- Email: [email protected]
Conjunctival oxygen tension monitoring during a controlled phlebotomy
Conjunctival Oxygen Tension Monitoring During a Controlled Phlebotomy MICHAEL KLEIN, MD, DEAN HESS, MEd, RRT, DAVID EITEL, MD, DENISE BAUERNSHUB, CRTT, NANCY SABULSKY, BS A decrease in the conjunctival oxygen tension (Pcjo,) and conjunctival index (Pcjo2/Pao2) has been shown to be an early marker of acute blood loss. We sequentially measured Pcjo,, Pcjoz/Pao,, blood pressure, and pulse rate in five healthy adults after controlled phlebotomy of 450 mL and after intravenous fluid repletion. No significant changes occurred in either the Pcjo, or PcjodPao, after phlebotomy or after fluid replacement. We conclude that a blood loss of 450 mL in healthy, euvolemic adults is insufficient to perturb the conjunctival index. The lower limits of sensitivity of changes in Pcjo2 and PcjoJPao, in response to acute blood loss remain to be established. (Am J Emerg Med 1988;5:11-13)
Miniaturization of a Clark-type oxygen electrode now permits continuous noninvasive measurement of tissue oxygen tensions. The palpebral conjunctiva, a thin nonkeratinized tissue bed whose blood supply is derived from the ophthalmic branch of the carotid artery, is uniquely suited for this purpose. In euvolemic humans, conjunctival oxygen tension (Pcjo,) correlates directly with arterial oxygen tension (Pao,), although the exact relationship between Pcjo, and Pao, varies among individuals.*F2 In hemodynamically compromised persons, Pcjo, no longer correlates with Pao, but instead correlates with tissue perfusion.3 The conjunctival index (obtained by dividing the Pcjo, by the concomitantly measured Pao,) behaves similarly to the conjunctival oxygen tension. In hemodynamitally stable, euvolemic humans, the conjunctival index varies from 0.41 to 0.98,1p2~4but it drops in hemorrhagic states.4J
From the Departments of Medicine and Emergency Medicine, York Hospital, and the School of Respiratory Therapy, York Hospital and York College of Pennsylvania, York, Pennsylvania. Manuscript 1987.
received 6 March 1987; revision accepted 8 May
Address reprint requests to Dr. Klein: York Hospital, Department of Medicine, 1001 South George Street, York, PA 17405. Key Words: Conjunctival phlebotomy.
index, hemorrhage,
0 1988 W.S. Saunders Company 0735-6757188 $0.00 + .25
oxygen tension,
During hemorrhage, tissue perfusion and oxygen delivery may be grossly perturbed before significant alterations are noted in the blood pressure and pulse.536 In this study, we examined the effect of acute loss of 1 unit (450 mL) of blood on Pcjo, and Pcjo,/ Pao, in otherwise normal persons. MATERIALS AND METHODS Five healthy volunteers (four men, one woman) aged 19 to 37 years without known cardiac disease, pulmonary disease, chronic anemia, recent blood loss, or conjunctivitis were studied. Informal consent was obtained from each person, and the study was approved by the Institutional Investigational Review Board. The conjunctival oxygen monitor (Insight Oxygen Monitoring System, NCC Division, Mallinckrodt Inc., Glenn Falls, NY) was prepared, membraned, and calibrated according to the manufacturer’s specifications. The sensor was inserted into the subject’s right lateral superior conjunctival fornix after the eye was anesthetized with two drops of 0.5% proparacaine hydrochloride. A 20-gauge arterial catheter (Arrow International Inc., Reading, PA) was inserted percutaneously into the radial artery using a modified Seldinger technique after collateral flow was confirmed by the Allen test. The catheter was intermittently flushed with a heparin flush solution (1 to 2 units of heparin/mL). Blood pressure was measured by auscultation, and the same sphygmomanometer was employed each time. Pulse rate was measured by palpation for 60 seconds. We chose to measure blood pressure by auscultation and pulse rate by palpation because these techniques are commonly used clinically. Mean arterial pressure (MAP) was calculated from a standard equation (MAP = [systolic + (2 x diastolic)]/3). One unit of whole blood (450 mL) was phlebotomized from an antecubital vein, according to standard blood banking procedures, within the hospital’s blood donation area. One liter of a lactated Ringer’s solution was then rapidly infused. Arterial blood was obtained anaerobically from the arterial catheter with a commercially available arterial blood gas kit (Concord Labs, Keene, NH). Samples were then packed in ice water, promptly delivered to 11
