system. It is easily performed with most continuous infusion systems used with Swan-Ganz catheters. The opening and sudden closing of the flush valve provides the assessment of free-flow. When checking the dy­namic response, observation of oscillations shou1d be identified such that: the time between each oscillation is less than 1 mm at 25 mm/sec. sweep speed, there is more than one oscillation, and the oscillation goes directly into the waveform without pegging on the bottom or top of the re­cording. This indicates an adequate underdamped system for measuring pulmonary and systemic pres­sures. In these systems, large errors can be made if a free-flow is not established. When the catheter tip is against a vessel wall, an in­creased pressure would be introduced by a flow resistive pressure drop across the partially oc­cluded tip. The balloon will commonly inflate assymetrically, forcing the tip of the catheter into the vessel wall (Lozeman, Powers, Older, Dutton, Roy, English, Marco & Eckert, 1974). This may occlude or partially occlude the cathe­ter and give false readings. Air bubbles, leaks, 17 a ball clot on the catheter tip, or blood in the system can also be identified by this maneuver. Correction may require removing air bubbles, ir­rigating the catheter, tightening loose connec­tions, withdrawing catheter 1-3 cm, deflating and reinflating the balloon, replacing pressure or fluid in pressurized flush solution, adjusting balloon volume and/or replacing the catheter if the problem is not correctable. Uncorrectable problems are usually due to assymetrical balloon inflation or, most commonly, a ball clot on the catheter tip. 3. Overinflation (Archer & Cobb, 1974) (See Figure 5). An increase in pressure during balloon inflation indicates overinflation of the balloon. It is due to occlusion of the distal catheter port by the balloon or vessel wall. It can be corrected by decreasing the amount of air in the balloon, with­drawing the catheter 1-3 cm or deflating and re­inflating the balloon. Overinflation usually in­dicates that the catheter has migrated to a more distal segment of the pulmonary artery and, if not withdrawn, may become permanently wedged, occlude the pulmonary blood flow distal to it, causing pulmonary infarction. 4. Suspect Pw ( Figure 6). Waveforms which appear 18 1 10 100 5 0 --~--~~-------'--1j~"';'--' 40 30 20 10 Figure 5. Ba 11 oon nflation. Ih i iL 1 (:1.1 COI'I'ec ll2d -1 ----7"- -- ~ - ! Balloon wi thdrawn 'j Examples of waveforms caused by balloon overinflation. ~'J i tl,,: j , .1'. ; ~ ;~ -; ~ - :~ - ~ ~ - .- ~ ..... +- -;- ,1 i r fro rn b all 00 n . - - ~- ! The t--' l::J A ECG 100 90 80 10 OrnlllHg. Fi B ~ c i In ~ IS t PAP I------r- i-I ~.! t~t re 6. Suspect or partial wedge catheter. The pulmonary artery pr'essure (PAP) tially occluded with balloon inflating resulting in a partial wedge (8). aft C:, d d v d n c i n ~J til(:: C d till: t (2 I' ~: I.i Ii \ I. j. I i \ . t ! C • --f ) is only par­Pw is obtained to be partially wedged or without an atrial looking waveform. When the balloon on· the catheter is inflated, the blood flow from the pulmonary artery may not be totally occluded. The pressure sensed is between PAP and Pw. Correction requires inflation of the balloon to maximum volume (0.8 to 1.5 ml), de­flation and reinflation of the balloon, or advancement of the catheter 1 to 3 cm. 5. No Pw (Pace, 1977) (See Figure 7). No waveform of atrial character or change in the pulmonary artery waveform can be obtained when the balloon is inflated. Correction requires advancement of the catheter or catheter replacement if the balloon is broken (See Appendix I for Protocol and step by step identification and correction of technical problems). One of the questions leading to this research was how much error do technical problems actually cause. Potentially large errors can be made, but most are easily identified and corrected if waveform analysis and flush testing is done. The confirming cri­teria outlined above were used to identify errors due to these problems. Technical problems such as no Pw and overinflation are obvious; others are more subtle. Routine lCU measurement only occasionally expose damped Pw and partial or suspect Pw and never a poor dynamiC response. 1 90 80 70 50 10 A PAP Figure 7. waveform. cJthcter is a B Wi th the ba 11 oon i nf1 a ted (V-) no PvJ via veform is observed (A). need 4 em to obtain a Pw waveform (B). The r0 N Catheter Position Errors caused by catheter position usually result from the catheter location in what West (1979) has identified as Zone I or II of the lung. In this area of the lung, alveolar, or airway pres­sure, rather than the Pw, or capillary pressure is measured (Tooker, Huseby & Butler, 1978). West's (1979) model of lung zones explains this phenomenon (See Figure 8). Blood flow is very high near the bottom or dependent area and decreases steadily up the lung. This is due to hydrostatic pressure, or the effect of gravity. The lun~ is divided into three zones, which are determined by the arterial (Pa), alveolar (PA) and venous (Pv) pressures. At the top of the lung (Zone I), the alveolar pressure is greater than the arterial pressure and the venous pressure, thus no blood flows through the capillaries because they are collapsed. In normal lungs, Zone I does not exist because the pulmonary arter_of pressure is just greater than the hydrostatic pressure. In Zone II the arterial pressure is greater than alveolar pre~­sure but the alveolar pressure exceeds venous pressure. In this case, collapsible tubes (capillaries) surrounded by a pressure chamber (the alveoli) are open to blood flow as long as the capii­lary pressure is greater than the chamber pressure. Near the end of the capillary, where capillary pressure becomes closer to veno~ than arterial pressure, the vessel narrows. The degree of narrow­ing is totally dependent upon the arterial pressure and upon how much r it is than the constant alveolar pressure. Zone III is that part of the lung where arterial pressure '::,,' :-: f:.,.:. '.;, :" ::' ~.,,:,::' ~.~ , : - :,---',:',-:~' ~,-.. ; ::.: ',::: , .. -:'.' "" "-:": '," ,,l , '- 8PA Zone 1 . " ..... - -, ", - -~ , ' , • . -' ' " ,.. P : _,; ': ~:, :, :," " : ' ' .- : , " : _ ~', -:,' ~: :. '~", ' ", . ," "v P a '~'~: ' " , : .. ' , .. , .... - - ' ' ,-- , . - ',-' ", " ,- . ' " ,: " : _ " ",: "...- .. ,' .. : - - __ ---.0 , , .; _' .. : :' , ' " " Z 0 n e 2 ,:' ,": .. \.., ' ~ ' .. '. . '. "..' " ' _ .. ,:, .. ,_ '" - .. ' Pa . ' :: ,-.. ' . , " ' ' /p I P A ; :: ~ : ~ I " ' ,', "':"::" -, ' ,,' ,< ' : _ A Zone 3 Pa > Pv > P ' ,' . A . .~ , ". Zone 2 ~ -- Blood flow Figure 8. Blood flow through the lung in the supine position. Model to explain the uneven distribution of blood flow ln the lung based on the pressure affecting the capillaries (Adapted from West et al., 1964). N +::. 25 exceeds venous pressure which exceeds alveolar pressure. Because venous pressu~e is higher than alveolar pressure throughout this zone, the pressure inside the collapsible vessels will be every­where higher than the pressure outside. The blood flow is increased in this zone and it is this location investigators cite as the opti­mum position for the Swan-Ganz catheter tip (Tooker et al., 1978; Cassidy, Robertson, Pierce & Johnson, 1978; Shasby, Dauber, Ander­son, Pfister, Carson, Manart & Hyers, 1981). Because of the high blood flow to this area, balloon tipped flow-directed catheters usually locate in this zone. West's (1979) model shows that blood flow in the lung is af­fected by the intrapleural, arterial, alveolar, and venous or left atrial pressures. A change in any of these pressures can affect the pressure measured as Pw. For example, when alveolar pressure is increased as with positive pressure breathing on a ventilator or with positive end-expiratory pressure (PEEP), the measured Pw may reflect alveolar pressure rather than pulmonary capillary pres­sure due to collapse of the capillary vessel. This occurs as alve­olar pressure is increased to greater than venous or arterial pres­sure, and Zone I or II results. Even when a patient does not have an artifically increased airway pressure, Zone II is of particular importance. The catheter may locate in this area, as there is blood flow, but as the balloon is inflated, occluding pulmonary artery blood flow, the arterial, or driving, pressure becomes zero, and since alveolar pressure remains constant, the capillary vessel collapses (Tooker et al., 1978) (See Figure 9). Zone II is present pul monary a t'tery Swan Ganz catheter >.... • pulmonary capillary inflated balloon (A) (8 ) Figure 9. Pw measurements in Zone II. Model to explain the effect occlusion of the lmonary artery pressure and blood flow has on the pulmonary capillary in Zone II. (A) Swan Ganz catheter in zone II with balloon deflated (Pa ~ P > Pv). (8) catheter in zone II with balloon inflated results in Zone I or callapse of capillary (P A > Pa > Pv). N ('"'f\ 27 in the anterior chest, or