AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 8, NUMBER 4 July 1991
IMPACT OF VENOUS CATHETERS ON PACKED RED BLOOD CELLS Richard 0. Oloya, M.D., Harriet J. Feick, M.D., and Mary Ellen A. Bozynski, M.D., M.S.
This study was designed to test the hypothesis that there would be increased hemolysis, as indicated by an increase in plasma free hemoglobin and potassium, when packed cells were infused through small gauge percutaneous catheters (24 and 28 g, PC) compared with short catheters (24 g; SC). We were unable to study the 28 g PC because after 1 hour, at a flow rate of 10 ml/hr, only 2.4 ml packed cells were infused. There was a significant increase in plasma free hemoglobin when pre- and postinfusion values were compared (SC, p = 0.005; PC, p = 0.009), but a statistically significant increase in potassium only for the SC (p = 0.008). There were no significant differences between the catheters for either potassium or free hemoglobin. For either catheter the quantity of free hemoglobin transfused could potentially cause a significant rise in serum bilirubin and hemoglobinuria.
Premature newborns often require continuous intravenous access for prolonged periods. Steel "butterfly" needles were initially used in infants because of their availability in small gauges and their ease of placement; however, they were dislodged easily and had short half-lives (7.5 hours).1 The introduction of plastic cannulas increased the mean duration (33 hours) of peripheral intravenous infusions.2 In the extremely premature infant, surgically placed venous lines have been required for long-term vascular access. Recently, an alternative method for such access has been introduced, the percutaneous central venous Silas tic catheter.3"5 These catheters have been used for infusion of parenteral alimentation solutions, lipids, blood products, and medications. Concerns regarding transfusion-related hemolysis of packed red blood cells have arisen as smaller diameter and longer catheters are being used. Previous studies have evaluated transfusion-related hemolysis resulting from various factors, including small needle bore, 6>7 rate of flow through needles or large bore catheters,7 type of infusion pump, and type of blood product.8 Varying degrees of hemolysis have been demonstrated under different conditions.9-12 This study was designed to test the hypothesis that there would be increased hemolysis, as indicated by an increase in plasma free hemoglobin and potassium,
when packed red blood cells were infused through small gauge percutaneous compared to short catheters. METHODS
Whole blood was obtained from three volunteer donors in the usual fashion by the American Red Cross 3 days prior to the study date. The units were spun according to the routine protocol used in our blood bank using a Sorvall RC 313 centrifuge (Dupont Sorvall; Wilmington DE) at 4300 rpm for 10 minutes to produce packed red blood cells. The hematocrits for each unit were 88%, 94%, and 92%. The packed cells were drawn through a PDF-20 Pediatric Transfusion Filter Set (Fenwal Products, Deerfield, IL) into an IVAC model 700 syringe (IVAC Co., San Diego, CA) for infusion and into 10 ml syringes for baseline determinations. After the first run was set up, the remaining syringes filled with packed cells were refrigerated until one-half hour before use. The catheters compared included a 28 gauge L-Cath percutaneous catheter (0.4 mm diameter, 20 cm length), a 24 gauge L-Cath percutaneous catheter (0.6 mm diameter, 30.5 cm length); both made by Lumed, Santa Ana, CA), and a 24 gauge Quicke
Aultman Hospital, Canton, Ohio, and Section of Newborn Services, University of Michigan Medical Center, Women's Hospital, Ann Arbor, Michigan Reprint requests: Dr. Bozynski, Section of Newborn Services, University of Michigan Medical Center, L3023 Women's Hospital, 200, E. Hospital Drive, Ann Arbor, MI 48109-0254
280
Copyright © 1991 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
ABSTRACT
IMPACT OF CATHETERS O N PACKED RBC/Oloya, Feick, Bozynski
Statistical Analysis
The Wilcoxon test was used to test whether there were significant changes in plasma free hemoglobin and potassium values when postinfusion values were compared with preinfusion values for each catheter. The Mann-Whitney and Kruskal-Wallis tests were used to test for differences between catheters, runs, and for any effect due to differences between donors on change in free hemoglobin or potassium concentrations. The donor variable served as a proxy for differences in hematocrit, red cell characteristics, etc. Simple linear regression analyses with change in free hemoglobin and potassium concentrations as the dependent variables and catheter type, donors, and initial plasma free hemoglobin and potassium concentrations as predictors were performed. Because the study design required revision, the a priori power calculations were not applicable and therefore additional analyses were performed to calculate the change in free hemoglobin and potassium detectable by this study. All statistical tests were two-sided.
