CS Spray – the critical study the Policing Board ignored

HEALTH AND SAFETY IN POLICING: LESSONS FROM THE REGULATION OF CS SPRAYS IN THE UK

INTRODUCTION

The use of weapons by security forces poses major health concerns. Those health professionals and social scientists writing about the effects of weapons typically consider them in relation to specific military conflict situations (e.g., Coupland, 1996). In the civilian context, few studies have been done of the health effects of policing weapons beyond the case of firearms in police-public encounters. Over time though, the collective effects of the use of force in police-civilian encounters can result in significant health problems for both parties.

The need to examine the regulation and effects of weapons in policing situations has intensified in recent years because of the introduction of a wide range of new force technologies. To resolve conflict situations encountered by officers during their duties, most countries have attempted to employ so-called 'non-lethal' weapons with the aim of minimising the risk of injury to both officer and assailant. The range of existing non-lethal weapons deployed by police officers include kinetic energy weapons; incapacitant chemicals in spray, aerosol and gas forms; and water cannons. In addition, there is an active search to develop the next generation of non-lethal weapons for routine and public order policing as well as military applications such as peacekeeping missions.

The development of such technology is supported by international agreements. The United Nations Basic Principles on the Use of Force and Firearms by Law Enforcement Officials encourage law enforcement agencies to develop new, 'carefully evaluated and controlled' non-lethal weapons in order to decrease the risk of death or injury from more conventional force options such as firearms or batons. What 'carefully evaluated and controlled' mean in practice though is not specified. There are then pressing questions about the regulatory framework in place for assessing and monitoring the effects of police weapons. This paper examines the management of uncertainty and disagreement in relation to one case study: the deployment of CS (o-chlorobenzylidene malononitrile) incapacitant sprays in the United Kingdom.

In general, various procedures and regulations exist to protect particular workforces and the public at large from health risks associated with innovations (McVeigh & Wheeler, 1992; Richardson, Ogus & Burrows, 1982). These measures include economic incentives, legal codes, administration procedures (e.g., licensing), and self-regulation (see Hawkins & Hutter, 1993). Few of these controls are relevant to the case of non-lethal weapons, which are often evaluated by policing agencies traditionally concerned with questions of operational effectiveness. With the increasing interest in such technologies though, this situation is arguably becoming inappropriate.

Following the work of John Abraham (1998), this paper advances a critical analysis on the basis of regulatory agencies failing to meet the standards of rigour they have declared for themselves. While not suggesting manufacturers or government officials have misled, acted dishonestly or in 'bad faith', it is argued here that the organisational structures in place have provided a less than adequate regime for assessing the effects of the sprays. This paper contends that the development of non-lethal weapons merits further attention by those interested in the intersection of health and regulatory issues.

INTRODUCTION OF CS SPRAYS

Much of the recent impetus for the adoption of incapacitant sprays began in the early 1990s when individuals in the British police began pressing for the adoption of personal incapacitant sprays. A wide variety of such devices had been used in the US and continental Europe for a number of years. The introduction of a spray was thought to be able to reduce the number of assaults on officers, help minimise injuries to officers and members of the public, and forestall the routine arming of the British police.

In 1995 the decision was taken by Home Office on advice from its Police Scientific Development Branch (PSDB) to adopt a 5 percent (weight/volume) CS spray. The common name of 'CS' derives from the two chemists who first synthesised it, Corson and Stoughton. Like most other chemical sprays, the British version consists of three components, in this case the primary chemical agent (CS), a liquid solvent (methyl isobutyl ketone [MIBK]) and a propellant gas. The French Gendarmerie had used the 5 percent spray formulation since the late 1980s and it was therefore deemed to be safe.

