Factors affecting flow through blood administration sets

Transcription

1 European Journal of Anaesthesiology 1997, 14, Factors affecting flow through blood administration sets M. D. Stoneham Department of Anaesthesia, Stoke Mandeville Hospital, Aylesbury, Bucks HP18 9AN, UK M. D. Stoneham (1997) European Journal of Anaesthesiology, 14, Factors affecting flow through blood Summary Factors affecting flow through blood administration sets in vitro were assessed under gravity-fed and pressurized conditions including an assessment of the influence of the intravenous (i.v.) cannula and Luer lock fitting. The fastest gravity-fed flow of ml s 1 was obtained through the largest internal diameter (ID=4.8 mm) blood administration set. Flow through blood with ID=3 mm was 50% of this. Flow increased over base-line through all the when the i.v. cannula was removed (range 18 50%) and increased further over base-line when the Luer lock fitting was removed from the distal end (range %), indicating that these are ratelimiting steps in the system. The Y-type trauma set with the largest diameter tubing facilitated the fastest flow, although flow through all the Y-type trauma sets produced lower flow rates than the corresponding blood, which may reflect their relative increased length. The ideal blood administration set should have an internal diameter at least 4mm and be less than 170 cm in length. Keywords:, therapy, methods;, intravenous. Introduction Accepted October 1996 Correspondence: Dr M. D. Stoneham, Nuffield Department of Anaesthesia, John Radcliffe Hospital, Oxford OX3 9DU, UK. Despite technological advances in automated methods for administering warmed intravenous (i.v.) fluids, there remain unresolved problems with stand- ard manual techniques for increasing flow. Laminar flow through cylindrical tubes depends on several factors as described by the Hagen-Poiseuille formula [1]. Maximum flow is restricted by the dia- meter of the narrowest part of the whole fluid ad- ministration apparatus, including: fluid bag, blood administration set tubing and the i.v. cannula, and by the length of the tubing. The major rate limiting step is believed to be the i.v. cannula [2]. Manual compression of the lower drip chamber of the fluid administration set using the ball valve is an inefficient method of increasing flow compared with external compression of the i.v. fluid bag and only marginally better than gravity itself because the pump chamber refills retrogradely after release [3]. Thus the requirement for blood with such a ball valve mechanism can be questioned. In addition, different diameter sets are available, which may affect the maximum flow through the set. There are other features of blood to consider also and these include: ease of priming; risk of air embolism when the system is pressurized; ease of injection into the system and cost. This study was designed to compare flow rates through currently available blood administration and trauma sets using an in vitro flow measurement tech- nique adapted from a clinical urological assessment system. Methods Six blood, manufactured by four companies listed in a current guide of equipment manufacturers [4], were evaluated (Table 1). A 500 ml bag of Gelofusine plasma substitute (B Braun Medical 1997 European Academy of Anaesthesiology 333