above the left atrium, when the patient is supine. If the Swan-Ganz catheter tip is positioned in this zone (See A Figure 9) the alveolar pressure and not the Pw or left atrial pressure is measured when the balloon is inflated (See B Figure 9). Investigations have shown that the Pw does not accurate­ly reflect LVEDP or left atrial pressure, and the Pw is of limited value when the catheter is located in Zone II (Pace, 1977; Tooker et al., 1978; Cassidy et al., 1978; Roy, Powers, Feustel & Dutton, 1977). Roy et al. (1977) and Shasby et al. (1981) recommended that lateral chest x-rays be taken to establish proper catheter tip po­sition, i.e., below the left atrium. Catheter placement informat~ can also be obtained by attempting to aspirate capillary blood frc ' the wedged position. Blood cannot be withdrawn from the wedged catheter in either Zone I or II because the vessel collapses when the balloon is inflated, the driving pressure is occluded. Thus important positioning problems can be identified by attempting to aspirate capillary blood. If the catheter is located in Zone I m- Il, it would be necessary to reposition it to obtain accurate Pw J • Rationale for Study All authors agree that errors in the dside Pw measurement are possible, and do occur (Shin et al., 1977). The incidence of technical problems has been reported to be 30% (Morris & Holt, 1980) in a research oriented lCU. There has been no research reporting the incidence in a regular lCU, and no studies to determine how la the errors are and if they are clinically significant. Confirming the Pw requires additional time, skill and expense. If errors at" significant, and result in incorrect clinical decisions, then this ditional effort is justified. There may also be other routes that are simpler and less expensive which can increase Pw accuracy, sucr as correcting technical problems without capillary blood aspiratior This investigation was designed to measure the Pw as routinely done in the lCU and then carefully re-measure the Pw using the criteria used by cardiologists in the cardiac catheterization laboratory. T routine Pw measurement (unconfirmed Pw) and the carefully analyzed measurement (confirmed Pw) were compared to determine the incider __ and significance of differences. A model (See Figure 10) illus-trates the major concepts and their interrelationships. Research Questions 1. Is there a clinically significant difference in the Pw measurement when errors are caused by easily correctable technical problems? 2. Does the aspiration of capillary blood (Criterion 5) improve the accuracy of the Pw measurement? 3. Is there a clinically significant difference in the confirmed Pw and the unconfirmed Pw? Role of the Nurse The answers to these questions are important to nurses and phys cians. Nurses make most Pw measurements. They titrate therapies u ing the Pw as a guide. In larger hospitals, nurses often teach fellows, residents and interns how to use monitoring equipment and make Pw measurements. In accepting these responsibilities, nurses must also take steps necessary to improve the quality of Pw meas-urements if indicated. Unconfirmed ~ ...... , 1 . < PAP ~ , 2. Change in waveform ......",,- ...... Confirrred Pw 1. Pw< PAP inically Significant Difference? I J / t7 ./" __ --ltC"./" ,/" ....... , ""- ,/" Pw-- /" Measure- -i>Patient Assessment- -t;>Cl inical 1. left ventri- Decisions cular function 1. Diag- 2. fluid status nosis 2. Ther­apy 2. Atrial waveform 3. Free-flow 5. Aspiration of capillary blood Identify and correct possible errors to: 1. technical problems 2. catheter position Fiqure 10. Conceptual model. The Pw measurement is clinically used as a reflection of LVEDP. Two methods for measuring the Pw with a Swan-Ganz catheter are described: One method, the unconfirmed Pw is the current ICU practice. The second method, the confirmed requires careful analysis of the Pw waveform and aspiration of capillary blood. confirmed Pw requires identification and correction of problems. These problems can be the source of error in fue recorded Pw value. The question raised in this study is ther or not there is a clinically significant difference between the confirmed and the unconfirmed Pw. Specifically, are errors being made in diagnosis and treat­ment of patients because clinical decisions regarding left ventricular function a fluirl status a}"e made using a method of PvJ measurement that is inaccurate? r0 1O CHAPTER II METHODOLOGY Design A single group pretest-posttest design was utilized and is schematically shown below. x One group pretest - posttest design PI is the unconfirmed Pw value. P2 is the confirmed Pw value. X represents the Pw. ps required to meet the criteria of a confirmed A sequential sample of patients requiring hemodynamic monitor-ing with a Swan-Ganz catheter was obtained. Patients were selectee from the Intermountain Respiratory Intensive Care Unit (IRICU) and the Shock-Trauma Intensive Care Unit (S-TICU) at the LDS Hospital in Salt Lake City, Utah. No patient was excluded on the basis of diagnosis, age, sex, or malfunctioning Swan-Ganz catheter. In all instances patients were assessed for stability. No mea­surements were included if: the time between PI and P2 was greater than three hours, blood pressure or heart rate changed by 20% or more, fluid intake or output changed by more than 20%, diuretics 31 were administered after PI or if a change was made in the rate or administration of any drug affecting vascular or cardiac dynamics. An attempt was made to make all P2 measurements immediately after Pl' These measures decreased the chance of intervening variables influencing the P2 values. Equipment Strip recorder (Model 2007, Gould, Inc., Instrument Division, Cleveland, OH 44114). Direct writing 2-channel pen brush recorders were util-ized for all pressure tracings. The recorders have a fixed gain with only a zero adjustment which el~minated the need to recali-brate prior to each measurement. A zero adjustment was made prior to each pressure measurement and was recorded on every pressure tracing. Transducers (P23ID-Gould, Inc., Statham Instruments Div., Oxnard, CA 93030 or Model 800-Bentley Trantec, Irvine, CA 92705). Strain gauge transducers were used which have a fixed sensitivity. They are routinely calibrated and checked for accuracy prior to sterili­zation between patients using a known 100 mmHg pressure on the transducer. Balloon tipped flow-directed catheters (Swan-Ganz) (Catalog No. 44166, Instrumentation Laboratory, Inc. Lexington, MA 02173). Standard 7 French four lumen thermal dilution 32 balloon tipped flow-directed catheters were used. They are rou­tinely used in the two critical care units at a rate of 24 catheter per month. The balloon was checked for symmetry and competency be-fore insertion. Continuous flush system (CFS - 03, Intraflow~ Sorenson Research Company, Salt Lake City, UT 84115). This system, commonly used in critical care units maintains a continuous fluid flow through the catheter and allows for dynamic response testing of the catheter system. The test for dynamic response (flush) verifies the unobstructed flow through tr.e catheter and the lack of air bubbles or leaks in the system. Blood gas machine (Radiometer ABL 3). All blood gases were analyzed in the pulmonary laboratory by trained blood gas technicians. Over 3,00) blood gases are analyzed monthly in this laboratory. The machine are calibrated every two hours according to manufacturer's in-structions. All hemodynamic monitoring equipment is maintained and cali­brated by an in-house biomedical instrumentation department on a three month preventative maintenance program. Definitions Conceptual Pulmonary capillary wedge pressure (Pw): the pressure in thl­pulmonary capillary bed. The pressure is equal to the left ven- tricular end-diastolic pressure when the patient has a competent mitral valve, a continuous column of blood is present from the left ventricle to the pulmonary capillary, and blood flow from the right heart is totally occluded. It is a reflection of left ventricular, pressure or filling, and fluid volume. Confirmed Pw: the measurement of Pw utilizing specific criteria which have been identified to verify the Pw. The criteria verify the separation of the pulmonary artery pressure, an unobstructed flow or communication from the catheter tip to the left ventricle during diastole, as well as freedom from technical problems with the measurement. Unconfirmed Pw: the measurement of Pw by the common ICU pro- . cedure of inflating the balloon on the Swan-Ganz catheter, looking for a change in the waveform and recording the value when the wave­form changes. Not all of the specific criteria to confirm a Pw are used. Operational Pulmonary capillary wedge pressure (Pw): measurement of the pressure in the pulmonary capillary bed as recorded with a balloon tipped flow-directed (Swan-Ganz) catheter. The pressure is mea­sured in mmHg with a strain gauge transducer properly zeroed to