Initial internal testing of sprays began in early 1995, but was called off due to a number of serious injuries to officers who later ended up claiming compensation. In March 1996 operational trials began in England and Wales. In August 1996, one month before the end of the trials, the Home Office approved the incapacitant. Then Home Office minister Michael Howard stated that they represented a 'dramatic improvement in police protective equipment...presenting no serious risk to health' (quoted from PCA, 2000, p. 1). The current Labour government later reaffirmed such sentiments. As then Home Office Minster Alun Michael said "CS spray has been scientifically tested to a level similar to that which would be required for a new pharmaceutical drug, and there is no evidence that it poses any significant threat to human health" (BBC News, 1998). To allay fears of their widespread use, the sprays were only supposed to be employed in highly specific contexts and specific ways for the defence of members of the police and the public or in handling highly dangerous situations. In addition, officers and police doctors were supposed to provide aftercare for recipients and pay particular attention to certain population groups (e.g., those with contact lenses, asthmatics).

Much debate has taken place on the short and long term health implications of the sprays. Before the trial began, CS itself was associated with a variety of acute effects including a severe burning discomfort, excessive lachrymation, cough, erythema, dyspnoea and sometimes blepharospasm. It is associated with conditions such as bronchospasm in asthmatics. The possible chronic effects are less well known.

The Himsworth Committee conducted the most comprehensive British public study of CS after the 1969 Derry riots. In light of widespread use of CS smoke grenades (so-called 'tear gas'), the Committee investigated possible negative effects. While generally supportive of the relative safety of CS smoke or aerosol against healthy persons in an open space, the Committee acknowledged limitations in knowledge about CS. Noting the uncertainties and speculative character of much of the evidence, it recommended that "…the medical and scientific research relevant to this decision should straight away be published in the appropriate scientific journals" (Himsworth Committee, 1971, p. 48). It further recommended such chemical agents should be regarded as being more akin to drugs than weapons. While the exact meaning of what 'akin to a drug' entails is a matter of some disagreement (see below), arguably any such process would consist of the pre-approval testing of known risks and the post-approval monitoring of effects.

Since the time of the Himsworth Committee, CS gas has been linked with permanent but non-lethal lung damage at a comparatively low dose (Jason-Lloyd, 1991), prolonged coughs and shortness of breath, heart failure and hepatocellur damage (Hu, Fine, Epstein, Kelsey, Reynolds & Walker, 1995) and those with aneurysms and those of an older age groups here particularly at risk (Ballantyne, Gall & Robson, 1976).

DESIGN AND METHODOLOGY

The introduction of CS sprays has been one of the most significant developments in relation to the use of force by British police officers since the Himsworth Committee. Given the majority of officers in mainland UK are not equipped with firearms, the introduction of the sprays has attracted quite a significant amount of public attention. This has led to a number of individuals and organizations commenting on the merits of the devices. As will be illustrated, much of the discussion of the sprays has been conducted against a general backdrop of limited information. As such, general statements about the sprays need to be scrutinised in some detail.

The case study approach adopted in this paper allows for a thorough consideration of the claims and counterclaims surrounding the effects of the sprays. This enables one to see how limited and conditional evaluative statements about their effects have been transferred between police, legal and media domains. As is argued here, the uncertainties and disagreements associated with the sprays have not simply resulted from scientific unknowns but have also been the consequence of inadequate and confused regulatory procedures as well as a lack of transparency.

In addition to a review of published medical literature, government-related research, journalistic accounts and other sources, 34 semi-structured interviews were conducted during 2000-1. Most interviews were done with those active in publicised cases of the use of sprays, organisational spokespersons, or other key individuals. Seven interviews were conducted with toxicologists and clinicians who had experience in treating cases of those sprayed or who had otherwise evaluated their safety. Two lawyers representing members of the public and the police gave detailed commentary on the legal standards of proof in relation to claims from exposure to the CS sprays. Two senior journalists who had written numerous articles about the devices provided additional evidence regarding particular claims in the media. Six interviews were conducted with non-governmental organisations regarding their assessments of the incapacitants. The use of force training officers in the two police forces in England that have not adopted the sprays were questioned in order to assess why their force had so declined. Four senior officers (rank of inspector and above) with responsibility for managing the devices were asked to assess their forces' experiences. Three interviews were done with street officers identified by other contacts who experienced serious reactions (i.e., requiring time off work) in order to assess how their claims for compensation were handled. Representatives of the Association of Chief Police Officers, the Police Complaints Authority, the Department of Health, the Committee on Toxicity, the Health and Safety Executive, a Member of Parliament (MP) and a manufacturer gave further interviews. Correspondence was also undertaken with MPs, the Police Federation, and lawyers representing individuals sprayed.