2 334 M. D. Stoneham Table 1. Details of blood Blood administration ID Total Drip chamber Priming Drip chamber Ball valve/ Filter size set (mm) length to Luer lock volume design manual pump (μm) (cm) distance (cm) (ml) Standard set Braun High Flow Flexible No Baxter C Flexible Yes 170 Portex A Flexible Yes 200 Codan Rigid No 200 Braun / Rigid No 170 Portex A Rigid No 200 Y-type trauma tubing Baxter C2159Y Flexible Yes 170 Braun Flexible No 170 Portex A Flexible Yes 200 flow. Thus Voiding time is equal to the time taken to empty the infusion bag and Voided volume is the total volume of fluid emptied from the infusion bag. Direct calibration of the uroflowmeter was per- formed before each experiment as described in the Urodyn 1000 Service Manual [5]. Calibration of the uroflowmeter involves pouring 400 ml of the fluid under test into the flow transducer at an even flow rate over s. If the volume measured differs from the actual volume (measured using a graduated meas- uring cylinder) by more than 1%, calibration is repeated. Since it is the volume of fluid which is the determining factor during calibration, calibration may be performed for fluids with different densities (relative density of Gelofusine=1017 g L 1 [personal communication, B. Braun Medical Ltd]) and viscosities. Each technique for increasing flow was repeated four times. Experiments were performed at room tem- perature (21 C) on one test day. Ltd, Aylesbury, England) was connected to the blood administration set under test, the distal end of which was then connected to a 14-gauge Braun Vasofix cannula mounted horizontally in a burette stand. The bag of fluid was suspended from a drip stand such that the vertical distance between the cannula and the junction of fluid bag and administration set was 100 cm (Fig. 1). Blood were changed between each set of tests and pre-filled before each test such that a full bag of fluid was emptied each time. The colloid Gelofusine was used as a blood substitute because of the difficulty and expense of obtaining sufficient quantities of bank blood. Fluid emerging from the distal end of the cannula was directed into the collecting funnel of a Urodyn 1000 uroflowmeter (Dantec Electronics Ltd, Bristol, England). This is an automated device used to measure the urine flow rate of patients with, for example, bladder outflow obstruction. It consists of a collecting funnel at the base of which is a metal disc kept rotating at constant speed by a Servo motor (Fig. 1). The amount of electrical energy required to keep the disc speed constant, when fluid lands on the disc, is proportional to the amount of fluid landing on the disc per second, in other words, to the flow rate of the Experiment 1. Comparison of different blood The flow rate of Gelofusine through a 14-gauge cannula was measured for each of six different giving fluid. A graphical representation is generated by the sets (see Table 1) under two different conditions: (1) machine, as well as the duration, maximum flow, gravity fed, 100 cm pressure head and (2) compression mean flow and total volume of fluid which has been within a rigid pressure box (PneuPac 1002, PneuPac infused (Fig. 2). The chart recorder labels are relevant Ltd, Luton, England) with constant 300 mmhg pres- to urology because this is a device for measuring urine sure provided by an automatic adjustable tourniquet

3 Flow through blood 335 Fig. 1. Equipment used for measuring flow. Fig. 2. Graphical display of flow from uroflowmeter. device (VBM Automatic Tourniquet 2X500/E, (2) Portex Avon brand A100; Medizintechnik GmbH, West Germany), monitored (3) Baxter C2071; with an aneroid pressure gauge. The six giving sets (4) Codan ; used were: (5) Braun Sangofix ES /7; (1) Braun High Flow ; (6) Portex Avon brand A122.