AMERICAN
JOURNAL
OF EMERGENCY
TABLE1. Data on Five Subjects
Subject 1 2 3 4 5 Mean
2 SD
MEDICINE
Who Underwent
Sex/Age
Weight
Blood Loss
(yr)
(kg)
(mukg)
Ml33 Ml37 M/35 F/19 Ml26 30 * 7
72.7 75.0 75.0 52.7 77.2 70.5 k 10.1
n Volume 6, Number 1 H January
Pcjo,/Pao, were not significant (p = 0.146, 0.277, and 0.072, respectively). The differences in the mean arterial pressure were small but significant (p = 0.015); the only significant difference between means was for pressures measured after phlebotomy supine and after phlebotomy erect. The differences in pulse rate were also small but significant (p = 0.001); significant differences between means was found for rates measured after phlebotomy supine and erect, after phlebotomy erect and after transfusion, and before phlebotomy and after phlebotomy erect. No subject experienced untoward side effects from the conjunctival oxygen sensor, indwelling arterial catheter, or intravenous lines. No orthostatic symptoms occurred.
Phlebotomy Percent Blood Loss
6.3 6.0 6.0 8.5 5.8 6.5 2 1.1
9.9 9.8 9.8 13.4 9.6 10.5 t 1.6
the blood gas laboratory, and analyzed in duplicate with the Corning Model 168 and 178 (Corning, Corning, NY) analyzers which were calibrated independently of each other. Data sets consisting of blood pressure, pulse rate, Pcjo,, and Pao, were obtained before phlebotomy (supine), after phlebotomy (supine and erect), and after fluid replacement (supine). Estimates of circulating blood volume were obtained by using previously determined regression equations derived from radiochromium erythrocytelabeling studies7p8 Analysis of variance with repeated measures was used to evaluate differences between matched variables in each data set and was performed with a commercially available software package (Human Systems Dynamics, Northridge, CA). Differences were considered statistically significant at the p < 0.05 level. When significant differences were found by analysis of variance, differences between pairs of means were evaluated with the T-method, with an experimentwide 0.05 error rate.9
DISCUSSION Diminution in Pcjo, and Pcjo,/Pao, has been shown to be an early sensitive marker of decreased tissue oxygenation in hemorrhagic states.*JOJ1 In our study, acute blood loss of 6.5 mL/kg (approximately 10.5% of estimated blood volume) in human subjects did not result in a drop in either Pcjo, or Pcjo,/Pao,. Additionally, assumption of an erect posture after phlebotomy did not affect these variables. Pulse rate did increase in the erect position after phlebotomy, although this variable was not clinically important. Likewise, a clinically insignificant increase in mean arterial pressure occurred in the erect position after phlebotomy. Abraham and coworkers noted that a reduced Pcjo,/Pao, accurately identified patients with occult blood loss, but only if the blood loss exceeded 15% of the total circulating blood volume.4 In sequentially phlebotomized dogs, both the Pcjo, and the Pcjo,/Pao, fell significantly before alterations in the heart rate or mean arterial pressure. However, this decrease occurred only after 15-mL/kg blood loss (approximately 18% of total blood volume).” During acute hemorrhage, circulating levels of catecholamines and angiotension increase, resulting in peripheral vasoconstriction and shunting of blood to the central circulation.12*13 Assumption of an erect posture after phlebotomy accentuates these responses
RESULTS Table 1 presents data for each subject, and Table 2 presents the mean Pcjo,, conjunctival indices, arterial pressures, and pulse rates before phlebotomy, after phlebotomy, and after fluid replacement. The mean blood loss was 6.5 t 1.1 (SD) mL/kg of body weight, corresponding to a calculated mean percent blood loss of 10.5 + 1.6%. The differences in Pcjo,, Pao,, and
TABLE 2. Conjunctival During Phlebotomy*
Oxygen
Tension
(Pcjo,),
Conjunctival
1988
Index (PcjodPao,),
Mean Arterial
Pressure
(MAP), and Pulse Rate
After Phlebotomy Before
Pcjo, (torr) Pao, (torr) PcjodPaO, MAP (mm Hg) Pulse (bpm) l
Data given as mean
12
Phlebotomy Supine
63 95 0.67 90 68 -t SD.