RESULTS
Preliminary observations. The occlusion-sensing disc required frequent adjustment when the L-Cath 28 gauge percutaneous catheter was tested. A blank sensing disc was used to bypass the pressure limit feedback system completely. Nevertheless, despite unlimited pressure, only 2.4 ml packed red blood cells was infused through the 28 gauge catheter in 1 hour at a flow rate of 10 ml/hr. Because of the length
of time required to infuse 10 ml of blood for each sample (about 4 hours), the 28 gauge L-Cath catheter was deleted from the study. The remainder of the study was performed comparing the 24 gauge percutaneous catheter (PC) to the 24 gauge short catheter (SC) in paired runs with pump-catheter combinations determined randomly. The occlusion-sensing disc required adjustment occasionally on the 24 gauge percutaneous catheter but was never bypassed completely. The study data are presented in Tables 1 and 2. There was a significant change when pre- and postinfusion plasma free hemoglobin values were compared for both the SC and the PC (SC, p = 0.005; PC, p = 0.009). Although the plasma potassium also increased when samples were infused through either catheter, statistical significance was achieved for only the SC (P = 0.008). No difference was found for the change in free hemoglobin when the two catheters were contrasted (median values for both catheters: SC 179, range 53 to 413, versus PC 133, range - 9 4 to 533, p = 0.86). Similarly, no difference for change in potassium was found (median values for both catheters: SC 1.2, range 0 to 3.0, versus PC 2.0, range -3.9 to 5.0, p — 0.75). Inexplicably, both plasma free hemoglobin and potassium decreased on one run comparing pre- and postinfusion values for the percutaneous catheter (Tables 1 and 2). This was probably a labeling error, but this could not be confirmed. The change in both free hemoglobin and potassium did not appear to be related to catheter type, donor, or the order in which the samples were run and processed. It required more time to infuse 10 ml packed cells through the PC than through the SC (PC me-
Table 1.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
Cath (0.6 mm diameter, 1.6 cm length; Travenol Labs, Deerfield, IL). The IVAC syringes were placed in the syringe pump. An extension set with an occlusion sensing device was placed between the syringe and the test catheter. The tip of the catheter was placed in a Vacutainer tube containing SST gel and clot activator (Bectin Dickinson, Rutherford, NJ) to facilitate collection and subsequent separation. The tube was surrounded by ice. A flow rate of 10 ml/hr was chosen, since this infusion rate is commonly used in premature infants. The preset pump pressure limit was 6 psi (41.4 kPa or 300 mmHg). If this pressure was exceeded, an alarm sounded. Pre- and post-transfusions samples were processed identically and spun at 3000 to 4000 rpm for 15 minutes using a Beckman Model TJ-6 centrifuge (Beckman Instruments, Brea, CA) to obtain plasma for potassium and serum hemoglobin measurements. The plasma potassium was measured using an ion-specific electrode on a Beckmann ASTRA with a standard deviation of ± 0.1 mmol/liter. The plasma hemoglobin was measured using a modified benzidine method13 with a error of ± 3 mg/dl at a free plasma hemoglobin of 25 mg/dl and ± 5 mg/dl at 100 mg/dl.