Interviewees provided the author with a range of documentation that, while being ostensibly 'public' in character, would be difficult to access otherwise. Although much of the data collected in the interviews is not explicitly presented in this paper, the responses given were vital in being able to formulate and validate the argument given below. Only after a number of years since the sprays' introduction has it become possible to piece together a comprehensive picture of the regulation procedures undertaken.

This paper is only concerned with a limited range of the effects of the weapons: those related to their direct health implications for users and recipients. There have been questions raised about the indirect effects of the sprays, such as whether their use has resulted in an escalation of force by the police. While these are important issues, they are beyond the scope of this paper (however, see Rappert, 2002). Operational questions about the use of the sprays are only considered in so far as they are relevant to assessments of their health implications. Furthermore, this paper is only concerned with the first five years of the spray's deployment (April 1996-2001). At the time of writing, much deliberation is still taking place within the UK about the relative merits of non-lethal weapons and it is likely to continue for some time. A consideration of these initial years though, provides an illustration of the range of safety tests and monitoring mechanisms established. These will provide the information on health effects that will inform future discussions.

THE EVALUATION OF CS SPRAYS

In light of the history of the research into CS given in the second section, let us now turn to a consideration of CS sprays in particular. This section assesses the basis of such claims. Following the basic procedures in place for the approval of drugs, it does so by examining both the preliminary safety testing of the sprays, their trials and the post-approval surveillance mechanisms in place.

Safety Testing and Research Reviews

To begin with the spray fell outside the jurisdiction of the Medicines Control Agency, the organisation responsible for granting drug approval in the UK and monitoring adverse reactions. Because of this and the failure of the Home Office to elaborate the types of tests carried out pre-approval, the basis for safety claims has been unclear. This uncertainty, in part, contributed to disputes about the merits of the sprays.

Various grounds for concern have been made in the medical literature about the experienced or potential effects of the sprays. These raise pressing questions about the exact nature of the tests done. It can be noted initially that the effects of CS depend on the dosage received and the method of its dissemination. A key point is that sprays entail far higher levels of irritant exposure than aerosols and lead to additional skin and eyes problems. As a simple means of distinguishing these two, the former spreads liquid and gas particles (e.g., as from a water sprayer for plants) while the latter disseminates mainly gaseous particles (e.g., as from a hairspray aerosol can). Thus many criticisms of the British CS spray have related to its specific product specifications. Jones (1997) has argued that the 5% CS concentration is far too high and much confusion has resulted due to equating the effects in spray and aerosol application forms. The French made spray was given a specification in the UK which demanded that it be a 5% solution and release 5 centilitres of fluid per burst which compares with US versions which contain a 1% solution and release a 1% burst. This means anyone targeted in the UK receives 25 times more irritant that in the US. Gray (1997) argued the sprays would not be used in situation imagined by the Himsworth Committee, and suggested that its supportive conclusions about the effects of CS were not applicable. In combination the effects of CS and MIBK were said to potentially be quite severe. Little research, for instance, exists about the carcinogenic potential of repeated exposures to the spray combination in relation to the skin and airways.

Besides these considerations, other worrying issues were raised during the time of the initial introduction of the sprays. While the French Gendarmerie has not systematically monitored the effects of its spray, cases of severe dermatitis and extensive blistering lasting for several days have been attributed to CS sprays in France (Pareix-Spake, Theisen, Roujeau & Revuz, 1993; Trevisick, 1996). In a report in 1996, the West Midlands Health and Safety Advice Centre called into question the safety of the sprays. The Centre found the manufacturer's documentation that claimed "In effect, the solution used has been retained for its harmlessness to the skin, mucous membranes and especially the eye" (WMHSC, 1996, p. 3) quite dubious. Among other potential effects, the report cited the possibility for chemical sensitisation and chromosomal damage (Schmid & Bauchinger, 1991).