atmospheric pressure at the mid-axillary line. A recording of the pressure is made with the zero point identified on the tracing. The pressure is read at end-expiration. Pw as measured using the following criteria: 34 a) t balloon is inflated (This causes the catheter tip to become wedged in a small pulmonary vessel); b) the Pw must be lower than the mean pulmonary artery pressure (PAP); c) the phasic Pw record­ing must be consistent with an atrial waveform; d) free-flow should be present with the catheter in the wedged position (Identified by three acceptable "flushes"); e) highly oxygenated blood is aspirated from the catheter when in the wedged position. Unconfirmed Pw; Pw as measured using the following criteria: a) the balloon is inflated; b) the phasic Pw is different than the pulmonary artery waveform; c) the Pw is lower than the mean pulmo­nary artery pressure (PAP). Data Collection Prior to each measurement the patient was placed in the supin: position and the transducer zeroed at the mid-axillary line. If the patient was agitated or physiologically unstable, measurements were postponed until the patient was quiet and stable. The trans­ducer was opened to air; the pressure amplifier and strip recorder were zeroed. PAP and Pw measurements were then made. Figure 11 is a flow chart of the steps taken to make Pw measurements. At Step A the PAP was recorded and the catheter balloon was inflated, (with not more than Ii ml 's of air) until the waveform met the cri­teria of an unconfirmed Pw. A recording was made for a period of at least four respiratory cycles, and the balloon deflated. After observing PAP, the balloon was reinflated and the Pw obtained again. The waveform was asses using confirmed Pw Criteria 1, 2, and 3. Figure 11. [ Confirmed ~TeChnical------ -{confirmed Problem Pw Not Confirmed Technical Problem Easily Corrected Not Confirmed Technical Problem -:\Chnical Pw L Pw Probler.1 measurements No Technical Probl A Step A L ~NO Technical Problem No Pw A No Pw !. Step B Step C Pw measurement Manipulation to 1st wed~ed Blood meet Criteria 1. Pw' PAP 2. Atrial v/aveform 3. Free fl 0,./ specimen Confirmed Not confirmed ! Step 0 Manipulation to meet Criterion 5. Aspiration capil­lal'y blood rmed Confirmed A Step E Subsequent wedged blood specimen(s) Flowsheet of steps taken to determine the incidence of clinically significant errors in the measurement of Pw. Step A. Unconfirmed Pw and analysiS of waveform utilizing con­firmed Pw criteria. Step B. Manipulations, when necessary, to meet criteria listed ude: removing air bubbles, tightening loose connections, withdrawing or advanci catheter. Step C. Attempt to aspirate blood in wedged position and obtain confirmed Pw. Step D. If capillary blood was not obtained the catheter was repositioned. S E. Subsequent attemots to confirm the Pw. w U1 36 At least four respiratory cycles were again recorded. Measure-ments were identified as belonging to one of three groups: techni-cal problem, no technical problem, or no Pw waveform. If any of the three criteria were not met an attempt was made to find the cause and correct it. Step B in the procedure consisted of manipu-lations to meet Criteria 1, 2, and 3. Blood was then aspirated from the distal catheter port while in the wedged position, in an at­tempt to meet confirmed Pw Criterion 5 (Step C). Fifteen ml IS of waste blood was removed in a sterile heparinized syringe (Suter et al., 1975) followed by 2 ml's of blood withdrawn for blood gas analysis. The waveform was then observed to make sure the cathete' was still in the wedged position and the Pw unchanged. The balloo: was then deflated, the 15 ml's of heparinized blood reinfused, and the catheter flushed. An arterial blood sample was obtained at th~ same time through an indwelling arterial catheter. Confirmed and unconfirmed Pw values as indicated on flowsheet (Figure 11), were then measured from the strip recordings at end-expiration and re-corded, as well as any technical problems which were identified and corrected. When technical problems could not be corrected without repositioning the catheter, a capillary blood sample was withdrawn, or attempted, before the catheter position was manipulated. The blood was considered capillary or wedged blood, and the Pw ........ , confirmed if Pw02 19 mmHg / Pa02; pHw .08 _ pHa; and/or PwC02 ./ 11 mmHg ~~ PaC02 (Brewster et al., 1973; Malstrom & Michas, 1953~ Chlup, Serf, Ourednik & Parkmannova, 1975). An example of capil-lary blood gas values is shown in Figure 12. It :~:; H U ::; F' I TAL H L. 0 I) [t Ci A::::, f~ E P C,t F: NJ:I. 25:~:5:=::":: 1 JUL 21 81 PH F'C02 Hcel:::: BE ; ::.IHMr-iL HI 7. 4~5 40.0 25.0 .-', c:- ..:: .. _' HB 21 1 ::::5 Gf:;:EC-i :::()2 ()2CT ~~Ct2 F,I"l 5i"i7 ';J P K .I F'L. / F' F' /"1 R / ::: F: 95 ~-: :~:~,. /~I NI~Jf.:M{~l_ lJIl-J "7. ::::5 :34.0 19.0 -2.5 17.0 1 ::;:.0 DR ELLIOT7 CO/MT r02i 0/ 1 6::: r71:~: 16.6 '~:O 1 J ,=, • () c· W l7 4 ,=, Ib Jf.J. ""1; t --::'1- rJ 4 .-::' 1 r= .-:. .-:. / (-) l~(-1) ':', 7 .-:. (-) J._ " i_I" .. '-.' • '_.' "_I _I • 40.-. _t II '_I • "_, ._'. ,_, ....:-. _.... ." ....:- _ • q,* 1_-J '.."J ...-:.;... .-.-.::... / 1 4' / (.-..) 11'.". ) . I ..."..::.. .• •r•: ~:I 21 18:05 A 7.28 71.0 32.6 3.5 15.4 1/ 0 42 J2 15.5 69 22/141 6 10/25 SAMPLE # 187 TEMP 37.0, BREATHING STArus IMV Figure 12. :::;EVERE M I XED CHRON I C AND ACUTE RE::::P I F:ATDRY AC I DO:;:; I::; ESTIMATED R-l SHUNT 58750,43~ (AT AV DIFF OF 4.81, 6.817 8.81) HYPOVENTIlATION IMPROVED A-A GRADIENT 295, A/A 12~ (EST ALV P02 337) :::;EVERE HYPOXEt-lIA BREATHING 02 AND C02 F<ETENT ION -.:t*CONT{~CT r'lD OF~: I~I\J Pw confirmation with capillary blood. Blood gas results illustrating confirmation of the Pw by blood withdrawn from the wedged catheter having a pHw· 08 ~ pHa; PwC02 11 mmHg. ~ PaC02; and Pwo2 2 19 mmHg Pa02 (w = wedged sample, A = arterial sampl p ). W --..J 38 If the capillary blood criteria were not met, the catheter was repositioned (Step D). Additional attempts were then made to conii' the Pw, using waveform analysis (Criteria 1, 2, and 3) and capillar~ blood aspiration (Criterion 5). Step E consisted of these subse­quent cathetermanipulations and Pw measurements until a confirmed Pw was obtained, if this could be done within three hours. When there was no technical problem and capillary blood obtained on the first attempt, only Steps A, B, and C were required. Demographic data and information about the patient's clinical state were also recorded. These included: age, sex, diagnosis, the critical care unit in which each patient was located, blood pressure, heart rate, cardiac and vasoactive drugs, intake and output. Data were collected daily when possible until three pair~ of Pw measurements were obtained from each patient. If a measure­ment pair could not be obtained due to patient care needs or in­stability problems, the data were discarded and collected the fol­lowing day_ CHAPTER III RESULTS Sample Characteristics Pulmonary capillary wedge pressures (Pw) were measured in 53 patients from the IRICU and S-TICU. Twenty-seven (51%) were from the IRICU and 26 (49%) were from the S-TICU. Mean age for all pa­tients was 57.6 (~ 19.79) with an age range from 15 to 85 years. Twenty-five patients were female (47%) and 28 were male (53%). Diagnoses varied considerably and are listed in Table 2. Forty-one patients had three measurements, eight patients had two measurements, and four patients had only one measurement. The patients who had fewer than three measurements did so because their catheters were in place for less than three days. Pw measurements A summary of the results of the Pw measurements as they oc­curred during each step of the procedure are illustrated in Figure 13. The results include, the incidence of technical problems, no Pw and confirmed Pw measurements. Incidence of Technical Problems One hundred and forty-three measurements of Pw were obtained from 53 patients. Evaluation of the Pw waveform and the dynamic Table 2 Diagnostic Classification and Placement of Study Patients: Shock-Trauma ICU (S-TICU) or Intermountain Respiratory ICU (IRICU) Diagnosis No. Patients No. Patients No. STICU IRICU Post-op 11 8 Heart Fa i 1 ure 6 4 COPD 0 4 ARDS 0 5 Trauma 6 0 Near Drowning 0 2 Sepis 4 0 Inhalation Inj. 0 2 Drug Overdose 0 1 Total Patients 27 26 40 Patients Total 19 10 4 5 6 2 4 2 1 53 ~confirmed o Technical -{confirmed Problem 12/15 (80) Pw 15/59 (25) Not Confirmed Technical Problem 15/15 (100) 59/143 (41) ~EaSi1Y Corrected Not Confirmed lechnical Problem 3/15 ( 44/59 (75~ . No Technical Conflrmed Pw measurements Pw No Technical Problem Problem 127 77/127 (61) . -{ conflrmed 143 78/143 (55) 34/50 (68) Not confirmed figure 13. , Step A Pw rreasurement Problem Not Confirmed ~NO Technical 50/127 (39) No Pw 5/6 (83) 16/50 (32) 6/143 (4) 1 , 4 • No PW.a .