Pre- and Postinfusion Plasma Hemoglobin Values by Catheter Type and Donor Plasma Free Hemoglobin (mg/dl)
Donor
Preinfusion
Short catheter (n = 10) 1 48 44 1 1 45 1 45 2 359 2 134 75 2 3 120 3 115 75 3 Percutaneous catheter (n = 10) 1 31 I 45 1 46 1 46 2 312 2 123 68 2 3 120 87 3 90 3
Postinfusion
Change
227 137 97 125 772 667 152 376 398 156
179 93 52 80 413 533 77 256
521 178 179 176 218 312
490 133 133 130 -94 189
121 376 367 264
53 256 280 174
283 81
281
AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 8, NUMBER 4 Pre- and Postinfusion Plasma Potassium Values by Catheter Type and Donor Plasma Potassium(mmol/l) Donor
Preinfusion
Short catheter (n = 10) 1 11.0 I 11.0 I 11.0 12.0 1 2 14.7 2 12.9 2 12.9 15.6 3 17.4 3 14.7 3 Percutaneous catheter (n = 10) 1 11.0 1 11.0 1 11.0 1 11.0 2 15.3 2 12.0 2 12.6 15.0 3 14.7 3 15.3 3
Postinfusion
seuille equation where the resistance to flow (Q) is related to the dimensions, length (L) and radius (r), of the vessel or tube and the viscosity of the fluid
Change
Q = APr4/8L-n 14.0 13.0 12.0 13.0 16.2 17.1 13.5 16.8 17.7 14.7
3.0 2.0
16.0 14.0 13.0 12.0 11.4 14.1 12.6 15.6 17.7 15.6
5.0 3.0 2.0 1.0 -3.9 2.1 0.0 0.6 3.0 0.3
1.0 1.0 1.7 4.2 0.6 1.2 0.3 0.0
dian, 68 minutes, range, 64 to 79; p = 0.007; SC median, 63 minutes, range, 62 to 65). The study had a 90% probability of detecting a change in free hemoglobin of 82 mg/dl and a change in potassium of 0.85 mmol/liter. DISCUSSION
The results of this project showed a significant increase in plasma free hemoglobin when packed red blood cells were infused through either the percutaneous or short catheters. Although the plasma potassium also increased when samples were infused through either catheter, statistical significance was achieved for only the short catheter. When the percutaneous and short catheters were contrasted, no significant differences for changes in either plasma free hemoglobin or potassium levels were noted. This would tend to indicate that catheter length was not a major factor, whereas hematocrit and gauge were. Although the statistical difference between the catheters was not significant, the observed increase in free hemoglobin concentration has potential clinical significance. The normal full-term newborn usually produces 6 to 8 mg/kg of bilirubin per day.14 This quantity of bilirubin may result from the breakdown of 170 to 230 mg/kg of hemoglobin, the amount contained in a packed red blood cell transfusion of 10 ml/kg. In the extremely preterm infant, that is, an infant weighing less than 1 kg at birth, this amount of free hemoglobin could cause a significant rise in serum bilirubin and hemoglobinuria. 282 The model for blood flow is based on the Poi-
From this equation, it is apparent that with no change in catheter gauge or blood viscosity, an increase in catheter length should cause a decrease in flow. In our study of 24 gauge catheters we noted that it required 64 to 79 minutes to complete the 10 ml transfusion through the percutaneous compared with 60 minutes for the short catheter. Since the volume of the percutaneous catheter was approximately 0.2 ml only an additional minute difference in transfusion time would have been predicted. The failure of the packed cells to flow through the 28 gauge percutaneous catheter is also predictable given the fact that flow is related to the fourth power of the radius and thus a small decrease in catheter diameter from 24 to 28 gauge resulted in a large negative effect on flow despite an increase in infusion pressure. Another factor in the Poiseuille equation is viscosity. Blood viscosity is related to hematocrit, red cell characteristics, and plasma constituents. Although factors affecting hemolysis are incompletely understood, an increase in shear stress will cause an increase in hemolysis.17 Since viscosity is defined as the ratio of shear stress over shear rate, in a catheter system such as ours, a rise in viscosity would be expected to be associated with an increase in shear stress and hemolysis. The relationship between hematocrit and viscosity is not linear, but exponential.16 For example, lowering the packed cell volume from 90 to 70% using saline results in a decrease in relative viscosity from about five times to 1.3 times that of whole blood.18 Thus, the increase in hemolysis reported in our study compared with earlier studies using 27 gauge needles may reflect the higher hematocrits of our packed cell units. Our study was limited, since it was performed in vitro and did not evaluate such variables as the hematocrit, catheter material, or other factors that may significantly affect hemolysis. Nonetheless, the study was performed simulating routine administration of packed red blood cells to preterm infants. This differs from previous studies, which tended to reflect laboratory conditions and not clinical practice, such as use of relatively high infusion rates or large bore catheters.10 This study showed that significant hemolysis occurs during infusion of packed red blood cells with both 24 gauge short catheters and long percutaneous catheters. In fact, the variability in pretransfusion plasma hemoglobin values suggest that minor differences in the technique used in filling syringes may lead to large differences in pretransfusion plasma free hemoglobin. The variability could also be due to minor differences in assay; however, the error in plasma free hemoglobin measurement at the hemoglobin values found in our study would only account
Downloaded by: Universite de Sherbrooke. Copyrighted material.