Still, despite such concerns it was repeatedly argued that the sprays were safe. No public statements, however, had been given out pre- or post- trial regarding the basis for such statements. Thus critics of the sprays were left openly wondering about such procedures and therefore the legitimacy of the critical medical commentary made in relation to the rigour of official tests (e.g., Wadham, 1996).

Drawing on Parliamentary questions and correspondence (Boateng, 1999ab; Jowell, 1999) it is now possible to describe the extent and nature of testing on which safety claims were based. The pre-trial scientific assessment consisted of three strands: a) expert opinion (not formal research) from the Chemical and Biological Defence Establishment (Porton Down) on the likely toxicity of the combination of MIBK and CS; b) Department of Health (DoH) advice on the safety of MIBK based on "data available in standard toxicological reference books" (Jowell, 1999); and c) a Porton Down review of riot control options. The latter was complied for Jill Tan of the Police Scientific Development Branch and evaluated CS against other agents for use in personal aerosol canisters (Rice & Jugg, 1994). As stated "The aim of this report is to review the readily available information regarding the toxicology of CS in the context of its use as an incapacitant delivered from small, personal pressurised canisters as a liquid aerosol" (emphasis mine). In focusing attention on aerosols the review would appear to have limited relevance for sprays or conflated them with aerosols. The review further specified the need to monitor those exposed to heavy doses of CS for some time. Although solvent options were not considered in any detail, the report did note in passing that while methylene chloride (MC) was (and still is) the solvent for CS canisters in the British military service, it was believed to be too hazardous for a civilian context.

As cited in support of the research process behind the sprays in Parliamentary correspondence (Boateng, 1999ab; Jowell, 1999), two further Porton Down studies were commissioned after the trials began. The studies were initiated following internal police training sessions with the CS sprays mentioned above that resulted in injuries including "delayed symptoms commencing approximately six hours after exposure and lasting for several days" (Rice, Dyson & Upshall, 1996). Only after Parliamentarian requests were both studies placed in the House of Commons Library in 1997 along with the 1994 Porton Down report and thus accessible to Members of Parliament, Lords and select others. The first study (A Review of the Toxicology of Methyl Iosbutyl Ketone and Methylene Chloride) was completed in July 1996 (the month before formal Home Office approval) and compared the toxicity of MC and MIBK for use in aerosol devices. MIBK exposures were associated with nausea, headaches, respiratory irritation and at high exposures can cause vomiting and diarrhoea. Although MC was ruled out due to health concerns, the study concluded that it posed a "significantly reduced risk" compared to MIBK. The authors "strongly recommended that any future trials of hand-held aerosol devices... should contain MC rather than MIBK." In 1998 it was reported that the DoH overrode this conclusion (Gillian & Evans, 1998, p. 1), yet former Home Office Secretary Michael Howard was reported not to have any knowledge of the review (Clark, 1998, p. 1).

In 1997 a second study (A Literature Review of Solvents Suitable for the Police CS Spray Device) reviewed available literature on the toxicity of possible solvents for aerosol devices, taking into account classified research conducted on CS at Porton Down and in the United States. The authors concluded that

"...it is possible to exclude a number of potential solvents on the basis of their toxicology. Several of these solvents are either confirmed or suspected carcinogens with associated mutagenic potential and clearly do not represent safe alternative solvents; we can, therefore, excluded methyl isobutyl ketone [MIBK]..."
(Rice, Jones & Stanton, 1997, p. 7).

The review recommended that di(propylene glycol) and polyethylene glycol were the only acceptable known solvent options.