f -+ f f f
7 11 0.04 5 10
68 92 0.75 87 70
If- 12 t 11 -c 0.11 zk 6 r 10
After Transfusion Supine
Erect
Supine
67 98 0.68 92 87
2 2 a f k
11 9 0.07 5 16
67 2 7 98 * 6 0.69 +z 0.07 8826 67 f 10
KLEIN ET AL n CONJUNCTIVAL OXYGEN TENSION MONITORING
and probably resulted in the small increase in mean arterial pressure and pulse rate we noted. The loss of 1 unit of blood in euvolemic adults is not sufficient, however, to cause a drop in Pcjo, or Pcjo,/Pao,, because these homeostatic mechanisms are able to maintain cerebral (and conjunctival) blood flow. The small increase in Pcjo, after phlebotomy we observed supports this assertion. No clinically important differences were seen between the data measured after transfusion and those measured before or after phlebotomy. Both mean arterial pressure and pulse returned to baseline after transfusion, but whether this return was secondary to volume replacement or merely to resumption of the supine position could not be determined from our data. Transfusion did not affect Pcjo, or Pcjo,/Pao, because of potent homeostatic mechanism that compensate entirely for a blood loss of 450 mL and adequately maintain conjunctival tissue perfusion. In the emergency department, adult patients are sometimes seen in whom there has been a small, but clinically important, blood loss. Unfortunately, conjunctival oxygen monitoring may not be useful in the identification of these patients until loss of much more than 1 unit of blood has occurred. CONCLUSION Acute blood loss of 450 mL in healthy, euvolemic adults is insufficient to decrease Pcjo, or Pcjo,/Pao, significantly. More study is required to determine the lower limit of sensitivity of Pcjo, or Pcjo,/Pao, in response to acute blood loss. The authors thank Sharon Blankenstein and Cindy Kraut of the York Hospital Blood Bank for help in phlebotomizing the subjects, and the York Hospital Laboratory for prompt arterial blood gas analysis.
REFERENCES 1. Chapman KR, Liu FLW, Watson RM, et al: Conjunctival oxygen tension and its relationship to arterial oxygen tension. J Clin Monit 1988;2:100-104 2. Hess Cl, Evans C, Thomas K, et al: The relationship between conjunctival PO, and arterial PO, in 18 normal persons. Respir Care 1988;31:191-198 3. Abraham E, Smith M, Silver L: Conjunctival and transcutaneous oxygen monitoring during cardiac arrest and cardiopulmonary resuscitation. Crit Care Med 1984;12:419421 4. Abraham E, Oye RK, Smith M: Detection of blood volume deficits through conjunctival oxygen tension monitoring. Crit Care Med 1984;i 2:931-934 5. Deszi JM, Kim SI, Shoemaker WC: Sequential hemodynamic changes in an experimental shock preparation designed to simulate clinical shock. Ann Surg 1969;170:157-165 6. Kho KL, Shoemaker WC: Cardiorespiratory changes in acute hemorrhage. Surg Gynecol Obstet 1967;124:826832 7. Wennesland R, Brown E, Hopper J, et al: Red cell, plasma and blood volume in healthy men measured by radiochromium (Cr151) cell tagging and hematocrit: influence of age, somatotype, and habits of physical activity on the variance after regression of volumes to height and weight combined. J Clin Invest 1959;38:1065-1077 8. Brown E, Hopper J, Hodges JL, et al: Red cell plasma and blood volume in healthy women measured by radiochromium cell-labeling and hematocrit. J Clin Invest 1962;41: 2182-2190 9. Sokal RR, Rohlf FJ: Biometry: The Principles and Practice of Statistics in Biological Research, ed 2. San Francisco, Freeman, 1981, pp 242-247 10. Abraham E, Smith M, Silver L: Continuous monitoring of critically ill patients with transcutaneous oxygen and carbon dioxide and conjunctival oxygen sensors. Ann Emerg Med 1984;13:1021-1026 11. Smith M, Abraham E: Conjunctival oxygen tension monitoring during hemorrhage. J Trauma 1986;26:217-224 12. Hall RC, Hodge RL: Changes in catecholamine and angiotensin levels in the cat and dog during hemorrhage. Am J Physiol 1971;221:1305-1309 13. Bond RF, Johnson G: Vascular adrenergic interactions during hemorrhagic shock. Fed Proc 1985;44:281-289
13