& Step B 1 Step C Step 0 Manipulation to 1st wedged Blood Manipulation to meet Criteria specimen meet Criterion 1. Pw < PAP 2. Atrial waveform 3. Free f1 01'1 5. A Step E Subsequent wedged blood specimen(s} The outcome of 143 Pw measurements during steps taken to confirm the Pw. One hundred and twenty-three were confirmed within three hours. At Step A, problems were identified and Pw measurements categorized. After an attempt to correct technical problems (Step B) Pw measuremen consisted of two groups; those ical problem and those with an unresolved technical problem. At a technical problem yielded capillary blood, while IS without a technical pr'oblem were confirmed. After repositioni catheter ( D) a high percentage of Pw measurements from both groups were confirmed ( tep E). All technical problems were resolved in Pwls con­rmed. Numbers in parentheses indicate percent of adjacent fraction. .f.::. I---' 42 response showed in the majority of measurements (55%) there were no technical problems. In 59 measurements (41%) a technical prob-lem was observed; and in six (4%) there was no Pw waveform (See Step A of Figure 13). The incidence of technical problems in the S-TICU and IRICU were compared. Of the 143 measurements, 66 were made on patients in the S-TICU and 77 in the IRICU. In the S-TICU 33 (50%) had technical problems, 28 (42%) had no technical problems and five (8%) would not wedge. By contrast, in the IRICU 26 (34%) had tech­nical problems, 50 (65%) had no technical problems and one (1%) would not wedge. Pw Measurements Following Correction of Technical Problems When a technical problem was observed or no Pw obtained when the balloon was inflated, steps were taken to correct the problem (Figure 13, Step B). These steps consisted of: removing air bub-bles, tightening loose connections, withdrawing or advancing the catheter 1 to 2 cm, increasing the balloon volume, or irrigating the catheter. Of the 59 Pw measurements assotiated with technical problems, 44 (75%) were easily corrected. Correction of these 44 minor technical problems, changed the Pw in 27 measurements; in 11 the change was greater than 4 mmHg. As illustrated in Figure 14, when these changes occurred the confirmed Pw was consistently lower than the unconfirmed Pw. Technical problems identified most fre-quently were: poor dynamic response 28 (20%), damped waveform 22 (1 ), and suspect Pw 19 (13%). Poor dynamic response, and su N 43 35 30 0 2 measurements 0 3 measurements 25 D 4 measurements V 5 measurements 20 1 5 10 5 o 5 10 15 20 25 30 35 Pw (1) Figure 14. Pw errors due to technical problems. Comparison of Pw values with an initial technical problem (Pw 1) and Pw values after resolution of easily correctible technical problems (Pw2). N = 44; p < .0023; r2 = 0.91. Ten (23%) corrections resulted in changes of ~ 4 mmHg; 7 (16%) 6 mmHg. The line of identi~ illustrates (Pw (2))is consistently lower after cor­rection of technical problem (Pw(l)). 44 Pw were most commonly associated with significant errors in the Pw. Some measurements had more than one problem associated with them. Table 3 shows the incidence of specific technical problems and the number of times the Pw changed 4 mmHg or more following their cor-recti on. In five of the six measurements where Pw was unobtainable, in-creasing the balloon volume to 1.5 cc, or advancing the catheter 1 to 3 cm easily corrected the problem. In one patient (whose cardiac output was less than 2.5 l/min) the Pw could not be ob-tained even though more than three hours was spent trying to wedge the catheter. Aspiration of blood from the wedged catheter Blood Aspiration After an attempt was made to correct technical problems, aspir-ation of blood from the wedged catheter was attempted in all Pw measurements (Figure 13 Step C). Fifteen measurements were associ­ated with technical problems which could not be easily resolved and in those a capillary blood sample was not obtainable. There were 127 measurements without known technical problems at the time of blood aspiration. Fifty (39%) of these measurements did not yield capillary blood. These 127 measurements included 83 Pw's with no technical problems initially and the 44 Pw's in which a technical problem was easily corrected. These 127 measurements are considered as one group because there was no significant Table 3 Distribution and Significance of Technical Problems in 143 Pw Measurements Total Measurements Associated with Pw change 143 > 4 mm Hg Technical Problem No. % Unresolved No. of % of tota 1 % Speci-without measure- measure- fic repositioning ments ments Problem Inadequate waveform 22 ( 1 5) 0 3 (2) (14 ) Poor dynamic response 28 (20) 9 6 (4) ( 16 ) Overinflation 6 (4 ) 0 0 Suspect Pw 19 (13 ) 6 10 (7) (53 ) ~ ( II 46 difference between the two groups using a student t-test (p = 0.51 in the outcome of the confirmed Pw (See Appendix II for categori­zation of each group). Aspiration of blood from the wedged catheter following manipulation When capillary blood confirmation was not made, the catheter was repositioned (See Figure 13 Step D) and wedged blood samples were then attempted again. As indicated in Figure 13, Step F, 12 of the 15 measurements with initially unresolved technical problems yielded capillary samples after repositioning. Three times the problem was not resolved within three hours and capillary blood could not be obtained. In the 12 measurements which were confirme~ (See Figure 15), the Pw changed ten times and six of the changes were greater than 4 mmHg. Of the 50 measurements with no technical problem in which cap~i-lary blood was not initially aspirated, 34 were subsequently con-firmed after catheter repositioning; 12 remained unconfirmed after three hours of manipulation and in four measurements follow-up was limited due to patient instability or care requirements. As illus-trated in Figure 16 when capillary blood confirmation was obtained in the 34 Pw measurements, the Pw changed in 18 measurements, six of these more than 4 mmHg. The confirmed Pw was consistently lower than the unconfirmed Pw. Probability of Accurately Measuring Pw One hundred and twenty-three (86%) of the Pw measurements were 47 35 o 2 measurements 30 25 20 N --15 3: Cl.. 10 5 o ~----~----~----~----~------~----------~----~ 5 10 15 20 p\v (1) 25 30 35 4C Figure 15. Pw errors due to technical problems related to catheter position. Comparison of Pw values with technical prob­lems which were not easily resolved and were uncon­firmed by capillary blood aspiration (Pw 1), with Pw values after repositioning the catheter to resolve the technical problems and confirm the Pw with capillary blood (Pw2). N = 12; P < .0056; r2 = .79. Six (50%) of the Pw 1 values were:> 4 mmHg; and 4 (33%) > 6 mmHg different than Pw~ values. The line of-rden­tity illustrates Pw 2 values consistently less than Pw 1 values. 48 2 measurements 3 measurements ~ 4 measurements ~20 3: (L 15 10 5 0 30 35 40 Figure 16. Pw errors identified by capillary blood aspiration. Comparison of Pw values without technical problems but unconfirmed by capillary blood aspiration and Pw values after repositioning catheter and confirmation with ca pi 11 a ry blood (Pw2). N = 34; P < .0049; r2 = 94. Five (l5~s) of Pw 1 values> 4 mm Hg different from Pw 2 values. The line of ldentity illustrates Pw 2 values are usually less than Pw 1 values. 49 ultimately confirmed. As summarized in Figure 17, the routine bed-side measurement of Pw was accurately measured within ± 4 mmHg 82% of the time (Accuracy for Pw values within ± 2 mmHg and ± 6 mmHg are also shown in Figure 17). When identification and cor­rection of technical problems were made, the ability to accurately measure the Pw increased. Correction of easily resolved technical problems identified errors of ~ 4 mmHg 23% (n = 10) of the time and improved the accuracy of the Pw to 91%. Correction of persis­tent technical problems by withdrawing the catheter to a central position and readvancing it, identified errors of ~ 4 mmHg 50% (n = 6) of the time, improving the ability to accurately measure the Pw to 96%. Aspiration of capillary blood identified otherwise undetected errors ~ 4 mmHg in 4% (n = 5) of the measurements. As demonstrated in Table 3, if careful waveform analysis and dynamic response testing are used (Confirmed Pw Criteria 1, 2, and 3) when making a Pw measurement, the probability of obtaining an accurate Pw can be as high as 96%. Poor dynamic response, and suspect Pw were problems identified as being sometimes difficult to correct and confirm with capillary blood, without catheter re­positioning. Although the overall incidence in measurement errors of ~ 4 mmHg is low (4 to 7%), when a poor dynamic response or suspect Pw problem is present, the incidence of an error is as high as 53% (See Table 3). 10 ~ 9 ::::I VI ~ 9 So­CL <l) ~ 8 01 C ""O<l) ~ ili 8 So­......,:;:, ~ ~ 7 So- <l) So- E 8 7 4- o o-~ I I '1--------_1 I I I y Bedside Routine Pw Measurement --.. "'" .. _l.~"",.'It:IIr,~_"~ ___________ ~~ 1,----------1 I ___ .