Table 2.
July 1991
IMPACT OF CATHETERS O N PACKED RBC/Oloya, Feick, Bozynski 4. 5. 6. 7. 8. 9. 10.
11. 12. 13.
REFERENCES 14. 1.
Batton DG, Meisels MJ, Appelbaum P: The use of peripheral intravenous cannula in premature infants: a controlled study. Pediatrics 70:487-490, 1982 2. Johnson RV, Donn SM: Life span of intravenous cannulas in a neonatal intensive care unit. Am J Dis Child 142:968— 971, 1988 3. Durand M, Ramanathan R, Martinelli B, et al: Prospective evaluation of percutaneous central venous Silastic catheters in newborn infants with birth weights of 510 to 3920 grams. Pediatrics 78:245-250, 1986
15. 16. 17. 18.
Dolcourt JL, Bose CL: Percutaneous insertion of Silastic central venous catheters in newborn infants. Pediatrics 70:484-486, 1982 Dietrich KA, Lobas JG: Use of single Silastic IV catheter for cystic fibrosis. Pediatr Pulmonol 4:181-184, 1988 Wilcox GJ, Barnes A, Modanlou H: Foes transfusion using a syringe infusion pump and small-gauge needle cause hemolysis? Transfusion 21:750-751, 1981 Angel JL, O'Brien WF, Knuppel RA, et al: Infusion of packed erythrocytes: An in vitro study of hemolysis. Obstet Gynecol 69:948-950, 1987 Gibson JS, Leff RD, Roberts RJ: Effects of intravenous delivery systems on infused red blood cells. Am J Hosp Pharm 41:468-472, 1984 Nand S, Bansal VK, Kozeny G, et al: Red cell fragmentation syndrome with the use of subclavian hemodialysis catheters. Arch Intern Med 145:1421-1423, 1985 Mateer JR, Perry BW, Thompson BM, et al: Effect of rapid infusion with high pressure and large-bore IV tubing on red blood cell lysis and warming. Ann Emerg Med 14: 966-969, 1985 Herrera AJ, Corless J: Blood transfusion: Effect of speed of infusion and of needle gauge on hemolysis. J Pediatr 99: 757-758, 1981 Schmidt WF III, Kim HC, Tomassini N, et al: Red blood cell destruction caused by micropore blood filters. JAMA 248:1629-1632, 1982 Crosby W, Furth F: A modification of the benzidine method for measurement of hemoglobin in plasma and urine. Blood 11:380-383, 1956 O'Dell GB: Neonatal hyperbilirubinemia. New York: Grune & Stratton, 1980, p 37 Bacher RP, Williams MC: Hemolysis in capillary flow. J Lab Clin Med 70:485-494, 1970 Black VD: Neonatal hyperviscosity syndromes. Curr Prob Pediatr 17:79-130, 1987 Bacher RP, Williams MG: Hemolysis in capillaryflow.J Lab Clin Med 76:485-496, 1970 Mollison PL, Engelfreit CP, Contreras M: Blood Transfusion in Clinical Medicine. 8th ed. Oxford: Blackwell Scientific Publications, 1987, p 131
Downloaded by: Universite de Sherbrooke. Copyrighted material.
for a difference of 5 to 15 mg/dl. Our observed differences in pretransfusion values ranged from 4 to 200 mg/dl. The differences between the catheters were not statistically significant. Thus, either appears suitable for infusion of packed red blood cells under conditions described in our study. Even though there were no significant differences between the catheters the quantity of plasma free hemoglobin infused could potentially cause a rise in serum bilirubin and hemoglobinuria in the very preterm infant. Small catheters (28 gauge) may not be adequate for transfusion, even at high infusion pressures, or, alternatively, infusion through smaller caliber catheters may require lower hematocrits. This option should be explored, since transfusion in the micropremie whose only vascular access is a 28 gauge catheter necessitates establishing a new line. Establishing vascular access is stressful for the infant and the gain by transfusing higher hematocrits may be minimal.