4 336 M. D. Stoneham ES set=2.5 ml s 1, Portex A122 set=2.55 ml s 1 ) (, P<0.0001). Constant external pres- surization to 300 mmhg caused an approximate threefold increase in flow through all the blood ad- ministration sets tested. Experiment 2. Influence of i.v. cannula and Luer lock fitting on flow rate To investigate the influence on flow of the i.v. cannula and Luer lock fitting, flow was measured through six blood using a gravity-fed infusion under 100 cm static pressure head. Three different conditions at the distal end were employed: (1) 14- gauge cannula; (2) cannula removed; (3) Luer lock attachment removed. Experiment 3. Comparison of Trauma (Y-tubing) blood The trauma tubing set with ID 4.8 mm (Braun) gave the fastest flow compared with the two with ID 3 mm (Baxter and Portex) during both pressurized and un- pressurized conditions (Fig. 4) (, P<0.0001). To assess the effectiveness of the available Y-type trauma blood, flow was measured for each of three currently available trauma Y-tubing sets (Braun , Baxter C2159 and Portex A13 see Table 1) under 2 conditions as for experiment 1. (a) Gravity-fed, 100cm pressure head, (b) automatic external pressurization to 300 mmhg. Statistical analysis was performed using the Arcus Pro-Stat statistical computer software program DOS version 3 (Medical Computing Ltd, Aughton, West Lancs, England). Statistical analysis was by analysis of variance ( ) comparing mean flow rates in each experiment. Individual comparisons between mean flow rates from different experiments were performed by modified Student s t-test with the Bonferroni correction. Results Experiment 1. Comparison of different blood Figure 3 shows the mean flow from six blood administration sets tested. Increasing the diameter of the blood administration set increased the mean flow rates obtainable through a 14-gauge i.v. cannula. Mean Experiment 2. Influence of cannula size and Luer lock fitting on flow rate Table 2 shows the changes in flow recorded when the cannula and Luer lock fitting were removed. Flow through all the increased when the cannula was removed (, P<0.0001) and increased further when the Luer lock fitting was removed from the administration set (, P<0.0001). With the Luer lock removed, flow through the 4 mm ID blood sets was greater than through the 4.8 mm ID Braun High Flow set (, P<0.0001). Experiment 3. Comparison of Y-trauma tubing blood Discussion There are several factors which contribute to a reduction in fluid flow from the i.v. fluid container, through a blood administration set and out through an i.v. cannula. The Hagen-Poiseuille formula describes the laminar flow of a Newtonian fluid through a smooth walled tube V = π.δp.r4 [V =flow, ΔP=pres- 8ηl sure change across tube, r=internal radius, η=viscosity, l=length] (1). Flow is proportional to the fourth power of the radius and the pressure change across the tube and inversely proportional to the viscosity of gravity-fed flow rates through the blood ad- the fluid (which is related to its temperature) and the ministration set with 4.8 mm internal diameter tubing length of the tube. If flow becomes turbulent, for (Braun High Flow set=4.78 ml s 1 ) was sig- example where the ID of the system changes abruptly, nificantly greater than that through sets with 4 mm then density of the fluid also becomes important in tubing (Baxter C2071 set=3.98 ml s 1, Portex A100 limiting flow. set=3.73 ml s 1 ) (ANOVA, P<0.0001) which in turn Flow through blood must be was significantly greater than that through sets with assumed to have both laminar and turbulent components 3 mm tubing (Codan set=2.55 ml s 1, Braun since there are several points in the system

5 Flow through blood 337 Fig. 3. Mean flow rates from 14-gauge cannulae through six different blood Table 2. Effects of i.v. cannula and Luer lock fitting on mean (SD) gravity-fed flow from blood 14-gauge Cannula % Luer fitting % cannula removed improvement removed improvement over 14 gauge over 14 gauge Braun High Flow 4.78 (0.15) 6.33 (0.15) (0.18) Baxter C (0.05) 5.78 (0.1) (0.1) 129 Portex A (0.05) 5.58 (0.17) (0.1) 81 Braun ES 2.5 (0.08) 3 (0.08) (0.05) 35 Codan 2.55 (0.06) 3 (0.08) (0.05) 26 Portex A (0.06) 3.1 (0.08) (0.05) 32 where the diameter changes abruptly. Both viscosity and density were constant factors in these experiments, thus the principle factors affecting flow should be the ID and length of the tubing. This assumption is supported by these data. From the results of experiment 1, flow through constructed from 3 mm ID tubing was 60 70% of that through 4mm sets, and only The discrepancy between experimental and theoretical calculation is explained not only by the fact that Gelofusine is not a perfect Newtonian fluid, but also by the presence of other rate limiting points in the apparatus such as the filter, the drip-forming aperture in the drip chamber and the Luer lock fitting as well as minor differences in length. In fact the results of experiment 2 suggest that the Luer lock fitting is the 50% of that through the 4.8 mm set. One can predict greatest rate-limiting step of the Baxter 4 mm ID blood (Hager-Poiseville formula) that the relative flow administration set, but not of the Braun 4.8 mm set through 3 mm ID tubing should be 31% of that through since removal of the Luer fitting had such a profound 4mm tubing and 15% of that through 4.8 mm tubing. effect on flow through the Baxter set.