-- ____ 1 After identification & correction of easi­ly resolvable tech­nical problems After correction of persistent technical problems by repositioning ca theter - ± 6 mmHg --- ± 4 mmHg -- ± 2 mmHg Confirmation by aspiration of capil­lary blood Figure 17. Probability of accurately measuring the Pw within ± 2 'mmHg, ± 4 mmHg, and ± 6 mmHg, The accuracy of routine Pw measurements are improved after identification and cor­rection of easily resolved technical problems is made. Pw accuracy is further improved as persistent technical problems are corrected by repositioning and confirmation by aspiration of capillary blood is made. U1 ("""'I CHAPTER IV DISCUSSION, RECOMMENDATIONS AND CONCLUSIONS The incidence and magnitude of errors in the bedside Pw measurement using catheterization laboratory criteria have been measured in this study. The causes of errors in Pw measurement and the steps required to correct them are also identified. Con-clusions as to the clinical significance of these findings can be made by answering the following questions: Is there a clinically significant difference in the Pw measurement when errors are caused by easily correctable technical problems? Does the aspiration of capillary blood improve the accuracy of the Pw measurement? And is there a clinically significant difference in the confirmed Pw and the unconfirmed Pw? Clinically Significant Errors Before attempting to discuss the significance of errors in the Pw measurement, a clinically significant error must be defined. Although unconfirmed Pw and confirmed Pw values may be statisti-cally significant, the more important clinical implications, may , be insignificant. Differences / 4 mmHg are considered clinically significant in this study. Errors of this magnitude can alter 52 clinical care, while changes of 1 or 2 mmHg can be attributed to normal variability in the Pw (Swan, 1975), even though they may prove to be statistically significant. The stability requirements and closeness of measurements were part of the research design to eliminate actual changes in the Pw of more than 4 mmHg. In a low pressure system such as the pulmonary capillary bed, an error of 4 mmHg is at least 30% of the normal value (normal Pw = 6 to 12 mmHg) where as in a high pressure system such as the systemic circulation a 4 mmHg is not significant. Paired confirmed wedge pressure measurements were available in 16 patients ,in this study. These measurements provide information regarding the Pw variability. As with Pw values used in this study~ measurements were made no more than three hours apart and patient stability was assured. Statistically, the first confirmed Pw measurements (Pw 1) were the same as the second confirmed Pw measurements (Pw 2). The mean of the first measurements (Pw 1) was 12.2 (± 6.18), while the Pw 2 measurements had a mean of 12.31 (± 6.22). Clinically, there was only one pair with a significant difference (5 mmHg) while 14 of the 16 pairs had a difference of <: 2 mmHg (See Figure 18). It is difficult to choose a specific value which would cause an error in patient therapy or diagnosis because the large number of possible situations which occur in patients whose Pw is being measured. Depending on the situation, errors of ± 4 mmHg mayor may not cause errors in clinical decisions. If the actual Pw was 6 mmHg instead of the measured 10 mmHg, there would be n,o change 53 5 10 15 20 25 30 35 40 Pw (1) Figure 18. Normal variation in Pw. Comparison of confirmed Pw measurements and confirmed Pw measurements made from one to three hours later. Stability requirements were met between Pw 1 and Pw 2. N = 16; p <: .669; r = .968. Pw 1 and Pw 2 are not statistically or clinically dif­ferent. Values are closely grouped, about line of identity. In one measurement (6%), the difference between Pw 1 and Pw 2 was 2. 4 mmHg. 54 in therapy, but if the Pw was actually 18 mmHg when it was mea­sured as 22 mmHg or 16 mmHg when it was measured as 20 mmHg, in-appropriate therapy may be instituted. Errors in excess of 4 mmHg would increase the chance of mismanagement. For example, one pa-tient in the study with a history of congestive heart failure was being treated with fluid restriction for a Pw of 16 mmHg. His cardiac output and urine flow were low. A poor dynamic response (flush) and venous blood were observed when an attempt was made to confirm the Pw. After the catheter was repositioned, his confirmed Pw was 9 mmHg. A unit of packed red blood cells was given, in-creasing his Pw to 12 mmHg and his cardiac output and urine flow improved. The error may be compounded. In many ICU·s left ventricular function curves are used as a guide to patient diagnosis and pro-gress. Left ventricular stroke work index is plotted against the Pw. As shown in Figure 19 errors of 4 mmHg in the Pw can alter ,- --_ -~-_ the diagnostic_~classification (LOS Hospital, 1982).--- Is There a Clinically Significant Difference in the Pw Measurement When Errors are Caused by Easily Correctable Technical Problems? The results indicate that technical problems do cause clinical­ly significant errors in the Pw measurement. Technical problems occurred 41% of the time, causing clinically significant errors ir 23% of those cases (See Figure 14). Errors due to technical pro-blems occurred more often than previously reported (Morris & Holt, 1980). The previous report from an investigation done in the IRICU 60 Ca.rd iac output interpretatlon map ---,- Hyperdynamic circulation 50-+- +-l ro OJ .J:l .......... C'-J E40-+- .......... E E 0') ~ Mild LV dysfunction Normal x 30-+- OJ u c Moderate LV dysfunction ~ ; 20 Severe LV dysfunction 3: OJ -~ 0 s- .p v/10-+- Hypovolemia and I I I Hypovolemic shock Cardiogenic shock I LV dysfunction I ( I ( I I I I I I r I 2 4 6 B 10 12 14 15 16 18 20 22 24 26 28 30 Mean Wedge Pressure (mmHg) Figure 19. Left ventricular stroke work index (LVSWI) plotted against Pw. An 4 mmHg can change the interpretation of left ventricular function. crease in LVSWI is due to the effect of Pw in the calculation: error in the Pw of The apparent in- ( Stroke volume LVSWI = .0136 X (mean systolic pressure Body suface area - PW)) U1 U1 56 showed a 30% incidence of technical problems. In the IRIeU, a small three-bed research unit, all Pw·s are measured from a strip recorder at end-expiration using the confirmed Pw Criteria 1, 2, and 3 to analyze the waveform before reading the Pw value. In the S-TICU a larger ten-bed unit, automated displays and oscilloscopes are observed to make unconfirmed Pw measurements. The routine use of careful waveform analysis in the IRICU probably led to the decreased incidence of technical problems. Twenty-five percent of the identified technical problems caused an error of 4 mmHg or more; 16%, 6 mmHg or more. A hi gh percent­age (75%) were easily correctable by irrigating the catheter, re­moving air bubbles, tightening loose connections, or adjusting bal­loon volume. The major factor seemed to be identifying the problem so it could be corrected. The most frequent technical problem was a poor dynamic response. As shown in Table 3, this problem was usually resolvable 19 of 28 times (68%). When it was not, the poor dynamic response was due to the tracing "peggingll on the bottom of the recorder and oscil­loscope before returning to the Pw waveform (See Figure 4). This response indicates that the catheter tip is occluded, or partially occluded, by either the vessel wall or a ball valve clot on the catheter tip. Correcting this problem required repositioning the catheter. If this type of dynamic response is due to a ball clot, it is usually present in the pulmonary artery waveform as well. Since in these cases blood cannot be withdrawn from the catheter in the pulmonary artery or wedged position, the cathether must be 57 replaced. An inadequate waveform was also frequently observed (15%). It was associated with a poor dynamic response and is often called a damped waveform in the clinical setting. Although common, it was not difficult to correct. Measuring pressure, drawing blood from the catheter, infusing medications, and manipulating the patient or the catheter may cause a damped waveform. Problems difficult to resolve were due to distally placed catheters or clotting. If left uncorrected, a clinically significant error in the Pw was measured 14% of the time. A suspect Pw was present in 12% of the initial measurements. This technical problem was associated with the highest percentage (54%) of clinically significant errors in the Pw. Without careful~ experienced observation, this problem is often difficult to identi7j. The catheter is only partially wedged, thus the measurement does net reflect the left atrial pressure; and the waveform is not atrial in appearance. The waveform may appear to have IIV" waves, however careful scrutiny utilizing the simultaneous ECG tracing, shows "V ll waves do not coincide with ventricular contraction. Suspect Pw is probably the most common cause of clinically significant errors in Pw because it is not as easily identified as other technical prob­lems. When identified, 65% were easily corrected by increasing the balloon volume to the maximum 1.5 cc or advancing the catheter 1 to 3 cm. The product information inserts supplied with Swan-Ganz cathe­ters may be a factor in the frequent occurrence of this problem. 