Continuous Oxygen Saturation Monitoring during Cardiac Catheterization in Adults

Continuous Oxygen Saturation Monitoring during Cardiac Catheterization in Adults· Steven R. Dodson, M. D.;t Frederick A Hensley] n, M. D.;+ Donald E. Martin, M.D.;§ David R. Larach, M.D., Ph.D.;+ and D. Lynn Morris, M.D., F.C.C.P.II

Arterial hypoxemia may contribute to morbidity during cardiac catheterization. Therefore, we measured arterial hemoglobin oxygen saturation (SaO.) continuously using pulse oximetry in 29 patients (age range, 21 to 83 years) undergoing cardiac catheterization. Baseline SaO. was 96 ± 0.4 percent. All patients had a decrease in SaO. at some time during the procedure. Eleven patients (38 percent) had episodes of arterial hypoxemia, defined as a decrease in SaO. below 90 percent. In these 11 patients, the mean number of episodes of hypoxemia was 16 ± 7, and the mean duration of each episode was 53 ± 25 seconds. Multiple stepwise regression analysis showed that the

minimum SaO. during catheterization for any patient was signi6cantly associated with the baseline SaOI , duration of the procedure, and end-diastolic volume (EDV) as described by the following regression equation: minimum SaO. 46.8 - 0.0580 (duration of procedure in minutes) + 0.5362 (baseline SaO.) - 0.0159 (EDV~ Based on our 6nding of arterial hypoxemia in greater than one third of our patients, we would consider continuous SaO. monitoring or supplemental oxygen during cardiac catheterization, especially for those patients with poor ventricular function or low resting SaO. or those expected to have long procedures.

Cardiac catheterization has been associated with a variety of complications, including arrhythmias.!" myocardial infarction," hypotension.P-" and acute pulmonary edema." Hypoxemia may occur during cardiac catheterization and could contribute to morbidity as well as potentially alter physiologic parameters measured at catheterization. We know of only two published studies in which arterial hemoglobin oxygen saturation (SaOJ was continuously monitored during cardiac catheterization.v" Both of those investigations were conducted on pediatric patients and were performed to determine the reliability of noninvasive Sa02 monitors. Using pulse oximetry we examined the continuous beat-to-beat changes in Sa02 in an adult population during cardiac catheterization. Measurements of Sa02 by noninvasive pulse oximetry have been shown to be accurate over a wide range of saturations when compared to in vitro blood measurements of Sa02 using a spectrophotometric oximeter (CO-Oximeter). 10-12

tion Committee, we studied 29 patients undergoing cardiac catheterization. Indications for catheterization included the diagnosis of coronary artery (N = 12)or valvular heart disease (N =41 assessment of ventricular function (N = 41 and percutaneous transluminal coronary angioplasty (N =9). With the exception of one patient, who was catheterized emergently for unstable angina, all procedures were scheduled electively Patients were excluded if they were receiving supplemental oxygen upon arrival at the catheterization laboratory or prior to the start of the procedure. The SaO, was continuously monitored throughout the catheterization by means of a pulse oximeter (Nellcor N-l(0) and was recorded at four-second intervals (Nellcor N-9000 recorder), The pulse oximeters probe was placed on the patients left index finger, with the hand at heart level, as soon as the patient was transferred to the catheterization table. Baseline measurements of SaOI were recorded after the patient had been lying supine on the catheterization table for two to five minutes prior to any intervention. The pulse oximeters display was continuously monitored during the procedure to assure accurate tracking of arterial pulsations. Any change in Sa02 while the pulse oximeter was not tracking arterial pulsations correctly as determined by the oximeters display of pulse strength and pulse rate when compared to a simultaneous electrocardiogram, was not included. From the continuous recording of Sa02' the following data were derived for each patient: (1) lowest Sa02 persisting for at least eight seconds during the entire procedure; (2) number and duration of episodes during which Sa02 decreased below 90 percent; (3) the Sa02 immediately before and the lowest Sa02 in the one minute following ventriculography coronary cineangiography or balloon inflation during percutaneous transluminal coronary artery angioplasty The patients age, sex, weight, hematocrit reading, smoking htstory extent of coronary artery disease, location and extent of valvular disease, and signs of congestive heart failure before catheterization were recorded. Sedation before and during the procedure was administered at the discretion of the attending cardiologist and was recorded.