Miniaturization of a Clark-type oxygen electrode now permits continuous noninvasive measurement of tissue oxygen tensions. The palpebral conjunctiva, a thin nonkeratinized tissue bed whose blood supply is derived from the ophthalmic branch of the carotid artery, is uniquely suited for this purpose. In euvolemic humans, conjunctival oxygen tension (Pcjo,) correlates directly with arterial oxygen tension (Pao,), although the exact relationship between Pcjo, and Pao, varies among individuals.*F2 In hemodynamically compromised persons, Pcjo, no longer correlates with Pao, but instead correlates with tissue perfusion.3 The conjunctival index (obtained by dividing the Pcjo, by the concomitantly measured Pao,) behaves similarly to the conjunctival oxygen tension. In hemodynamitally stable, euvolemic humans, the conjunctival index varies from 0.41 to 0.98,1p2~4but it drops in hemorrhagic states.4J
From the Departments of Medicine and Emergency Medicine, York Hospital, and the School of Respiratory Therapy, York Hospital and York College of Pennsylvania, York, Pennsylvania. Manuscript 1987.
received 6 March 1987; revision accepted 8 May
Address reprint requests to Dr. Klein: York Hospital, Department of Medicine, 1001 South George Street, York, PA 17405. Key Words: Conjunctival phlebotomy.
index, hemorrhage,
0 1988 W.S. Saunders Company 0735-6757188 $0.00 + .25
oxygen tension,
During hemorrhage, tissue perfusion and oxygen delivery may be grossly perturbed before significant alterations are noted in the blood pressure and pulse.536 In this study, we examined the effect of acute loss of 1 unit (450 mL) of blood on Pcjo, and Pcjo,/ Pao, in otherwise normal persons. MATERIALS AND METHODS Five healthy volunteers (four men, one woman) aged 19 to 37 years without known cardiac disease, pulmonary disease, chronic anemia, recent blood loss, or conjunctivitis were studied. Informal consent was obtained from each person, and the study was approved by the Institutional Investigational Review Board. The conjunctival oxygen monitor (Insight Oxygen Monitoring System, NCC Division, Mallinckrodt Inc., Glenn Falls, NY) was prepared, membraned, and calibrated according to the manufacturer’s specifications. The sensor was inserted into the subject’s right lateral superior conjunctival fornix after the eye was anesthetized with two drops of 0.5% proparacaine hydrochloride. A 20-gauge arterial catheter (Arrow International Inc., Reading, PA) was inserted percutaneously into the radial artery using a modified Seldinger technique after collateral flow was confirmed by the Allen test. The catheter was intermittently flushed with a heparin flush solution (1 to 2 units of heparin/mL). Blood pressure was measured by auscultation, and the same sphygmomanometer was employed each time. Pulse rate was measured by palpation for 60 seconds. We chose to measure blood pressure by auscultation and pulse rate by palpation because these techniques are commonly used clinically. Mean arterial pressure (MAP) was calculated from a standard equation (MAP = [systolic + (2 x diastolic)]/3). One unit of whole blood (450 mL) was phlebotomized from an antecubital vein, according to standard blood banking procedures, within the hospital’s blood donation area. One liter of a lactated Ringer’s solution was then rapidly infused. Arterial blood was obtained anaerobically from the arterial catheter with a commercially available arterial blood gas kit (Concord Labs, Keene, NH). Samples were then packed in ice water, promptly delivered to 11
AMERICAN
JOURNAL
OF EMERGENCY
TABLE1. Data on Five Subjects
Subject 1 2 3 4 5 Mean
2 SD
MEDICINE
Who Underwent
Sex/Age
Weight
Blood Loss
(yr)
(kg)
(mukg)
Ml33 Ml37 M/35 F/19 Ml26 30 * 7
72.7 75.0 75.0 52.7 77.2 70.5 k 10.1
n Volume 6, Number 1 H January
Pcjo,/Pao, were not significant (p = 0.146, 0.277, and 0.072, respectively). The differences in the mean arterial pressure were small but significant (p = 0.015); the only significant difference between means was for pressures measured after phlebotomy supine and after phlebotomy erect. The differences in pulse rate were also small but significant (p = 0.001); significant differences between means was found for rates measured after phlebotomy supine and erect, after phlebotomy erect and after transfusion, and before phlebotomy and after phlebotomy erect. No subject experienced untoward side effects from the conjunctival oxygen sensor, indwelling arterial catheter, or intravenous lines. No orthostatic symptoms occurred.
Phlebotomy Percent Blood Loss
6.3 6.0 6.0 8.5 5.8 6.5 2 1.1
9.9 9.8 9.8 13.4 9.6 10.5 t 1.6
the blood gas laboratory, and analyzed in duplicate with the Corning Model 168 and 178 (Corning, Corning, NY) analyzers which were calibrated independently of each other. Data sets consisting of blood pressure, pulse rate, Pcjo,, and Pao, were obtained before phlebotomy (supine), after phlebotomy (supine and erect), and after fluid replacement (supine). Estimates of circulating blood volume were obtained by using previously determined regression equations derived from radiochromium erythrocytelabeling studies7p8 Analysis of variance with repeated measures was used to evaluate differences between matched variables in each data set and was performed with a commercially available software package (Human Systems Dynamics, Northridge, CA). Differences were considered statistically significant at the p < 0.05 level. When significant differences were found by analysis of variance, differences between pairs of means were evaluated with the T-method, with an experimentwide 0.05 error rate.9
DISCUSSION Diminution in Pcjo, and Pcjo,/Pao, has been shown to be an early sensitive marker of decreased tissue oxygenation in hemorrhagic states.*JOJ1 In our study, acute blood loss of 6.5 mL/kg (approximately 10.5% of estimated blood volume) in human subjects did not result in a drop in either Pcjo, or Pcjo,/Pao,. Additionally, assumption of an erect posture after phlebotomy did not affect these variables. Pulse rate did increase in the erect position after phlebotomy, although this variable was not clinically important. Likewise, a clinically insignificant increase in mean arterial pressure occurred in the erect position after phlebotomy. Abraham and coworkers noted that a reduced Pcjo,/Pao, accurately identified patients with occult blood loss, but only if the blood loss exceeded 15% of the total circulating blood volume.4 In sequentially phlebotomized dogs, both the Pcjo, and the Pcjo,/Pao, fell significantly before alterations in the heart rate or mean arterial pressure. However, this decrease occurred only after 15-mL/kg blood loss (approximately 18% of total blood volume).” During acute hemorrhage, circulating levels of catecholamines and angiotension increase, resulting in peripheral vasoconstriction and shunting of blood to the central circulation.12*13 Assumption of an erect posture after phlebotomy accentuates these responses
RESULTS Table 1 presents data for each subject, and Table 2 presents the mean Pcjo,, conjunctival indices, arterial pressures, and pulse rates before phlebotomy, after phlebotomy, and after fluid replacement. The mean blood loss was 6.5 t 1.1 (SD) mL/kg of body weight, corresponding to a calculated mean percent blood loss of 10.5 + 1.6%. The differences in Pcjo,, Pao,, and
TABLE 2. Conjunctival During Phlebotomy*
Oxygen
Tension
(Pcjo,),
Conjunctival
1988
Index (PcjodPao,),
Mean Arterial
Pressure
(MAP), and Pulse Rate
After Phlebotomy Before
Pcjo, (torr) Pao, (torr) PcjodPaO, MAP (mm Hg) Pulse (bpm) l
Data given as mean
12
Phlebotomy Supine
63 95 0.67 90 68 -t SD.
