Antireflux valves in patient-controlled analgesia

Transcription

1 Anaesthesia, 199, Volume 45, pages APPARATUS Antireflux valves in patientcontrolled analgesia M. T. KLUGER AND H. OWEN Summary Antireflux valves are widely used in conjunction with patientcontrolled analgesia devices. It is important to appreciate the limitations and dangers of these systems. They can achieve a potential stored volume if occluded and they may, as part of the administration set, retard.puid administration. Seven antireflux systems currently available were tested in conjunction with three patientcontrolled analgesia pumps. The systems volume, time to occlusion alarm and flow rates were measured. Results showed that the sets with low stored volumes were less eficient as administration sets. A potentially dangerous bolus could result after release of occlusion iy sets with large stored volumes were used in conjunction with pumps that utilised concentrated solutions of opioid. This study has identijted the ideal antireflux valve system. Key words Equipment; infusion systems, patientcontrolled. The need for antireflux valves during patientcontrolled analgesia (PCA) was recently highlighted by the Emergency Care Research Institute (ECRI). The valves are necessary to prevent retrograde pumping of analgesic agent into the tubing of a parallel, gravitydriven infusion should the intravenous cannula become obstructed. The feedback loop is opened if this happens, the technique fails and a dangerous dose of drug could be delivered as a bolus when the blockage is cleared to restart the infusion. The ECRI recommendation for PCA systems is that if complete occlusion occurs the pump should alarm within three dose activations at every dose setting. The pressure in the system required to activate the occlusion alarm is determined by the manufacturer; it is generally high (over 3 mmhg) to overcome the internal friction of the syringe and to accommodate the pressure required to overcome the flow restriction of smallbore cannulae and delivery tubing. The effect of adding an antireflux valve to the drug delivery system has not been studied. The volume of drug solution expelled from the PCA device and delivery tubing after releasing an occlusion (which we have called the stored volume) is determined by the setting of the occlusion pressure alarm and the compliance of the system (tubing plus valve). One study reported that surges of up to.5 ml can be delivered from an infusion pump connected to standard tubing after release of an obstruction.2 The number of milligrams of opioid contained in this stored volume is then a product of the concentration of the solution. Some manufacturers of PCA devices also produce dedicated givingsets for their equipment, whereas others either suggest or recommend a particular product. One PCA device is in widespread use at this hospital and two others were considered and used with several antireflux systems. It was clear from discussion with colleagues using PCA, and some equipment suppliers, that the clinical relevance of stored volume had largely been ignored and so was worth further investigation. It was decided formally to compare some PCA and antireflux systems, paying particular attention to the stored volume. rate through the antireflux valves and the potential for misconnexion were also studied. Methods The characteristics of seven different antireflux systems (Fig. l), when used with three PCA pumps, were investigated. The PCAS (Graseby, Watford, UK) and MDS I1 (Bionica, Castle Hill, NSW) were tested with the following antireflux devices (see Appendix for list of manufacturers): Tuta in two configurations; Lifemed, Cardiff (v.y.c. Con), Abbott PCA, Abbott PCA Mini Bore with antisiphon (Abbott 2) and Quest Medical. The Lifecare 41 (Abbott, North Chicago) was tested only in conjunction with its own dedicated giving set for comparison. M.T. Kluger, MB, ChB, FCAnaes, Research Fellow, H. Owen, MB, ChB, FCAnaes, FFARACS, Senior Staff Specialist, Department of Anaesthesia and Intensive Care, Flinders Medical Centre, Bedford Park, Adelaide, South Australia 542, Australia. Accepted 15 February /9/ The Association of Anaesthetists of Gt Britain and Ireland 157