283
IMPACT OF VENOUS CATHETERS ON PACKED RED BLOOD CELLS Richard 0. Oloya, M.D., Harriet J. Feick, M.D., and Mary Ellen A. Bozynski, M.D., M.S.
This study was designed to test the hypothesis that there would be increased hemolysis, as indicated by an increase in plasma free hemoglobin and potassium, when packed cells were infused through small gauge percutaneous catheters (24 and 28 g, PC) compared with short catheters (24 g; SC). We were unable to study the 28 g PC because after 1 hour, at a flow rate of 10 ml/hr, only 2.4 ml packed cells were infused. There was a significant increase in plasma free hemoglobin when pre- and postinfusion values were compared (SC, p = 0.005; PC, p = 0.009), but a statistically significant increase in potassium only for the SC (p = 0.008). There were no significant differences between the catheters for either potassium or free hemoglobin. For either catheter the quantity of free hemoglobin transfused could potentially cause a significant rise in serum bilirubin and hemoglobinuria.
Premature newborns often require continuous intravenous access for prolonged periods. Steel "butterfly" needles were initially used in infants because of their availability in small gauges and their ease of placement; however, they were dislodged easily and had short half-lives (7.5 hours).1 The introduction of plastic cannulas increased the mean duration (33 hours) of peripheral intravenous infusions.2 In the extremely premature infant, surgically placed venous lines have been required for long-term vascular access. Recently, an alternative method for such access has been introduced, the percutaneous central venous Silas tic catheter.3"5 These catheters have been used for infusion of parenteral alimentation solutions, lipids, blood products, and medications. Concerns regarding transfusion-related hemolysis of packed red blood cells have arisen as smaller diameter and longer catheters are being used. Previous studies have evaluated transfusion-related hemolysis resulting from various factors, including small needle bore, 6>7 rate of flow through needles or large bore catheters,7 type of infusion pump, and type of blood product.8 Varying degrees of hemolysis have been demonstrated under different conditions.9-12 This study was designed to test the hypothesis that there would be increased hemolysis, as indicated by an increase in plasma free hemoglobin and potassium,
when packed red blood cells were infused through small gauge percutaneous compared to short catheters. METHODS
Whole blood was obtained from three volunteer donors in the usual fashion by the American Red Cross 3 days prior to the study date. The units were spun according to the routine protocol used in our blood bank using a Sorvall RC 313 centrifuge (Dupont Sorvall; Wilmington DE) at 4300 rpm for 10 minutes to produce packed red blood cells. The hematocrits for each unit were 88%, 94%, and 92%. The packed cells were drawn through a PDF-20 Pediatric Transfusion Filter Set (Fenwal Products, Deerfield, IL) into an IVAC model 700 syringe (IVAC Co., San Diego, CA) for infusion and into 10 ml syringes for baseline determinations. After the first run was set up, the remaining syringes filled with packed cells were refrigerated until one-half hour before use. The catheters compared included a 28 gauge L-Cath percutaneous catheter (0.4 mm diameter, 20 cm length), a 24 gauge L-Cath percutaneous catheter (0.6 mm diameter, 30.5 cm length); both made by Lumed, Santa Ana, CA), and a 24 gauge Quicke
Aultman Hospital, Canton, Ohio, and Section of Newborn Services, University of Michigan Medical Center, Women's Hospital, Ann Arbor, Michigan Reprint requests: Dr. Bozynski, Section of Newborn Services, University of Michigan Medical Center, L3023 Women's Hospital, 200, E. Hospital Drive, Ann Arbor, MI 48109-0254
280
Copyright © 1991 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
ABSTRACT
IMPACT OF CATHETERS O N PACKED RBC/Oloya, Feick, Bozynski
Statistical Analysis
The Wilcoxon test was used to test whether there were significant changes in plasma free hemoglobin and potassium values when postinfusion values were compared with preinfusion values for each catheter. The Mann-Whitney and Kruskal-Wallis tests were used to test for differences between catheters, runs, and for any effect due to differences between donors on change in free hemoglobin or potassium concentrations. The donor variable served as a proxy for differences in hematocrit, red cell characteristics, etc. Simple linear regression analyses with change in free hemoglobin and potassium concentrations as the dependent variables and catheter type, donors, and initial plasma free hemoglobin and potassium concentrations as predictors were performed. Because the study design required revision, the a priori power calculations were not applicable and therefore additional analyses were performed to calculate the change in free hemoglobin and potassium detectable by this study. All statistical tests were two-sided.