6 338 M. D. Stoneham Fig. 4. Comparison of flow rates through Y-type trauma tubing blood. With this unequivocal evidence that big is best when it comes to blood administration set tubing, why then are sets constructed from 3 mm tubing? British Standard BS 2463 pt 2 (1989) states that all fluid for use with blood should have an ID not less than 3.95 mm [6], although the International Organization for Standardization ISO (1987) specifies that the ID must be not less than 2.7 mm [7]. In addition to this, European consumers have apparently expressed a preference for 3 mm tubing (personal communications, B Braun Medical Ltd and Portex Ltd). There is indirect evidence from the data on flow through Y-type trauma tubing that length is also an important consideration in determining flow. Flow through all the trauma sets tested was slower than though the fastest of the standard blood administration sets. This is not perhaps surprising in two of the sets (Baxter and Portex) which are constructed from 3 mm ID tubing. However, the Braun trauma set is made with 4.8 mm ID tubing, yet mean flow through the Baxter standard C2071 blood administration set (4 mm ID) was 5 62% greater than that through the Braun trauma set over the different testing conditions. These trauma sets are all 225 cm in length or longer Any improvement in flow through a blood administration set facilitated by a reduction in length will be offset if an extension set is also used. Personal experience and observation confirm that such extension sets, which are typically 100 cm in length, made from 3 or 4 mm tubing and when added, include a three-way tap if the site of i.v. access is covered by surgical drapes during an operation and an i.v. in- jection port is required. One further difference between the sets is in the design of the drip chamber. Blood with a drip chamber constructed using rigid plastic (Portex A122, Braun and Codan sets) have less risk of air embolism when the fluid bag is pressurized and are subjectively easier to fill than those with flexible drip chambers. The value of the pump chamber in a blood administration set has been shown previously to be limited because the drip cham- ber fills retrogradely immediately after release rather than anterogradely from the fluid bag [3]. In Australia and the United States, blood are available with a different type of manual pump chamber which has a non-return valve incorporated to prevent retrograde filling of the drip chamber when at least 32% greater than the Baxter set. Of course the pump is released. However, these sets are much there are other possible reasons why flow should be longer for example 250 cm and therefore this limits lower in these dual chambered giving sets, such as a the maximum flow obtainable. narrow aperture in the drip chamber or turbulent flow Limitations of this survey, in particular the fact that where the two limbs of the set meet, however, length no assessment was made of the contribution of venous is one obvious reason for the observed reduced flow. resistance [8,9], and that a synthetic colloid rather than

7 Flow through blood 339 blood was used means that care must be taken when Measurement in Anaesthesia, 4th Edn. Oxford: But- extrapolating the results of this work to clinical use. terworth-heinemann, In conclusion from the results of this in vitro survey, 2 Kestin IG. Flow through intravenous cannulae. Anaesthesia 1987; 42: it is recommended that the ideal blood administration 3 Stoneham MD. An evaluation of methods of increasing set is constructed from tubing with an ID of at least the rate of intravenous fluid administration. Br J Anaesth 4 mm throughout. The length should be reduced to 1995; 75: the minimum compatible with practicality, which is 4 Arthurs G. User Guide to Anaesthetic and Intensive Care probably 170 cm, yet perhaps incorporating an in- Equipment. Herts, UK: Saldatore Ltd, jection port to avoid the requirement for an extension 5 Dantec Electronics. Urodyn 1000 Service Manual. Bristol: set. The fluid chamber should be of rigid construction Dantec Electronics Ltd, and a manual pump chamber is unnecessary. Further 6 British Standards Office. British Standard Transfusion work is required to determine the value of Y-type Equipment for Medical Use. Part 2. Specification for Ad- trauma tubing. ministration Sets. London UK: British Standards Institution; International Organization for Standardization. Transfusion Acknowledgment equipment for medical use. Part 4: Transfusion sets for The author would like to thank B Braun Medical for single use. International Organization for Standardization. 1987; Berlin, Germany: International Orproviding equipment used during this study and the ganization for Standardization, Department of Radiology, Stoke Mandeville Hospital 8 Goodie DB, Philip JH. An analysis of the effect of venous for loan of the Urodyn 1000 uroflowmeter. resistance on the performance of gravity-fed intravenous infusion systems. J Clin Monit 1994; 10: Elad D. Intravenous infusion: understanding the technical References side can improve clinical performance. J Clin Monit 1994; 1 Davis PD, Parbrook GD, Kenny GNC. Basic Physics and 10: