A survey carried out in july last year by Personnel Today, in association with the charities DrugScope and Alcohol Concern, revealed that 31 per cent of the 306 respondent organisations were considering introducing alcohol and drug testing in the coming year. The question of drugs testing in the workplace raises concerns, on the one hand, about civil liberties, scientific accuracy and the effectiveness of existing management practices. On the other, it can be seen as a preventative measure in terms of public safety, company liability and shortfalls in productivity. It also leads us to consider underlying social attitudes towards illegal drugs and, indeed, the legislation surrounding them.

Derek Mason, Workplace Development Officer for Alcohol Concern and DrugScope, argues that "testing is not a substitute for having a proper alcohol and drugs policy. Employers are in a very good position to provide information to employees about alcohol and drug use and to train managers to handle these issues". This is validated by the fact that while 79 per cent of the respondents in the survey said they had a drug and/or alcohol policy, only 25 per cent regularly communicated these policies to staff. Mason does not entirely oppose testing but underlines the need for balance between employers' responsibilities for safety and productivity and the civil liberties and rights of the employee. Tom Mellish of the TUC also supports the need for implementing workplace policies and for distinguishing what is primarily a health and safety concern from the disciplinary procedure.

From the civil liberties perspective, arguments against testing - particularly on a random or, in the correct legal terminology, "unannounced" basis - point to its invasiveness, not only in terms of the process of obtaining supervised urine samples, but also in its implied judgement of people's lifestyles. Steve Rolles of Transform, which campaigns for the legalisation of all drugs, regards drug testing as "persecuting people for their lifestyles rather than their competency" and states that the aim should be to ascertain whether drug use is affecting competence. This opens up the grey area surrounding social and political attitudes to drugs and the uncertainties regarding their use. The increasing recreational use of drugs leads us to question how far employers have the right to know what employees do in their leisure time. Furthermore those arguing against testing often point out that responses to alcohol use differ to those regarding drugs, which is linked to their respective legality and illegality. For someone to come into work complaining of a hangover raises less eyebrows than saying they took an ecstasy tablet or a line of cocaine at the weekend.

Without well-considered policies and a greater understanding of drug use, testing can be seen by employers to provide a simple "solution" to a "problem" clouded by doubt. As Derek Mason points out, "the attraction of testing is that it produces results" and the extent of this attraction must be considerable when we consider that 67 per cent of the respondents admitted that they were not confident that they could recognise the signs of drug misuse. It is, of course, far easier to detect whether somebody is drunk - you can usually smell it --- and a breathalyser clearly indicates whether somebody is over the limit. But a positive result in a drugs test is not as clear cut and here the technology as well as the principle is open to question. The most common form of testing for drugs is by urine sample but traces of cannabis, for example, can remain in the system for up to seven days and up to 30 for regular users. Thus, if somebody tests positive after an accident, there is still no clear evidence of their being "under the influence" of drugs at the time of the event. How, then, do we define "misuse"?

Also urine tests can produce "false positives" or "false negatives" and certain medicines, including some cold remedies or cough syrups, may give a positive result. Again the argument is that if somebody is taking medication, such as anti-depressants, they should not be obliged to disclose this to their employer. Alternatives are blood and hair tests - but the former is far more expensive and, in the opinion of some, more invasive. Hair tests, while increasingly popular, cannot eliminate the possibility of traces being absorbed from a person's surroundings. As Mellish puts it, 'it is not cheap to get it right' and he questions the extent to which British companies can afford the kind of precision that is needed. There is a far more persuasive call for investing in proper counselling and advice services and encouraging employees to use these without fear of dismissal. This allows for a more informed consideration of the factors, such as stress and personal circumstances, which could be affecting someone's performance at work or causing them to use alcohol or drugs.

Given these arguments, the prospect of unannounced testing across different employment sectors is just cause for alarm. But, for a more nuanced representation of employers' views it is useful to turn to an in-depth survey carried out by the Chartered Institute of Personnel Development. Published in December last year, this survey breaks down its results into private and public sectors and, more importantly, safety and non-safety critical fields. Fifty-two of the 281 respondents reported the use of one or more types of testing, out of which only nine were non-safety critical. Furthermore just six per cent conducted random tests. Meanwhile 16 per cent of the organisations were considering the introduction of drugs testing, with twice as many in the private than public sector - the incidence of safety critical jobs being higher in the former. It is noteworthy however, that the most frequently given reason for this was employee absenteeism (55 per cent). Only three per cent cited rising accident levels.

Those arguing against testing do, in general, concede that in safety critical jobs, such as train driving and airline piloting, testing is less objectionable - given the responsibility for public safety and company liability. Here, it is worth citing the example of London Underground, which introduced mandatory testing in 1993, being one of the first British organisations to do so. According to a London Underground spokesperson, the policy was introduced in full consultation with the Trade Unions and a new employment contract had to be signed by each employee who was subject to unannounced testing (all safety critical staff). Pre-employment testing for safety critical staff and post-incident ("for cause") testing for all staff involved also takes place. All testing (breath test for alcohol and urine sample for drugs), except pre-employment, is contracted out and conducted at the workplace and samples are sent to an approved laboratory. Drug tests are subject to medical review before confirming a positive result, which leads to a disciplinary hearing. London Underground describes its policy on alcohol and drugs as using a "stick and carrot approach”. If an employee admits to a problem "they will get help, but it is too late to own up when the testing team have arrived".

Questioned as to employee reactions to the policy, London Underground admit that "staff do sometimes complain about the effect that it can have on their social behaviour" as they are required "to be aware of their behaviour even when they are off duty" and "never to come to work with drugs or alcohol in their system". However employees are said to understand how important this is and both staff and management appreciate the underpinning of this policy by a Drug and Alcohol Advisory service and the close monitoring which takes place during staff rehabilitation at work.

Although it is "expensive both in terms of money and management time", London Underground is confident of the policy's successfulness. This they assess "by reviewing the views of management and staff, the results of testing, the outcomes of those referred to the advisory service and the fact that we are constantly being asked by other companies, agencies and the media to talk about it".

In such a case, the grounds and procedures for testing appear to be well founded and the policy provides some reassurance of public safety. However, for legal reasons, London Underground would not comment on the incidence of positive results after accidents, which would provide more concrete evidence on how far drugs being "in the system" threatens safety or impairs performance.

Whether or not testing should be mandatory in certain professions needs to be considered in proportion to the job concerned and based on clear evidence of competence impairment. This may seem straightforward, but the demarcations of such criteria are very much open to debate. There is a clear and pressing need for further research and careful considered policy as well as a greater, rational understanding of the role drugs play in society and the effects they have. While the science of testing is still subject to scrutiny, a highly discriminatory approach to its implementation is required. In other words, it needs to be firmly grounded in a broader policy that, in the first instance, provides support. If testing is to be mandatory, its guidelines should be as well. These in turn should also seek to define the professional areas where testing is justifiable and not just a "quick fix" for a vaguely defined "problem".