RESULTS
Preliminary observations. The occlusion-sensing disc required frequent adjustment when the L-Cath 28 gauge percutaneous catheter was tested. A blank sensing disc was used to bypass the pressure limit feedback system completely. Nevertheless, despite unlimited pressure, only 2.4 ml packed red blood cells was infused through the 28 gauge catheter in 1 hour at a flow rate of 10 ml/hr. Because of the length
of time required to infuse 10 ml of blood for each sample (about 4 hours), the 28 gauge L-Cath catheter was deleted from the study. The remainder of the study was performed comparing the 24 gauge percutaneous catheter (PC) to the 24 gauge short catheter (SC) in paired runs with pump-catheter combinations determined randomly. The occlusion-sensing disc required adjustment occasionally on the 24 gauge percutaneous catheter but was never bypassed completely. The study data are presented in Tables 1 and 2. There was a significant change when pre- and postinfusion plasma free hemoglobin values were compared for both the SC and the PC (SC, p = 0.005; PC, p = 0.009). Although the plasma potassium also increased when samples were infused through either catheter, statistical significance was achieved for only the SC (P = 0.008). No difference was found for the change in free hemoglobin when the two catheters were contrasted (median values for both catheters: SC 179, range 53 to 413, versus PC 133, range - 9 4 to 533, p = 0.86). Similarly, no difference for change in potassium was found (median values for both catheters: SC 1.2, range 0 to 3.0, versus PC 2.0, range -3.9 to 5.0, p — 0.75). Inexplicably, both plasma free hemoglobin and potassium decreased on one run comparing pre- and postinfusion values for the percutaneous catheter (Tables 1 and 2). This was probably a labeling error, but this could not be confirmed. The change in both free hemoglobin and potassium did not appear to be related to catheter type, donor, or the order in which the samples were run and processed. It required more time to infuse 10 ml packed cells through the PC than through the SC (PC me-
Table 1.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
Cath (0.6 mm diameter, 1.6 cm length; Travenol Labs, Deerfield, IL). The IVAC syringes were placed in the syringe pump. An extension set with an occlusion sensing device was placed between the syringe and the test catheter. The tip of the catheter was placed in a Vacutainer tube containing SST gel and clot activator (Bectin Dickinson, Rutherford, NJ) to facilitate collection and subsequent separation. The tube was surrounded by ice. A flow rate of 10 ml/hr was chosen, since this infusion rate is commonly used in premature infants. The preset pump pressure limit was 6 psi (41.4 kPa or 300 mmHg). If this pressure was exceeded, an alarm sounded. Pre- and post-transfusions samples were processed identically and spun at 3000 to 4000 rpm for 15 minutes using a Beckman Model TJ-6 centrifuge (Beckman Instruments, Brea, CA) to obtain plasma for potassium and serum hemoglobin measurements. The plasma potassium was measured using an ion-specific electrode on a Beckmann ASTRA with a standard deviation of ± 0.1 mmol/liter. The plasma hemoglobin was measured using a modified benzidine method13 with a error of ± 3 mg/dl at a free plasma hemoglobin of 25 mg/dl and ± 5 mg/dl at 100 mg/dl.