2 QJ M.T. Kluger and H. Owen Fig. 1. Antireflux valves used; left to right; Abbott 1, Abbott 2, Tuta 1, Tuta 2, Cardiff, Lifemed (Above) and Quest (Below). Stored volume The pumps were set to deliver the maximum bolus dose at their highest rate of delivery. Each pump was tested in turn four times with each of the seven antireflux valves, using either the integral part of the antireflux valve (Tuta 1 and 2) or a standard Tuta low compliance tubing (length 75 cm, deadspace 1.25 ml, part no. 538) in a configuration used clinically. Each administration set had a threeway valve at its termination which was used to create an occlusion. The pumps were primed with.9% saline and activated to deliver a bolus with the threeway tap in the closed position. The time taken for the occlusion alarm to sound was measured using an analogue stopwatch. When the occlusion alarm sounded, it was silenced and the tap opened. The fluid expelled from the tubing ('stored volume') was collected in a preweighed beaker on a calibrated toploading electric balance (Model PL 12, Mettler Instruments, AG, Zurich, Switzerland). The volume was determined from the mass of fluid and the specific gravity of.9% saline at ambient temperature (1.6 at 2225 C). The time to postocclusion alarm and the 'stored volume' released were recorded for all combinations of the antireflux valves with the Graseby and Bionica pumps. The Abbott was tested solely with its dedicated giving set. A new antireflux device was used for each determination, to prevent bias from any hysteresis in compliance, which may occur with cyclical pressurisation of the set. Retardation of fluid administration A protocol, similar to that used to investigate the efficiency of the Cardiff valve, was carried out in order to assess the adequacy of the valves as part of the administration set.) Each valve system was connected to the Luer connector of a standard intravenous infusion giving set primed with.9% saline. The dripcontroller was turned on and the time for 2 ml of saline to flow into a measuring cylinder was measured when the surface of the saline was 5, 1 and 15 cm above the outlet of the oneway valve system. This provided a range of pressures from 515 kpa, which corresponds to that found in normal clinical practice, between the reservoir and the patient's venous system. The measurements were repeated three times at each height, and also without any valve present. Results The time to alarm with the PCAS (Fig. 2) ranged from 5 seconds (Abbott 2) to 15 seconds (Quest). The stored volume expelled after release of occlusion ranged from 1.3 ml (Abbott) to 2.6 ml (Tuta 2). The MDS 11 (Fig. 3) had " 'y ; s o 3 + E :;. t E g 1. E Tuta I Abbott I Cardiff Quest Tuta 2 Abbott 2 Lifemed Fig. 2. Time to occlusion alarm and stored volume with the Graseby PCAS. n = 4 for each group.

3 I 25 v) p 225 : 2 v) 175 E & 15 c 125 P ; 1 g i.7 t E 2.4 ' 3.3 V $.2. I. Tuta I Abbott I Cardiff Quest Tuta 2 Abbott 2 Liferned Fig. 3. Time to occlusion alarm and stored volume with the Bionica MDS 11. n = 4 for each group. wider range of occlusion times, 24 seconds (Abbott 2) to 21 seconds (Tuta 2). The volumes associated with this pump were much smaller,.8 ml (Abbott) to.6 ml Tuta 2. The Lifecare 41 with its special set had a mean time to occlusion of 59.5 seconds and stored volume of 1.24 ml. The results of the flow studies (Fig. 4) show that the most efficient administration set, as demonstrated by the most rapid flow rate, were the two Tuta sets. These proved to have a rate of delivery equal to the control drip set with no antireflux valve present. The Abbott PCA set reduced the AntireJlux valves in PCA 159 maximum flow rate to one third of the more efficient sets and although the newer Abbott 2 was more efficient at higher pressures, it was not an improvement over the PCA or other sets at lower pressures. The Cardiff, Quest and Lifemed were intermediate in their performance. Discussion The effectiveness of antireflux valves was tested to find out whether their intrinsic compliance and possible 'stored volume' led to the delivery of a bolus of drug when an occlusion was released. We found that the Tuta sets, with their widebore tubing and large, floppy, compliant valve leaflets, have consistently higher stored volumes than the other antireflux devices. This becomes clinically relevant in the context of PCA and the pump used in conjunction with the antireflux device. We routinely use morphine sulphate at a concentration of 1 mg/ml. Under these conditions in the Graseby PCAS, an accidental dose of approximately 2 mg could be administered after an occlusion was cleared. This is within an acceptable dose range, although undesirable. It should be noted that this pump can be programmed to provide a concentration range from 1 pg/ml' to 99.5 mg/ml, so more concentrated solutions could be used. However, the Bionica MDSI 1 PCA has a minimum concentration setting of 1 mg/ml and so would have the potential of administering a 6mg dose of opioid. This could have potentially serious effects and would be unacceptable in clinical practice. The bolus administered is still significant, even in conjunction with other other valves such as the Cardiff, Lifemed and Abbott; 2.5 mg, 3.1 mg and 2.5 mg respectively. The differences in time to occlusion alarm, although longer with the antireflux devices used in conjunction with the Bionica pump, were clinically insignificant. It was also questioned whether the addition of one of the valves hindered intravenous fluid administration. It seems from the results that the most effective antireflux valves, that is with respect to low stored volume and short occlu Pressure (kpa) Fig. 4. rate characteristics of the seven antireflux devices when tested at three different pressures., tuta 1;, Tuta 2; AA, Cardiff; AA, Abbott 1; VV, Abbott 2; ww, Lifemed; ou, Quest; DV, no valve.

