From Mr T. Delaney, MPSI
SIR,—I cannot agree with Mr Freedman (PJ, September 4, p349) about the use of colour on labels. The problem with his attitude is that it does not take account of the environment in which those who administer drugs in hospitals must operate - an environment of frequent interruptions, stressful working conditions, and overly long working hours (especially for doctors). It is not uncommon for medical staff to be working while on the brink of exhaustion. In these circumstances, it is not good enough to say "force them to read the labels". Instead, we must give those on the front line all the help we can.
In my discussions with many front-line junior staff in nursing and medicine, who, unlike pharmacists, actually prepare and administer the vast majority of doses administered in our hospitals, I have never met one who held the views of Mr Freedman, ie, that they should be "forced to read the label" by the elimination of colour. Virtually without exception, they agree that the use of colour is a help.
Indeed, recently a consultant anaesthetist at our hospital contacted the pharmacy department to ask that we ensure that any potassium chloride concentrate injections stocked should be a brand with a coloured label. This was because he had encountered non-colour labelled potassium chloride injections in a private hospital where he worked, and considered them to be very dangerous because of the similarity of appearance with water for injections and normal saline ampoules. In our hospital, we have special storage requirements for potassium chloride concentrate and a policy which encourages use of this only as a last resort. But we also use the colour-labelled ampoules as an extra safety feature.
While I do not agree with the idea of colour-coding, meaning the use of a particular colour to denote a particular drug, I do consider that the use of colour on labels can be a valuable additional safeguard to complement the use of appropriate fonts and letter size in the wording on packages.
A recent medication error at our hospital serves to further illustrate the point. Luckily, our patient, who had to be resuscitated, recovered fully from this misadventure.
An antibiotic was inadvertently diluted with a solution of lignocaine 2 per cent rather than the intended water for injections. Both products came in identically sized and shaped vials from the same manufacturer. Both had the kind of black and white labels proposed by Mr Freedman. Root cause analysis of the error revealed seven causative factors, one of which was the similarity in appearance of the products. (Two other factors were the tiredness of the doctor who had worked over seven hours without a break, and an interruption while the doctor was preparing the dose.)
When we advised the manufacturers of the lignocaine 2 per cent (and the water for injections) of the error, they advised us that, in the United Kingdom, they used colour on the lignocaine 2 per cent label as an extra differentiating factor. They also provided correspondence from 1996 with the Irish regulatory authorities in which they had sought permission to use the UK style label in our market, precisely because of fears of the error which occurred at our hospital. The response of the Irish regulatory authorities was as Mr Freedman would have it - users should be forced to read the label. The manufacturers were denied permission to change the label.
Research shows that 85 per cent of errors are due to process causes and 15 per cent to human error. Managers of the medication-use process are therefore responsible for most errors. Rather than face this uncomfortable fact, too often the easy option is to ascribe all errors to the human failings of the junior staff who are usually on the front line, and, therefore, the ultimate instruments of the errors.
As head of pharmacy, I bear responsibility in part for the lignocaine error described above, because I did not see to it that we used different brands of water for injections and lignocaine 2 per cent injection which were different in appearance. The Irish Medicines Board, which refused to allow the manufacturer to make these products more distinguishable, likewise bears some responsibility.
Our approach to errors is to seek to learn from them and to try to prevent a future recurrence by eliminating or addressing all the contributory factors. In the real world, some factors can be more easily addressed than others. For example, we would be unrealistic to suppose that the disgraceful treatment of doctors, so far as working hours are concerned, will end inside the next 10 years, even though it has been outlawed for all other categories of worker by European Union working time legislation. On the other hand, to allow the use of colour, not for coding, but as an additional distinguishing feature, would be a realistic process improvement. Of course it will not be of help to the 10 per cent of the population which is colour blind, but that is no reason not to use it to help prevent errors by the majority of us, which is not colour blind. In his recent book ‘Medication errors' (Washington DC: American Pharmaceutical Association, 1999), Dr Michael Cohen comments: "The argument is specious; that the rate of adverse events might be reduced as much as 90 per cent through the use of colour seems not to have been understood."
Pharmacy staff are mostly in the fortunate position that even when a dispensing error gets through our checking systems, there is still one more possibility that the mistake will be found before the drug is given to the patient. Those who administer medicines have no such safety net.