Mount syringes vertically …

EDITOR,—D A Southern and M S Read report on a patient who received an overdosage of opiate while using patient controlled analgesia.1 Such an overdosage is not, however, a specific consequence of this technique but can occur with any infusion syringe, so it is not only anaesthetists who need to be aware of this risk. Infusion of insulin, for example, could be equally dangerous and might similarly be given in relatively low dependency areas.

Such events must be rare, but the fact that they may happen with a damaged syringe has been unequivocally shown.2 In this latest case the syringe was “normal on close examination.” When I duplicated the administration set used I found that perforating the 50 ml syringe plunger with a 20 gauge cannula resulted in virtually free flow, given a gravitational advantage of as little as 5 cm. Such a small leak might not be apparent on visual inspection, but ideally the syringe that was used should also be tested under pressure and the remainder of that batch looked at by the manufacturers. Failure to examine formally the relevant equipment means that the one reported lethal case of overdosage with patient controlled analgesia remains unexplained.3

Infusion syringes are frequently mounted horizontally and at some height above the patient. In such an orientation an airlock should develop before the syringe is competely emptied. The outcome would probably have been far worse if the syringe had been mounted pointing downwards. This situation is often exacerbated by the omission of antisyphon valves.

The solution is always to mount syringes vertically with the outlet uppermost at or even below the level of the heart, but despite such precautions an apparently fail safe system can still fail. Perhaps we should consider reverting to one of the early precepts of patient controlled analgesia—namely, that the syringe should contain only a “survivable dose.”2 In Nottingham, for example, when using morphine we use 60 mg in 60 ml, though more frequent changes of syringe then carry their own potential for operator error.

Patient controlled analgesia (and, by implication, other infusion techniques) has a good safety record,4 but current practice is far from ideal. We should recognise that syphoning is a practical as well as a theoretical risk with any infusion syringe. Despite these problems, if the measures suggested above are used we can still hope to see this technique proved to be more effective than, or at least as safe as, the alternatives.