Pre- and Postinfusion Plasma Hemoglobin Values by Catheter Type and Donor Plasma Free Hemoglobin (mg/dl)
Donor
Preinfusion
Short catheter (n = 10) 1 48 44 1 1 45 1 45 2 359 2 134 75 2 3 120 3 115 75 3 Percutaneous catheter (n = 10) 1 31 I 45 1 46 1 46 2 312 2 123 68 2 3 120 87 3 90 3
Postinfusion
Change
227 137 97 125 772 667 152 376 398 156
179 93 52 80 413 533 77 256
521 178 179 176 218 312
490 133 133 130 -94 189
121 376 367 264
53 256 280 174
283 81
281
AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 8, NUMBER 4 Pre- and Postinfusion Plasma Potassium Values by Catheter Type and Donor Plasma Potassium(mmol/l) Donor
Preinfusion
Short catheter (n = 10) 1 11.0 I 11.0 I 11.0 12.0 1 2 14.7 2 12.9 2 12.9 15.6 3 17.4 3 14.7 3 Percutaneous catheter (n = 10) 1 11.0 1 11.0 1 11.0 1 11.0 2 15.3 2 12.0 2 12.6 15.0 3 14.7 3 15.3 3
Postinfusion
seuille equation where the resistance to flow (Q) is related to the dimensions, length (L) and radius (r), of the vessel or tube and the viscosity of the fluid
Change
Q = APr4/8L-n 14.0 13.0 12.0 13.0 16.2 17.1 13.5 16.8 17.7 14.7
3.0 2.0
16.0 14.0 13.0 12.0 11.4 14.1 12.6 15.6 17.7 15.6
5.0 3.0 2.0 1.0 -3.9 2.1 0.0 0.6 3.0 0.3
1.0 1.0 1.7 4.2 0.6 1.2 0.3 0.0
dian, 68 minutes, range, 64 to 79; p = 0.007; SC median, 63 minutes, range, 62 to 65). The study had a 90% probability of detecting a change in free hemoglobin of 82 mg/dl and a change in potassium of 0.85 mmol/liter. DISCUSSION
The results of this project showed a significant increase in plasma free hemoglobin when packed red blood cells were infused through either the percutaneous or short catheters. Although the plasma potassium also increased when samples were infused through either catheter, statistical significance was achieved for only the short catheter. When the percutaneous and short catheters were contrasted, no significant differences for changes in either plasma free hemoglobin or potassium levels were noted. This would tend to indicate that catheter length was not a major factor, whereas hematocrit and gauge were. Although the statistical difference between the catheters was not significant, the observed increase in free hemoglobin concentration has potential clinical significance. The normal full-term newborn usually produces 6 to 8 mg/kg of bilirubin per day.14 This quantity of bilirubin may result from the breakdown of 170 to 230 mg/kg of hemoglobin, the amount contained in a packed red blood cell transfusion of 10 ml/kg. In the extremely preterm infant, that is, an infant weighing less than 1 kg at birth, this amount of free hemoglobin could cause a significant rise in serum bilirubin and hemoglobinuria. 282 The model for blood flow is based on the Poi-
From this equation, it is apparent that with no change in catheter gauge or blood viscosity, an increase in catheter length should cause a decrease in flow. In our study of 24 gauge catheters we noted that it required 64 to 79 minutes to complete the 10 ml transfusion through the percutaneous compared with 60 minutes for the short catheter. Since the volume of the percutaneous catheter was approximately 0.2 ml only an additional minute difference in transfusion time would have been predicted. The failure of the packed cells to flow through the 28 gauge percutaneous catheter is also predictable given the fact that flow is related to the fourth power of the radius and thus a small decrease in catheter diameter from 24 to 28 gauge resulted in a large negative effect on flow despite an increase in infusion pressure. Another factor in the Poiseuille equation is viscosity. Blood viscosity is related to hematocrit, red cell characteristics, and plasma constituents. Although factors affecting hemolysis are incompletely understood, an increase in shear stress will cause an increase in hemolysis.17 Since viscosity is defined as the ratio of shear stress over shear rate, in a catheter system such as ours, a rise in viscosity would be expected to be associated with an increase in shear stress and hemolysis. The relationship between hematocrit and viscosity is not linear, but exponential.16 For example, lowering the packed cell volume from 90 to 70% using saline results in a decrease in relative viscosity from about five times to 1.3 times that of whole blood.18 Thus, the increase in hemolysis reported in our study compared with earlier studies using 27 gauge needles may reflect the higher hematocrits of our packed cell units. Our study was limited, since it was performed in vitro and did not evaluate such variables as the hematocrit, catheter material, or other factors that may significantly affect hemolysis. Nonetheless, the study was performed simulating routine administration of packed red blood cells to preterm infants. This differs from previous studies, which tended to reflect laboratory conditions and not clinical practice, such as use of relatively high infusion rates or large bore catheters.10 This study showed that significant hemolysis occurs during infusion of packed red blood cells with both 24 gauge short catheters and long percutaneous catheters. In fact, the variability in pretransfusion plasma hemoglobin values suggest that minor differences in the technique used in filling syringes may lead to large differences in pretransfusion plasma free hemoglobin. The variability could also be due to minor differences in assay; however, the error in plasma free hemoglobin measurement at the hemoglobin values found in our study would only account
Downloaded by: Universite de Sherbrooke. Copyrighted material.