58 The Edwards Laboratories, Inc. insert suggests inflation of the balloon, only until the waveform changes from that of the pulmo­nary artery to avoid overinflation. IIInflation must be stopped immediately when the pulmonary artery pressure tracing is seen to change to pulmonary capillary wedge pressure 'l (Edwards Laboratories, Inc., 1980). Although rupture of pulmonary vessels has been re­ported, complications due to overinflation of the catheter balloon when only the maximum volume of air (1.5 cc) is used are not spe­cifically reported (Shin et al., 1977; Lemen, Jones & Cowan, 1975; Chun & Ellestad, 1971; Lapin & Murray, 1972). Rupture and hemor­rhage are usually associated with distally placed catheters. Thesl. complications can be controlled if the waveform is observed while the balloon is inflated. If the waveform indicates overinflation, the catheter is too distal and should be withdrawn a few centimete s. More commonly however, partially wedged waveforms which look dif­ferent from the pulmonary artery waveform, but are not a true Pw are the difficulty. If the catheter is not completely wedged, the recorded pressure will be higher than the actual Pw because part or all of the pressure from the pulmonary artery will be measured instead of the Pw illustrated in Figure 20. When there is a ques­tion about a catheter being only partially wedged, aspiration of blood from the wedged catheter for confirmation is useful. The blood will not be capillary if blood flow from the pulmonary artery is only partially occluded. The other technical problem which occurred, ballon overinfla­tion (4%), was not associated with clinically significant errors. 30- 20- 10- 0- GB 13M LDSH#328392 10 July 1978 2300 HRS VT= 0.76 P= 76/72/25 END EXPIRATION BALLOON VOLUME (CC) o 0.75 1.0 1.25 Figure 20. Partially wedged catheter. Incomplete occlusion of the pulmonary artery pressure due to inadequate amounts of air used to inflate the balloon (.75 ee and 1.0 ee). Pw obtained when 1.25 eels of air is used to inflate balloon. Ul tD 60 This problem is so obvious, demanding correction before any reason-able Pw can be obtained, that it is probably not a factor in Pw measurement errors. Overinflation indicates the catheter is located too distally in the pulmonary circulation and should be withdrawn to a more central position to avoid permanently wedging and occlud-ing pulmonary bloodflow, or the complication of rupture. The most significant errors in the Pw occurred when technical problems could not be easily corrected. Fifty percent of the mea-surements had a difference of 4 mmHg or greater. None of these Pw measurements could be confirmed with capillary blood without withdrawing the catheter to the right ventricle or main pulmonary artery and readvancing it. Since this group showed the highest incidence of clinically significant errors it is important that when technical problems are present and not easily resolved, the catheter be repositioned until the problem is corrected. When re-positioned, a confirmed Pw was obtained 80% of the time, indicating the problem was due to improper catheter position. It is important that technical problems be identified and cor-rected. Routinely utilizing the confirmed Pw Criteria 1, 2 and 3 or careful waveform analysis is relatively easy, inexpensive, and can increase the accuracy of the Pw measurement to 95%, if errors due to respiratory variation, zeroing and gain are also eliminated. Does the Aspiration of Capillary Blood "Improve ~~e Accuracy of the Pw Measurement? A significant number of measurements (35%) were not confirmed 61 with capillary blood (Criterion 5) after analysis of the waveform and confirmed Pw Criteria 1, 2, and 3 had been met. However, ulti­mate confirmation of Pw resulted in clinically significant errors which would not have been identified by waveform analysis in only five measurements (3% of the total 143 measurements or 4% of the 123 measurements which were ultimately confirmed). The incidence of clinically significant errors identified by the addition ofcapil­lary blood aspiration does not warrant blood gas analysis with each Pw measurement. In cases where major clinical decisions are to be made on the basis of the Pw or the Pw does not fit with the patients clinical assessment and other parameters, capillary blood confirm­ation eliminates the 4 to 5% chance of a clinically significant er­ror, and should be done. Although the aspiration of capillary blood to confirm the Pw is physiologically sound and used in the cardiac catheterization laboratory, there have been no formal studies comparing the con­firmed Pw to the left atrial pressure (LAP). Eight simultaneous measurements of Pw and LAP in five post-operative cardiac patients were made (Chapman, 1981). These patients were being monitored with a Swan-Ganz catheter and a left atrial pressure line. The results suggest that when capillary blood confirmation is made, the Pw does reflect the pressure in the left atrium (See Table 4). However, when blood could not be withdrawn or when it was venous, the Pw did not reflect the pressure in the left atrium. 62 Table 4 Correlation of Confirmed Pw and Left Atrial Pressure in Five Post Cardiac Surgery Patients Pa t i ent Number LAP Pw Wedged Blood Gas A 17 18 capillary Bl 15 23 venous B2 17 22 unable to withdraw B3 10 13 arterial Cl 23 24 capillary C2 17 15 capi llary 0 21 20 capi lla ry E 10 10 capillary Note. Pw confirmation attempted with aspiration of blood from wedged catheter. Is there. a Clinically Significant Difference in the Confirmed Pw --a-nd the Unconfi rmed PvJ? 63 These findings have shown that clinically significant differ­ences do occur in the Pw measurement when using the unconfirmed Pw criteria as routinely used in most critical care areas. The incidence is 18%. This value may be an underestimate assuming as it does, that there are not errors due to improper zeroing, cali-bration, or respiratory variation. It also assumes that the prob­lems which were identified and corrected (90) are representative of the unconfirmed Pw measurements which could not be confirmed. The impact of these measurements is not known. The high percentage of accuracy (83%) in the Pw measurement, without identifying or correcting problems, may explain why it is used successfully as a guide in patient management. However, using Pw confirmation criteria can increase the accuracy of the Pw mea-surement. The waveform analysis criteria used in this study showed the incidence of significant errors could be decreased by 9% when simple problems were corrected and 16% when all technical problems were corrected. This could increase the accuracy of the Pw measurement to 95%. An additional 5% of clinically significant errors were detected by capillary blood aspiration when no problem was identified in the Pw waveform. Areas for Further Study 1. A study assessing whether capillary blood aspiration en-sures Pw is a reflection of the left atrial pressure needs 64 to be done to confirm the observations suggested by only five patients (Chapman, 1981) and Scharf et a1 IS., (1977) findings in dogs. This investigation is currently in pro­gress at LOS Hospital. 2. Blood gas data need to be reviewed and capillary blood criteria for use in critically ill patients identified. The criteria currently available has been derived from pa­tients being treated in the cardiac catherization labora­tory. It has been suggested that the pH and PC02 may be better indicators of capillary blood than the oxygen par­ameters (Brewster & McIlroy, 1973). Observations made dur­ing this research support this suggestion. When only one of the three criteria for capillary blood confirmation was met it was almost always the pH. There were 32 of the 123 capillary blood samples in which only one or two of the established criteria were met. When this occurred, pH met the criteria in 25 samples; PC02 in 13; and P02 in 12. 3. Research has shown that the posterior or dependent zone of the chest is the optimum catheter position for accurately measuring the Pw. It is suggested that lateral x-rays be taken to confirm catheter tip position and increase the 1ike1ihoodthatthe Pw is actually reflecting LAP (Tooker et a1., 1978; Shasby et a1., 1981). Further study is needed to document this theory and whether capillary blood con­firmation correlates with catheters located dependently. 4. Further studies are also warranted to determine the effect of pulmonary shunts, FI02 and PEEP, on the ability to accurately measure and confirm the Pw. 65 5. Computer programs to analyze the frequency response of the PAP and Pw waveforms should be developed to provide prac­titioners information regarding the waveform; specifically to determine freeflow of catheters and damped waveforms. Recommendations 1. Pw measurements with a Swan-Ganz catheter in the critical care setting should include: careful waveform analysis, including assessment of dynamic response, in the Pw position. This reduces the chance of a clinically signifi­cant error to 5%. Optimum analysis requires the use of a strip recorder. If a recorder is unavailable, an oscil­loscope may be used in making these assessments. 