Shortly before the 1997 study was reported in the media in November 1998 and in response to persistent criticisms of the sprays (see below), in October of that year the Department of Health referred the safety of the sprays to its Committees on Toxicity (COT), Mutagenicity and Carcinogenicity for a review of existing research. The review considered CS, MIBK, and the combination of the two in the existing spray form. After numerous delays, almost a year later the committees led by COT found

"the available data did not, in general, raise concerns regarding the health effects of CS spray itself...It must be noted that no comprehensive investigation of the effects of CS spray in humans was available, nor has there been any systematic follow-up of individuals who have been sprayed with CS spray"
(COT, COM & COC, 1999, p. 11 - emphasis in original).

The Committee further

"noted the sparsity of data on the combination of CS dissolved in MIBK. There are no data available on the metabolism, kinetics, acute toxicity, or skin irritancy of CS when administered in MIBK solution."

Concerns were expressed for those suffering from asthma, chronic obstructive airways, hypertension, or other forms of cardiovascular disease. The potential for dermatitis from multiple exposures was also acknowledged. All of the Committees' conclusions were based on the use of the sprays in accordance with the operational guidelines, though no attempt was made to determine whether this was the case. Given numerous uncertainties about effects, they suggested that follow-up studies be conducted on those sprayed to see if delayed effects did take place. At the time of writing (earlier 2002), no such information had been made available. This is despite the identification of delayed symptoms before the trial in 1996 as mentioned above.

In a press release the Home Office welcomed the committees' findings and said the report supported their position that there was "no reason to prevent the police service from using CS spray" (Home Office, 1999). Although the 1996 and 1997 Porton Down studies formed central planks of the government safety research process, neither were analysed or acknowledged. Little has been said about these Porton Down reports in public. In July 1999 the Home Office did comment that the 1997 study was only a 'theoretical' exercise (BBC, 1999). If accurate this raises questions about why the study was commissioned at all. The new Scientific Secretary of the COT appointed since the report indicated that the committee has a policy of not drawing the conclusion of literature reviews and instead seeks to analyse 'primary' research material (Benford, 2001). Yet, despite this there are commercial studies cited in the Porton Down reports that were not cited in the 1999 COT study. Given that the findings of the 1996 Porton Down report were overridden by the DoH, the 1997 study was downplayed while the 1994 study referred to CS aerosols rather than sprays, it appears there is little research that can justify the safety statements made since the CS sprays' introduction regarding the rigorousness of the testing.

Monitoring of operational use

Because of the uncertainties, unknowns, and questions raised by the review of the safety research above, the provisions made for monitoring the operational use of the sprays become all the more important in terms of justifying the benign effects of the sprays. As already suggested, little systematic information exists about their implications and what is available offers grounds for concern.

The official operational trials began in March 1996 and involved approximately 4,000 officers in 16 forces. The analysis of the trials of the spray, A Review of Police Trials of the CS Aerosol [sic] Incapacitant (Kock & Rix, 1996), found a relatively acceptable risk of injury from CS spray for officers and the public. It did determine though that 78% of officers in trials experienced some amount of cross contamination due to problems of aiming the sprays, the persistence of CS (e.g., on recipients clothing), and the release of pressured gas particles in addition to the spray liquid. The review did not monitor the duration of effects or delayed complications, despite the Home Office being aware of the possibility of such reactions (see as well Gregory & Knill, 1998). The review was also unable to comment on the effects of multiple exposures. Figures from the trial indicated the average officer holding the sprays would use it once every 32 months and be exposed to some degree to the CS and MIBK mixture every six months because of cross contamination. This would suggest that the deleterious effects from multiple exposures would only arise after some time.

The health conclusions of the trial review stand in sharp contrast to a number of other studies. In 1996 a Dispatches television programme surveyed a sample of those sprayed during the trials and found two-thirds suffered from skin blistering and breathing difficulties three days after exposure (Trevisick, 1996). Only 2 of the 34 people interviewed reported recovering in the 15 minutes time period. The study also found the sprays used quite frequently outside of the guidelines established to ensure their safety: few verbal warnings were issued, two-thirds of recipients were sprayed below one metre (and thus exposed to an increased risk of eye damage), only one-third were promptly moved away from area after use, no one was questioned as to whether or not they wore contact lenses even though this was associated with increased health risks in the guidelines, and finally only one in six were given information on after care.