Table 2.
July 1991
IMPACT OF CATHETERS O N PACKED RBC/Oloya, Feick, Bozynski 4. 5. 6. 7. 8. 9. 10.
11. 12. 13.
REFERENCES 14. 1.
Batton DG, Meisels MJ, Appelbaum P: The use of peripheral intravenous cannula in premature infants: a controlled study. Pediatrics 70:487-490, 1982 2. Johnson RV, Donn SM: Life span of intravenous cannulas in a neonatal intensive care unit. Am J Dis Child 142:968— 971, 1988 3. Durand M, Ramanathan R, Martinelli B, et al: Prospective evaluation of percutaneous central venous Silastic catheters in newborn infants with birth weights of 510 to 3920 grams. Pediatrics 78:245-250, 1986
15. 16. 17. 18.
Dolcourt JL, Bose CL: Percutaneous insertion of Silastic central venous catheters in newborn infants. Pediatrics 70:484-486, 1982 Dietrich KA, Lobas JG: Use of single Silastic IV catheter for cystic fibrosis. Pediatr Pulmonol 4:181-184, 1988 Wilcox GJ, Barnes A, Modanlou H: Foes transfusion using a syringe infusion pump and small-gauge needle cause hemolysis? Transfusion 21:750-751, 1981 Angel JL, O'Brien WF, Knuppel RA, et al: Infusion of packed erythrocytes: An in vitro study of hemolysis. Obstet Gynecol 69:948-950, 1987 Gibson JS, Leff RD, Roberts RJ: Effects of intravenous delivery systems on infused red blood cells. Am J Hosp Pharm 41:468-472, 1984 Nand S, Bansal VK, Kozeny G, et al: Red cell fragmentation syndrome with the use of subclavian hemodialysis catheters. Arch Intern Med 145:1421-1423, 1985 Mateer JR, Perry BW, Thompson BM, et al: Effect of rapid infusion with high pressure and large-bore IV tubing on red blood cell lysis and warming. Ann Emerg Med 14: 966-969, 1985 Herrera AJ, Corless J: Blood transfusion: Effect of speed of infusion and of needle gauge on hemolysis. J Pediatr 99: 757-758, 1981 Schmidt WF III, Kim HC, Tomassini N, et al: Red blood cell destruction caused by micropore blood filters. JAMA 248:1629-1632, 1982 Crosby W, Furth F: A modification of the benzidine method for measurement of hemoglobin in plasma and urine. Blood 11:380-383, 1956 O'Dell GB: Neonatal hyperbilirubinemia. New York: Grune & Stratton, 1980, p 37 Bacher RP, Williams MC: Hemolysis in capillary flow. J Lab Clin Med 70:485-494, 1970 Black VD: Neonatal hyperviscosity syndromes. Curr Prob Pediatr 17:79-130, 1987 Bacher RP, Williams MG: Hemolysis in capillaryflow.J Lab Clin Med 76:485-496, 1970 Mollison PL, Engelfreit CP, Contreras M: Blood Transfusion in Clinical Medicine. 8th ed. Oxford: Blackwell Scientific Publications, 1987, p 131
Downloaded by: Universite de Sherbrooke. Copyrighted material.
for a difference of 5 to 15 mg/dl. Our observed differences in pretransfusion values ranged from 4 to 200 mg/dl. The differences between the catheters were not statistically significant. Thus, either appears suitable for infusion of packed red blood cells under conditions described in our study. Even though there were no significant differences between the catheters the quantity of plasma free hemoglobin infused could potentially cause a rise in serum bilirubin and hemoglobinuria in the very preterm infant. Small catheters (28 gauge) may not be adequate for transfusion, even at high infusion pressures, or, alternatively, infusion through smaller caliber catheters may require lower hematocrits. This option should be explored, since transfusion in the micropremie whose only vascular access is a 28 gauge catheter necessitates establishing a new line. Establishing vascular access is stressful for the infant and the gain by transfusing higher hematocrits may be minimal.
283