4 16 M.T. Kluger and H. Owen P b Lifemed Quest ond Abbott I Cardiff Tuta I and 2 Abbott 2 Fig. 5. Schematic illustration (not to scale) of the antireflux valves used in the study. Shaded areas correspond to stored volume. \ sion alarm times, in particular the Abbott sets, significantly hinder the maximum flow rates associated with fluid administration. Conversely, the sets with the potential for large stored volumes, for example the Tuta sets, are able to deliver fluids as rapidly as the normal drip sets. The stored volume has two components, the compliance of the tubing and deformability of the valve. Most of the antireflux valves are small flap valves, but the prototype Tuta sets have large duckbilled valves with large valve leaflets which during occlusion testing can be seen to invert, but under normal conditions do not retard rapid flow (Fig. 5). The flapvalve in the original Cardiff valve could be rotated so that in one position it did not function as an antireflux device, whereas at another setting it completely occluded flow. The potential for misuse is immediately apparent and the version of Cardiff valve we tested does not have this facility. The infusion can still be incorrectly connected to devices such as the Cardiff valve, Abbott PCA and Abbott 2 sets, since there is only one valve present on a twolimbed infusion system? This again may defeat the purpose of using these systems. The Tuta sets do not have this potential problem, because they have only one access port. Both the Lifemed and Quest products have valves on both limbs so that identification of the protected limb is unnecessary. This is especially useful when PCA is first introduced and personnel are inexperienced in its use. The Lifemed device is sterilised by ethylene oxide and therefore the protecting plugs at the ends of each limb have a central hole. Thus, if only one limb is to be used, for example during initiation of intravenous therapy before induction of anaesthesia, the end plug needs to be replaced. The systems sterilised by gammairradiation do not have this irritating problem. An antireflux valve in place intraoperatively prevents blood being forced backwards along the infusion line by a noninvasive blood pressure monitor on the same arm. The technical specifications of modern PCA pumps are increasingly complex to ensure greater accuracy and safety. However the patient/machine delivery interface, that is the connecting set and antireflux device, was shown to have significant effects on the function of the overall PCA system. The disposables must be considered as important as the pumps themselves for the safe administration of PCA, and the manufacturers should take account of this. Moreover, when in use with pumps that rely on small volumes of concentrated solutions, for example the MDS 11 PCA, they can present a potential problem. It was previously shown that occlusion of syringe pumps can generate substantial inline pressures that range from mmhg and deliver a bolus of.5 m1;* we have shown similar results in the context of PCA machines and antireflux devices. We suggest that, for safety, dilute solutions be used in PCA systems for postoperative pain control. This might avoid recurrence of some of the accidents that have been reported with PCA.5 Nursing staff need to be informed of these dangers and, in particular, must ensure that if the occlusion alarm has been activated the system is disconnected before the line is unblocked. PCA requires a drugdelivery system, designed to prevent incorrect configuration, with low intrinsic compliance and a wide bore to allow rapid fluid administration. It is also

5 Antireflux valves in PCA 161 suggested that the ideal system should be sterilised by irradiation and be inexpensive. Currently, if the antireflux valve system is to be incorporated in the intravenous line, for example, to facilitate drug infusion during surgery, then the Tuta sets will not retard fluid administration. However, the Abbott sets would be safest if the system is added after operation specifically for PCA administration. It appears that the ideal system is still to be marketed. Ackaowledgmemts The authors are particularly grateful to E. Loveday of Abbott Australasia, T. Sofu of Tuta Laboratories, G. Middleton of Implox for providing samples of their products and Dr J. Plummer for reviewing the manuscript. Note added at proof The Bionica pump has since been modified so that drug dilutions down to 1 mdml can be programmed. References 1. Anonymous. Patientcontrolled analgesic infusion pumps. Heulth Devices 1988; 17: RUNCIMAN WB, ILSLEY AH, RUTTEN AJ, BAKER D, FRONSKO RLL. An evaluation of intravenous infusion pumps and controllers. Anaesthesia and Intensive Care 1987; ROSEN M, WILL B. The valvedyconnector (V.Y.C. Con). Anaesthesia 1979; 34: WHITE PF. Mishaps with patientcontrolled analgesia. Anesthesiology 1987; 66: moms DW, OWEN H. Patientcontrolled analgesia: the need for caution. Anaesthesia 1988; 43: 772. Appendix List of manufacturers of antireflux valves Tuta, Lane Cove, New South Wales, Australia. Lifemed, Biospectrum, Chatswood, New South Wales, Australia. Cardiff, Vygon, Uxbridge, UK. Abbott PCA Set, Abbott Laboratories, N. Chicago, Illinois, USA. Quest, Quest Medical, Carolton, Texas, USA.