MATERIALS AND METHODS

After receiving approval from our institutional Clinical Investiga*From the Department of Anesthesia and the Division ofCardiolo~ Department of Internal Medicine, Pennsylvania State University College of Medicine, Hershey tResident in Anesthesia. +Assistant Professor of Anesthesia. IAssociate Professor of Anesthesia. IlAssistant Professor of Medicine. Manuscript received August 3; revision accepted January 4. Reprint requests: Dr. Hensley, Milton S. Hershey Medical Center, Pennsylvania State University, Hershey 17033 28

=

Oxygen saturation dUringCardiac Catheterization (Dodson st aI)

Central venous pressure, pulmonary arterial pressure, pulmonary capillary wedge pressure, left ventricular pressure, and arterial pressure were measured and recorded as the initial part of the catheterization. Cardiac output was measured by the thermodilution method, except in three cases in which the indocyanine green dye method was used. Due to the potential effects of intravascular dye on the accuracy of pulse oximeter readings, 13 no Sa02 measurements were considered for five minutes after green dye injection. Left ventricular volumes and ejection fraction were calculated from biplane ventriculography (right and left anterior oblique views~14 Contrast used for angiography was either Hypaque-76 or Omnipaque-350. Results are reported 84i the mean±SEM. The paired t-test was used to evaluate the changes in Sa02 occurring for each patient. The differences between patients were evaluated using multivariate regression analysis. Changes in Sa02 from reference values are reported as absolute percent changes. Statistical significance was defined as p
The patient population had a mean age of 53 ± 3 years (range, 21 to 83 years) and mean weight of 83 ± 3 kg. Nineteen patients were men, and ten were women. Eighteen patients (62 percent) had a history of smoking. Nineteen patients received sedation before or during catheterization (14 with diazepam [2 to 15 mg], two with morphine [3 to 9 mg], and three with combinations of diazepam and morphine). The mean duration of the procedure was 96 ± 5 minutes. The baseline Sa02 was 96±0.4 percent (range, 93 to 100 percent). All patients had a decrease in Sa02 from baseline at some time during catheterization. The lowest Sa02 for the population was 90 ± 1 percent (range, 82 to 96 percent; p
Patient No. 1

Longest No. Episode Total TIme Baseline Lowest Episodes Sa02<90% Sa02<90% Sa02 Sa02 (%) Sa02<90% (min) (min) (%)

1

2

9 6

5

86

3 4

6 7 8 9 10 11 Mean±SEM

2

1 10 4

27 8 22 16±7

4.77 4.77 93 0.75 3.15 95 98 11.83 24.00 0.27 0.47 96 93 23.08 1.72 97 0.40 0.40 94 0.33 2.27 95 2.23 0.97 91 40.58 5.00 95 0.20 1.13 97 4.92 13.95 2.83± 1.08 10.55±4.01 95±1

89

88 87 88 82 88 85

89 87

89

87 87±1

Table 2-Determinants ofMinimum SaO. during Cardiac Catheterization Regression CoefficientMinimum Sa02 -0.228 Age 0.013 Sex 0.023 Weight 0.555 (p <.01)· Baseline Sa02 -0.165 Intraoperative diazepam (mg) -0.661 (p<.OOI)· Duration of procedure (min) 0.224 Number of diseased coronary arteries -0.197 Valvular heart disease (yes/no) Symptoms of right ventricular failure (yes/no) -0.248 -0.298 Symptoms of left ventricular failure (yes/no) - 0.601 (p <.01)· End-diastolic volume -0.399 Pulmonary artery diastolic pressure -0.372 Pulmonary capillary wedge pressure 0.193 Cardiac index ·Identified as significant independent predictors of minimum SaO! by stepwise linear regression analysis.

of episodes of Sa02 below 90 percent among the 11 patients was 16 ± 7 (range, 1 to 86). The mean duration of these episodes was 53 ± 25 seconds (range, 8 to 240 seconds). The longest cumulative time below Sa02 of 90 percent for anyone patient was 40.58 minutes. The minimum Sa02 for any patient was found by stepwise multiple regression analysis to be significantly related to the baseline Sa02 (p<0.01), the duration of the procedure (p<0.001), and end-diastolic volume (EDV) (p<0.01), according to the following regression equation: minimum Sa02=46.8-0.0580 (duration of procedure in minutes) + 0.5362 (baseline SaOJ - 0.0159 (EDV in milliliters). This equation describes 69 percent of the variation in minimum Sa02. Other indices of left ventricular function, although related to minimum Sa02, were not independent predictors of minimum Sa02because of their close correlation to end-diastolic volume. The patients age, sex, weight, hematocrit reading before catheterization, smoking history extent of coronary artery or valvular heart disease, and dose of diazepam administered had no significant relationship to the minimum Sa02 (Table

2).

We evaluated patients for signs of congestive heart failure before catheterization. On physical examination before catheterization, five patients had evidence of right ventricular failure (peripheral edema, jugular venous distention, or hepatojugular reflux), while one patient had evidence of left ventricular failure (rales, S3' or abnormal chest x-ray Hlm), and two patients had evidence ofbiventricular failure. When the effect of coronary cineangiography left ventriculography and balloon inflation during percutaneous transluminal coronary artery angioplasty on Sa02were measured, two patients were excluded, one because he was having multiple episodes of desaturaCHEST / 94 / 1 / JUL~

1988

29

Table 3-SaOs Decreases during Angiography and Angioplasty

Coronary arteriography Angiocardiography Coronary angioplasty

No. Patients

Total No. Procedures*

27 17

182 21 67

9

No. Procedures* Associated with Absolute Decrease in Sa022:3%

No. Procedures* Associated with Fall in Sa02 <90%

Largest Single Decrease in SaO! (% Absolute)

10 6

0

4

5

4

0

5 4

*Coronary arteriography left ventricular angiocardiographg coronary angioplasty balloon inflation,

tion with such frequency that we were unable to establish accurate readings for 8a02 before angiography or baseline values before angioplasty and another because his peripheral pulses were not adequate to register accurate pulse oximeter readings during contrast injection. This population of 27 patients had 10 ± 1 (range, 4 to 20) single coronary artery contrast injections per patient, with 5 to 7 ml of contrast per injection. After combining contrast injections occurring within one minute of each other into one effective injection, these 27 patients had 7 ± 1 (range, 1 to 18) effective contrast injections per patient, for a total of 182. The largest decrease in 8a02 observed within the one minute following an injection was 5 percent in one patient, and another nine injections (in eight patients) were followed by a decrease in 8a02 of 3 percent or more. Of these ten injections, in four the 8a02 decreased below 90 percent, but for no more than 12 seconds (Table 3). Eighteen patients in this population of 27 had left ventricular angiography with two having supravalvular aortic angiography in addition. The amount of contrast injected was 42 ± 1 ml (range, 40 to 50 ml). We were able to accurately measure changes in saturation during angiography in 17 of these patients (one patient had too much motion artifact during angiography). These 17 patients underwent 21 contrast injections. The largest decrease in 8a02 observed within the one minute following an injection was 5 percent in one patient, and another five injections (in five patients) were followed by a decrease in 8a02 of 3 percent or more. During none of these six episodes did the 8a02 fall to below 90 percent (Table 3). Nine patients in this population underwent angioplasty of coronary lesions. These patients had 8 ± 2 (range, 2 to 14) balloon inflations per patient, for 30 to 60 seconds per inflation, Of71 total balloon inflations, the largest decrease in 8a02 following an inflation was four percent in one patient, and another three inflations (in two patients) were followed by a decrease in 8a02 of 3 percent. During none of these four episodes did the 8a02 fall to below 90 percent (Table 3). DISCUSSION

Arterial hypoxemia may be associated with arrhyth30

mias, myocardial ischemia, myocardial infarction, or congestive heart failure in the patient with coronary artery or valvular heart disease. It is difficult to predict, in any particular patient, the oxygen tension necessary to prevent morbidity; however, an arterial oxygen tension (Pa02) less than 60 mm Hg is considered by most authorities to be hypoxemic. A value for 8a02 of 90 percent corresponds approximately to a Pa02 of 60 mm Hg. In addition, 8a02 of 90 percent is at the steep portion of the oxygen-hemoglobin dissociation curve, where a small decrease in Pa02 will cause a large decrease in the amount of oxygen carried by hemoglobin. Therefore, we considered 8a02 less than 90 percent to represent arterial hypoxemia in our population. Previous studies have examined arterial blood desaturation in cardiac patients after receiving preoperative medication," cardiac patients undergoing preoperative placement of invasive monitoring lines, 16 and patients undergoing arterial and venous angiography 17 Various reasons have been suggested for the occurrence of arterial hypoxemia. In our study 11 (38 percent) of 29 patients had episodes of arterial hypoxemia. Minimum 8a02 was likely to be lower with increasing duration of procedure, lower baseline 8a02, and higher end-diastolic volume. No patient experienced angina or arrhythmias during these episodes, and no other clinical signs of morbidity were observed. Despite this lack of clinical signs of morbidity such episodes of arterial hypoxemia in other monitored settings, such as a coronary care unit, would warrant intervention. Further, supplemental oxygen administered using a nasal cannula has been shown to increase 8a02 in sedated patients undergoing invasive procedures in other settings. 16 Coronary cineangiography left ventriculography or coronary artery percutaneous transluminal balloon inflation rarely caused a decrease in 8a02 (Table 3). The tendency for 8a02 to increase following these maneuvers was remarkable. This increase could be due to the ventilatory maneuvers associated with angiography" or the effect of the contrast agent on oxygen-hemoglobin dissociation. 18 In an attempt to determine the factors that would be most predictive of arterial oxygen de saturation, we examined factors related to minimum Sa02 • Three Oxygen saturation during Cardiac Catheterization(Dodson at 81)

factors were found to be independent predictors: (1) low baseline Sa02 ; (2) the duration of the procedure; and (3) indices of left ventricular failure such as increased pulmonary arterial diastolic pressure, pulmonary capillary wedge pressure, end-diastolic volume, and decreased cardiac index. Five (63 percent) of the eight patients with symptoms or signs of heart failure before catheterization had episodes of arterial hypoxemia (Sa02 <90 percent) during the procedure. Thus, patients with congestive heart failure before catheterization or ventricular dysfunction found at catheterization are at increased risk of hypoxemia. The duration of the procedure was associated with the occurrence ofhypoxemia. Episodes of desaturation occurred in our population when sedated patients were lying quietly without stimulation, for up to 20 minutes, during preparation for catheterization. In our population, arterial hypoxemia occurred in more than one-third of the patients during cardiac catheterization. These results were observed in patients predominantly undergoing elective cardiac catheterization (the one patient who was not elective did not have episodes of arterial hypoxemia). A more acutely ill population may have a different incidence of arterial hypoxemia. Noninvasive monitoring of Sa02 by pulse oximetry does not require any calibration of instruments before it is used. Less than a minute is required to apply the noninvasive probe to the patients digit and determine the patients Sa02 • In the light of our findings of the frequency and duration of hypoxemia during cardiac catheterization and the factors which appear significant in causing this hypoxemia, we would recommend considering use of pulse oximetry during cardiac catheterization and suggest supplemental oxygen or continuous pulse oximetric monitoring during cardiac catheterization in patients who (1) have decreased left ventricular function evidenced by a history of congestive heart failure, an increased left ventricular enddiastolic volume, increased pulmonary capillary wedge pressure, or decreased cardiac index; (2) have low baseline Sa02 ; or (3) are expected to have a long procedure.

REFERENCES

1 Kennedy JW Complications associated with cardiac catheterization and angiograph~ Cathet Cardiovasc Diagn 1982; 8:5-11 2 Adams OF, Fraser DB, Abrams HL. The complications of coronary arteriography Circulation 1973; 48:609-18 3 Davis K, Kennedy M Kemp KG, Judkins MD, Gosselin AJ, Killip T. Complications of coronary arteriography from the collaborative study in coronary artery surgery (CASS). Circulation 1979; 59:1105-11 4 Adams OF: Abrams HL. Complications of coronary arteriography: a follow-up report. Cardiovasc Rad 1979; 2:89-96 5 Braunwald E. Severe hypotension. Circulation 1968; 37(suppl 3):54-6 6 Slack JD, Slack LA, Orr C. Recurrent severe reaction to iodinated contrast media during cardiac catheterization. Heart Lung 1982; 11:348-52 7 Bourassa MG, Noble J. Complication rate of coronary arteriography: a review of 5,250 cases studied by percutaneous femoral technique. Circulation 1976; 53:106-14 8 Krenkel R, Liappis N, Redel 0, Hildenbrand G. Non-invasive determination of oxygen saturation with the oxygenmet pulse wave oximeter: comparison with the invasive reftectometric method. lOin Paediatr 1981; 193:315-17 9 Bucher HU, Menner K, Huch A, Huch R. 'franscutaneous pO.measurement during heart catheterization in childhood. Z KardioI1977; 66:598-600 10 Yelderman M, New W Evaluation of pulse oximetry Anesthesiology 1983; 59:349-52 11 Swedlow DB, Stern S. Continuous non-invasive oxygen saturation monitoring in children with a new pulse oximeter (abstract). Crit Care Med 1983; 11:228 12 Mihm FG, Halperin BD. Noninvasive detection of profound arterial desaturations using a pulse oximetry device. Anesthesiology 1985; 62:85-7 13 Scheller MS, Unger RJ, Kelner MJ. Effects of intravenously administered dyes on pulse oximetry readings. Anesthesiology 1986; 65:550-52 14 Sandler H, Dodge H'I: Hay RE, ReckIey CEo Quantitation of valvular insufficiency in man by angiocardiography Am Heart J 1963; 65:501-13 15 Kopman EA, Ramirez-Inawat RC. Arterial hypoxaemia following premedication in patients with coronary artery disease. Canad Anaesth Soc J 1980; 27:132-34 16 Hensley FA, Dodson DL, Martin DE, Stauffer RA, Larach DR. Oxygen saturation during preinduction placement of monitoring catheters in the cardiac surgical patient Anesthesiology 1987; 66:834-36 17 Neagley SR, Vought MB, Weidner WA, Zwillich CW 'Iransient oxygen desaturation following radiographic contrast medium administration. Arch Intern Med 1986; 146:1094-97 18 Rosenthal A, Litwin B, Laver M. Effect of contrast media used in angiocardiography on hemoglobin-oxygen equilibrium. Invest Radio11973; 8:191-98

CHEST I 94 I 1 I JUL'(. 1988

31