2. Technical problems, especially suspect or partial wedge, damped waveform, and poor dynamic response should be cor­rected before the Pw measurement is made. The data shows that the presence of technical problems predicts a rela­tively high incidence of error. 3. The balloon of the Swan-Ganz catheter should be inflated with the maximum recommended volume while observing the waveform for Pw and overinflation. If overinflation occurs, the volume should be decreased and the catheter withdrawn a few centimeters. Maximal inflation can decrease the in­cidence of the most clinically significant error: partial or suspect Pw. 66 4. If a partial wedge is suspected, the catheter should be repositioned and/or a blood sample should be drawn from the wedged position. Withdrawing capillary blood indicates complete occlusion of venous blood from the pulmonary ar­tery and that the catheter is properly wedged. 5. Capillary blood aspiration should be done to confirm the Pw when making a clinical decision in which the Pw value is critical, or when the Pw value does not correlate with the patients other measured parameters or clinical state. There are clinically Significant errors which are undetected unless capillary confirmation is made. 6. Nurses and physicians making Pw measurements should be educated in the use of the catheter and monitoring equip­ment, waveform analysis, potential problems and errors in the Pw measurement and how to correct problems when they occur. The incidence of clinically significant errors is underestimated in this study since errors caused by zero­ing, calibration and respiratory variation are not included. Clinically significant errors due to respiratory variation have been found to be 40% in spontaneously breathing pa­tients and 8% in patients being ventilated with a mechani­cal ventilator (Cengiz, 1980). With careful measurement technique and equipment management these errors can be minimized. Conclusions Clinically significant errors occur in the bedside measurement 67 of Pw with methods currently used in the critical care setting. When the criteria from the cardiac catheterization laboratory are used in the critical care area, the incidence of errors can be re­duced. When errors occur they usually resul t in an overestimation of Pw. The automated equipment currently used in critical care areas does not identify or correct for technical, catheter position, or respiratory variation problems. Waveform analysis, including dynamic response testing by medical personnel, can significantly improve the accuracy of the Pw mea surement when problems are i dent i­fied and corr~ted. Aspirating blood from the wedged catheter does identify unde­tected clinically significant errors in 4% of the Pw measurements, but this additive confirmation is not warranted with each measure­ment of Pw in the critical care setting. These findings indicate that bedside measurement of Pw can re­main relatively inexpensive and simple. Medical personnel must be educated, however, in the aspects of waveform analysis and trouble shooting when making Pw measurements. Critical care standards should be set requiring Pw waveform analysis to be as routine as it is with the electrocardiogram. APPENDIX I PROTOCOL FOR HEHODYNAMIC MEASUREMENTS The protocol for data collection once the subjects are iden­tified is: 1. Ascertain criteria for acceptance into study. 1.1 Admission to the IRICU, S-TICU 1.2 Placement of a Swan Ganz catheter in one of the above units. 2. Determine and record 2.1 Date/Time of Catheter insertion 2.2 Demographic data 2.2. 1 2.2.2 2.2.3 2.2.4 2.2.5 2.2.6 Age Sex Diagnosis (admission and current) Thoracic compliance Cardiac output Pulmonary R/L shunt 3. Measurement procedure 3.1 Unconfirmed Pw measurement 3.1.1 Measure BP, HR 3.1.2 Record current cardiovascular drugs, fluid infusions, diuretic therapy, cardiac output, ventilatory support and output. 3.1.3 Place patient in supine position. 3.1.4 Zero transducer at mid-axillary line 3.1.5 Record zero on 2-channel strip recorder 3.1.6 Record pulmonary artery waveform 3.1.7 Inflate balloon on Swan Ganz catheter with 1 1/4 to 1 1/2 cc of air 3.1.8 Observe waveform on recorder. When waveform changes record at least three consistent stable respiratory cycles. 3.1.9 Deflate balloon 3.2.0 Measure the Pw at the end-expiratory phase of the respiratory cycle on the waveform. 3.2.1 Record value. Record time. 3.2 Confirmed Pw measurement 3.2.1 Repeat steps A-l through A-5 3.2.2 Observe the pulmonary artery (PA) waveform and "flush ll three times. 3.2.2.1 Identify a good PA waveform and 69 IIflush li then move to 3.2.3 if a good PA waveform cannot be identified, continue to 3.2.2.2. 3.2.2.2 Identify damped waveform or flush Correct with one or more of the follow­ing: 3.2.2.2.1 Check for adequate pressure and fluid in flush system. 3.2.2.2.2 Flush catheter several times with the Intraflo. 3.2.2.2.3 Irrigate catheter with a 3cc syringe of normal saline. 3.2.2.2.4 Look for and remove air bubbles from transducer dome or Intraflo. 3.2.2.2.5 Look for and correct leaks at connections 3.2.2.3 Identify atrial or partially wedged waveform. Correct by one or more of the following: 3 . 2 . 2 . 3 . 1 Ma k e sur e b all 0 0 n i sin f 1 at­ed. 3.2.2.3.2 Flush catheter with Intra­flo. 3.2.2.3.3 Pull catheter back 1-2 cm or until PA waveform appears. 3.2.2.4 Identify right ventricular waveform. Correct by: 3.2.2.4.1 Inflate balloon 3.2.2.4.2 Advance the catheter to the wedged position. 3.2.2.4.3 Deflate balloon. 3.2.3 Inflate balloon with 1 1/4 to 1 1/2 cc air while observing waveform. 3.2.3.1 Identify atrial waveform and Pw less than PAP. If present move to step 3.2.4, if not proceed to 3.2.3.2. 3.2.3.2 Identify "overinflation -- grad-ual increase of pressure off the top of scale. Correct by one or more of the following: 3.2.3.2.1 Deflate and reinflate the balloon. 70 3.2.3.2.2 Let a small amount of air out of the balloon until an atrial waveform appears. 3.2.3.2.3 Deflate balloon and pull the catheter back 1-2 em then reinflate. 3.2.3.3 Identify partial wedge, Pw greater than PAP or systolic-diastolic looking wave­form. Correct by one or more of the following: 3.2.3.3.1 Determine if the balloon is intact. 3.2.3.3.1.1 Able to withdraw to­tal volume of air inserted into bal­loon. 3.2.3.3.1.2 Resistance to infla­tion is felt. NOTE: If balloon is broken the catheter must be replaced by a physician if further Pw measurements are to be made. 3.2.3.3.2 Inflate balloon and advance until atrial waveform is ob­served. 3.2.3.4 "Flush" catheter three times 3.2.3.4.1 Identify an adequate flush Move to 3.2.3.5, if not, continue to 3.2.3.4.2. 3.2.3.4.2 Identify poor flush. Cor­rect by one or more of the following: 3.2.3.4.2.1 Deflate and re­inflate bal­loon. 71 3.2.3.4.2.2 Flush catheter again with Intraf10 3.2.3.4.2.3 Flush catheter with 3 cc syringe of nor­ma 1 sa 1 ine. 3.2.3.4.2.4 Pull catheter back 1-2 cm. 3.2.3.5 Record at least three consistent, stable respiratory cycles. 3.2.3.5.1 If variable end-expiratory pressures are observed, correct by one of the fol­lowing: 3.2.3.5.1.1 Make sure the patient is stable and not agi­tated or coughing 3.2.3.5.1.2 Deflate and rein­flate the balloon. 3.2.3.5.1.3 Make sure that there is not a slow leak in the balloon (See 3.2.3.3.1). 3.2.3.6 With balloon still inflated, aspirate 15 cc of fluid and blood (waste) from the stopcock connected to the distal port of the catheter with a 20 cc sterile heparinized syringe. This clears the catheter and pulmonary artery of saline and venous blood (21). Move on to step 3.2.3.7. 3.2.3.6.1 If unable to aspirate blood from wedged position, cor­rect by one or more of the following: 3.2.3.6. 1 . 1 Defl a te and re i n­f1ate balloon. 3.2.3.6.1.2 Irrigate catheter with 3 cc of nor­mal saline in syringe. 3.2.3.6.1.3 Deflate balloon and attempt to aspirate blood. If unable to get blood, there is probably a blood 72 clot on the catheter tip. NOTE: A physician must replace the catheter if Pw is to be measured. 3.2.3.6.1.4 Reposition catheter. 3.2.3.6.1.5 Decrease PEEP 3-5 cm. 3.2.3.6.1.6 Change patient's po-sition. 3.2.3.7 Aspirate 3 cc of blood from wedged cathe­ter with a heparinized blood gas syringe. 3.2.3.8 Flush by opening valve to clear blood from catheter. 3.2.3.9 Check to see that the Pw waveform is still atrial. If it is not, repeat step 3.2. 3.2.4.0 Deflate the balloon and reinsert waste fluid. 3.2.4.1 Measure and record the Pw as in 3.2.0 and 3.2.1. Note time. 3.2.4.2 Immediately place the blood sample in an iced slurry and send to Pulmonary Lab for analysis. (NOTE: These measurements will be done in conjunction with the need for as­sessment of arterial blood gases so that a comparison between the arterial and capillary blood oxygen can be made). 