f -+ f f f
7 11 0.04 5 10
68 92 0.75 87 70
If- 12 t 11 -c 0.11 zk 6 r 10
After Transfusion Supine
Erect
Supine
67 98 0.68 92 87
2 2 a f k
11 9 0.07 5 16
67 2 7 98 * 6 0.69 +z 0.07 8826 67 f 10
KLEIN ET AL n CONJUNCTIVAL OXYGEN TENSION MONITORING
and probably resulted in the small increase in mean arterial pressure and pulse rate we noted. The loss of 1 unit of blood in euvolemic adults is not sufficient, however, to cause a drop in Pcjo, or Pcjo,/Pao,, because these homeostatic mechanisms are able to maintain cerebral (and conjunctival) blood flow. The small increase in Pcjo, after phlebotomy we observed supports this assertion. No clinically important differences were seen between the data measured after transfusion and those measured before or after phlebotomy. Both mean arterial pressure and pulse returned to baseline after transfusion, but whether this return was secondary to volume replacement or merely to resumption of the supine position could not be determined from our data. Transfusion did not affect Pcjo, or Pcjo,/Pao, because of potent homeostatic mechanism that compensate entirely for a blood loss of 450 mL and adequately maintain conjunctival tissue perfusion. In the emergency department, adult patients are sometimes seen in whom there has been a small, but clinically important, blood loss. Unfortunately, conjunctival oxygen monitoring may not be useful in the identification of these patients until loss of much more than 1 unit of blood has occurred. CONCLUSION Acute blood loss of 450 mL in healthy, euvolemic adults is insufficient to decrease Pcjo, or Pcjo,/Pao, significantly. More study is required to determine the lower limit of sensitivity of Pcjo, or Pcjo,/Pao, in response to acute blood loss. The authors thank Sharon Blankenstein and Cindy Kraut of the York Hospital Blood Bank for help in phlebotomizing the subjects, and the York Hospital Laboratory for prompt arterial blood gas analysis.
REFERENCES 1. Chapman KR, Liu FLW, Watson RM, et al: Conjunctival oxygen tension and its relationship to arterial oxygen tension. J Clin Monit 1988;2:100-104 2. Hess Cl, Evans C, Thomas K, et al: The relationship between conjunctival PO, and arterial PO, in 18 normal persons. Respir Care 1988;31:191-198 3. Abraham E, Smith M, Silver L: Conjunctival and transcutaneous oxygen monitoring during cardiac arrest and cardiopulmonary resuscitation. Crit Care Med 1984;12:419421 4. Abraham E, Oye RK, Smith M: Detection of blood volume deficits through conjunctival oxygen tension monitoring. Crit Care Med 1984;i 2:931-934 5. Deszi JM, Kim SI, Shoemaker WC: Sequential hemodynamic changes in an experimental shock preparation designed to simulate clinical shock. Ann Surg 1969;170:157-165 6. Kho KL, Shoemaker WC: Cardiorespiratory changes in acute hemorrhage. Surg Gynecol Obstet 1967;124:826832 7. Wennesland R, Brown E, Hopper J, et al: Red cell, plasma and blood volume in healthy men measured by radiochromium (Cr151) cell tagging and hematocrit: influence of age, somatotype, and habits of physical activity on the variance after regression of volumes to height and weight combined. J Clin Invest 1959;38:1065-1077 8. Brown E, Hopper J, Hodges JL, et al: Red cell plasma and blood volume in healthy women measured by radiochromium cell-labeling and hematocrit. J Clin Invest 1962;41: 2182-2190 9. Sokal RR, Rohlf FJ: Biometry: The Principles and Practice of Statistics in Biological Research, ed 2. San Francisco, Freeman, 1981, pp 242-247 10. Abraham E, Smith M, Silver L: Continuous monitoring of critically ill patients with transcutaneous oxygen and carbon dioxide and conjunctival oxygen sensors. Ann Emerg Med 1984;13:1021-1026 11. Smith M, Abraham E: Conjunctival oxygen tension monitoring during hemorrhage. J Trauma 1986;26:217-224 12. Hall RC, Hodge RL: Changes in catecholamine and angiotensin levels in the cat and dog during hemorrhage. Am J Physiol 1971;221:1305-1309 13. Bond RF, Johnson G: Vascular adrenergic interactions during hemorrhagic shock. Fed Proc 1985;44:281-289
13