3.2.4.3 Pw02 19 mmHg higher than Pa02; pH\'1 .08 greater than pH ; PCO? (arterial) pwCO? less than PCO? tvenous) PaC02 criterionof capillaryb100l1. Ifit is not, repeatsteps 1 through 12 above. 3.3 The total time spent trying to confirm the Pw will not exceed three hours during anyone day. 3.3.1 Stability Requirements BP HR C.O. Fluids Output No change > 20% Cardiovascular Drugs: no new or no change Diuretics: none administered within three hours or between measurements. Ventilatory support: no change 73 3.4 All information will be recorded on the data flow sheet at the time of measurement. A copy of all blood gas reports will be attached to the flow sheet. 3.5 The measurements will be made: 3.5.1 3.5.2 Within 12 hours of insertion. Daily to coincide with routine measurements and a-v blood gases until three random mea­surements are obtained. APPENDIX II PW MEASUREMENTS CATEGORICALLY ILLUSTRATED 15 Tech. Prob STICU 6) 59 IRICU 9) Tech Prob (STlCU 33) (IRICU 26) ® 44 Resolved TP (STrCU 27) (IRICU 17) 5 Reso 1 ved (STlCU 1) 6 (IRICU 4) 143 Pw No Pw (STICU 66) ~STICU Sf (IRICU 77) (IRICU 1) Unresolved (STlC 1 U 1) 78 No Tech Prob. {STICU 28! (IRICU 50) o Con fl rmcd 23 lLncon firmed 21 Conf 1 rmed Unconfirmed 23 3 Confirmed Unconf 1 rmed 2 53 Confl nned Uncon finned 23 12 Con firmed ® 3 Unconfirmed 13 Confirmed © L Uncon ( i rmcd 10 Uncon! irmed 2 21 ,C.on-f -1 rf.l"'o.Cd- --- ® UnconfIrmed 4 Figure 21. Pw measurements categorically illustrated with actual Pw values and differences after corrections were made at Points A,B,C,D. 75 Figure 21 Continued Inter- Difference in # Measure- % Actual Pw Values vention Pw after Cor- ments rection ~O mmHg = 44 100 6-6 36-30 14-14 2-2 > 2 mmHg = 12 27 18-18 18-18 10-9 11-11 ® 2-4 mmHg = 10 23 18-18 16-16 14-12 12-13 L6 mmHg = 8 18 14-14 10-10 23-30 11-11 L8 mrnHg = 3 7 7-7 20-20 8-8 26-22 9-8 10-10 12-12 28-20 16-16 18-12 2-2 10-7 6-6 10-5 20-20 14-14 12-12 12-5 14-14 14-10 22-14 10-10 10-10 14-r) 6-5 30-30 > 0 mmHg = 12 100 17-14 18-20 ?2 mmHg = 9 75 12-12 30-20 ® ..:L4 mmHg = 6 50 10-6 24-22 L6 mmHg = 4 33 10-6 20-12 L8 mmHg = 2 17 14-13 12-12 ..:Ll0mmHg = 1 8 16-9 16-9 LO mmHg = 13 100 16-14 10-8 8-8 CD L2 mmHg = 6 46 18-20 10-10 30-28 2-4 mmHg = 1 8 10-6 11-10 14-14 L6 mmHg = 0 0 20-18 20-20 10-10 14-14 ~O mmHg = 21 100 10-2 11-12 22-14 L2 mmHg = 8 38 13-14 6-6 18-16 ® L4 mmHg = 4 19 3-3 18-17 14-13 L6 mmHg = 2 9 28-26 14-14 16-15 L8 mmHg = 2 9 18-18 30-26 10-10 2-10mmHg = 0 0 8-8 14-10 14-14 20-20 12-10 26-28 ~O mmHg = 34 100 CD & :L2 mmHg = 14 41 :L4 mmHg = 5 15 ® ~6 mmHg = 2 5 L8 mmHg = 2 5 REFERENCES Archer, G. & Cobb, L.A. Long term pulmonary artery pressure moni­toring in the management of the critically ill. Annals of Surgery, 1974, 180, 747-752. Brewster, H. & McIlroy, M.B. Blood gas tensions and pH of pulmo­nary "wedge" samp1es in patients with heart disease. Journal of Applied Physiology, 1973, 34, 413-416. Brown, R.S. Hemodynamic monitoring in the community hospital ICU. Critical Care Medicine, 1977, ~, 101-104. Cassidy, S.S., Charles, H., Robertson, J.R., Pierce, A.K., & Johnson, R.L., Jr. Cardiovascular effects of positive end­expiratory pressure in dogs. Journal of Applied Physiology, 1978, 44, 743-750. Cengiz, M.C. Effects of ventilation of accuracy of automated monitoring systems for the monitoring of pulmonary artery and wedge pressures. Unpublished master1s thesis. University of California, San Ftancisco, 1980. Chapman, R. Comparison of pulmonary capi11ary wedge pressure and left arterial pressure with capillary blood confirmation. Un­published manuscript (in progress). Salt Lake City: LOS Hospital, Pulmonary Division, 1982. Chlup, J., Serf, B., Ourednik, A., & Parkmannova, A. Partial pres­sures of oxygen and carbon dioxide, and pH of blood sampled from wedged pulmonary artery. Clinical Science and Molecular Medicine, 1975, 48, 47-50. Chun, G.M., & Ellestad, M.D. by a Swan-Ganz catheter. 1971, 284, 1041-1042. Perforation of the pulmonary artery New England Journal of Medicine, Dexter, L., Hayes, F.W., Burwell, C.S., Eppinger, E.C., Sagerson, R.P., & Evans, J.M. Studies of congenital heart disease II. The pressure and oxygen content of blood in the right auricle, right ventricle, and pulmonary artery in control patients, with observations on the oxygen saturation and source of pulmonary capillary blood. Journal of Clinical Investigation, 1947, 26, 554-560. I I Forrester, J.S., Diamond, G., Chatterjee, K., & Swan, H.J.C. Medi­cal therapy of acute myocardial infarction by application of hemodynamic subsets (First of two parts). New England Journal of Medicine, 1976, 295, 1352-1356. Gardner, R.M. Direct blood pressure measurement - dynamic response requirements. Anesthesiology, 1981, 54, 227-236. Gardner, R.M. Computer help sector. (Unpublished computer pro­gram). Salt Lake City: LOS Hospital, Department of Bio­physics, 1982. Gardner, R.M., Bond, E.L. & Clark, S.S. Safety and efficacy of continuous flush systems for arterial pulmonary artery cathe­ters. Annals of Thoracic Suroery, 1977, ~, 534-538. Hurst, J.W., Logue, R.B., Schlant, R.C., & Wenger, N.K. The Heart (4th Ed.). San Francisco: McGraw-Hill Book Company, 1972, pp. 479-514. Lalli, S.M. The complete Swan-Ganz (Reprint of article). RN, September 1978, 1-15. Langerlof, H., & Werko, L. Studies on the circulation of blood in man. VI. The pulmonary capillary venous pressure pulse in man. Scandinavian Journal of Clinical Laboratory Investiga­tion, 1947,1, 147-161. Lapin, E.S., & Murray, J.A. Hemoptysis with flow-directed bal­loon- tipped cardiac catheterization. Journal of the American Medical Association, 1972, 220, 1246. Lemen, R., Jones, J.G., & Cowan, G. A mechanism of pulmonary artery perforation by Swan-Ganz catheters. New England Journ­al of Medicine, 1975, 292, 211-212. Lozeman, J., Powers, S.R., Older, T., Dutton, R.E., Roy, R.S., English, M., Marco, D., & Eckert, C. Correlation of pulmo­nary wedge and left atrial pressures. Archives of Surgery, 1974, 109, 270-277. Malstrom, G., & Michas, P.A. Oxygen and carbon dioxide partial pressures in blood from the systemic circulation and the pulmonary capillaries. Acta Medica Scandinavia, 1953, CXLV, 91-96. Mendel, D. A practice of cardiac catheterization. Oxford: Black­well Scientific Publications, 1968, p. 234. Morris, A.H., & Holt, R. Pulmonary wedge pressure: A difficult measurement. American Review of ~espiratory Disease, 1980, -" 10 Pace, N.L. A critique of. flow directed pulmonary artery catheteri­zation. Anesthesiology, 1977, 47, 455-465. Rapapport, E., & Dexter, L. Pulmonary IIcapill aryll pressure. Methods in Medical Research, 1958, I, 85-93. Roy, R., Powers, S.R., Jr., Feustel, P.J., & Dutton, R.E. Pulmo­nary wedge catheterization during positve end-expiratory pressure ventilation in the dog. Anesthesiology, 1977, 46, 385-390. Scharf, S.M., Caldini, P., & Ingram, R.H., Jr. Cardiovascular ef­fects of increasing airway pressure in the dog. American Journal of Physiology, 1977, 232, H35-H43. Shapiro, H.M., Smith, G., Pribble, A.H., Murray, J.A., & Cheney, F.H. Errors in sampling pulmonary arterial blood with a Swan-Ganz catheter. Anesthesiology, 1974, 40, 291-295. Shasby, D.M., Dauber, I.M., Anderson, J.T., Pfister, S., Carson, S.B., Manart, F., & Hyers, T.M. Swan-Ganz catheter location and left atrial pressure determine the accuracy of the wedge pressure when positive end-expiratory pressure is used. Ches:, 1981, 80, 666-670. Shin, B., Ayella, R. J., & McAslan, T.C. Pitfalls of Swan-Ganz catheterization. Critical Care Medicine, 1977, ~, 125-127. Suter, P.M., Lindauer, J.M., Fairley, H.B., & Scholobohm, R.M. Errors in data derived from pulmonary artery blood gas values. Critical Care Medicine, 1975, 1, 175-181. Swan-Ganz flow-directed monitorinq catheters. Edwards Laboratory, American Hospital Supply Corporation. Anasco, Puerto Rico, 1980 .. Swan, H.J.C. The role of hemodynamic monitoring in the management of the critically ill. Critical Care Medicine, 1975,1,83- 89. Swan, H.J., Ganz, W., Forrester, J., Marcus, H., Diamond, G., & Chonette, D. Catheterization of the heart in man with the use of a flow-directed balloon tipped catheter. New England Journal of Medicine, 1970, 283, 447-451. Tooker, J., Huseby, J., & Butler, J. The effect of Swan-Ganz catheter height on the wedge pressure left atrial pressure relationship in positive pressure ventilation. American Re­view of Respiratory Disease, 1978, 117, 721-725. 79 Vij, D., Babcock, R., & Magilligan, D.J., Jr. A simplified con­cept of complete physiological monitoring of the critically ill patient. Heart and Lung, 1981, lQ, 75-82. Walinsky, P. Acute hemodynamic monitoring. Heart and Lung, 1977, .§.' 838-844. West, J.B. Respiratory physiology - the essentials (2nd Ed.), Baltimore: Williamsand Wilk"ins, 1979, p. 32-50. Woods, S.L. Monitoring pulmonary artery pressures. American Journal of Nursing, 76, 1765-1771.