Various grounds have been raised for not only doubting the compliance of officers with the operational guidelines, but the quality of the after care advice give to officers. The 1996 CS spray guidelines stated: '…The person who has been sprayed should be removed to an uncontaminated area where they can be exposed to cool fresh air. This will permit the CS particles to be blown off the body' (ACPO, 1996, p. 6). As pointed out by Jones (1998), while this advice may be appropriate for those exposed to CS aerosol or smoke devices, such prescriptions are inappropriate for sprays as large amounts of chemicals are disseminated through sprays and these are left deposited on the skin. In this case, CS particles will not simply blow away but must be physically or chemically removed. Further precautions are thus needed than those traditionally associated with CS in other forms. Despite the differences between CS in spray and aerosol/smoke forms, the two have been frequently confused in debates about the merits of the sprays (as in Yih, 1995).

With a limited knowledge of the testing done and due to concerns about the sprays' operational deployment, in July of 1998 the medical journal The Lancet called for a moratorium on their use until the further details were published regarding the basis for evidence about the sprays' safety (The Lancet, 1998). Due to the sorts of uncertainties and known problems identified with the sprays, Northamptonshire and Sussex police chiefs have so far refused to introduce them and their uptake by Scottish police forces remains patchy.

A variety of studies into the operational effects of the sprays since 1996 have established additional grounds for concern. The National Poisons Information Service (London) has observed delayed severe skin reactions from CS sprays (Euripidou, 2000) consistent with those witnessed by Ro and Lee (1991). Instances of ocular damage (Gray, 2000) and delayed drying, flaking, and blistering (Worthington & Nee 1999) have also been documented. One officer who experienced multiple exposures was reported as suffering from allergic sensitisation and no longer able to work (Jenkins, 1999).

Such concerns have come along with repeated media and anecdotal claims about the questionable uses of the sprays. Kossoff (1998), for instance, claimed that mentally ill persons, pensioners and children have all been on the receiving end of the sprays, in situations that posed little threat to police officers. The Police Complaints Authority (2000) has conducted an analysis of the operational use of the sprays. In an examination of a limited range of complaints against the police in one year, it found significant deviations from the original guidelines designed to minimise safety risks and evidence of the spray being used in situations for which it was not intended (e.g., in 30 percent of cases the sprays had been used in a distance less than one meter (thus raising the possibility of ocular damage and amount of chemical exposure). In its report, the PCA recommended in the future all police forces should record injuries to the public in use of force reports. This has not taken place. The devolved governance of the UK police means that what information is held varies significantly across forces. Even if injuries were recorded, unless that data included delayed effects the results would be of limited utility in evaluating CS sprays. The National Poisons Information Service plays a surveillance role with regard to this sort of chemical technology, but the centres under this organisation only work as a responsive service for hospitals and are not in a position to evaluate the overall safety of the sprays. Individual accident and emergency departments do not collect data on injuries from police CS sprays (or any incapacitant sprays more generally) in a manner that is available for analysis.

Since the time of the PCA report, new guidance (rather than previous guidelines) has been issued to officers regarding the appropriate use of the sprays. There are two major changes that give greater latitude in the spray's scope of deployment. The first is that the basic rationale for the legitimate use of the sprays has gone from self-defence of officers and members of the public to the perception of violence. The second is that the guidance given is meant to be suggestive to officers and forces, rather than being binding or prescriptive.

Due to potential for long term problems arising from multiple exposures and the high rate of cross-contamination, the most at risk individuals from the sprays are police officers. Unfortunately no centrally held public information exists on the number of legal proceedings started by the public or officers involving CS sprays (Boateng, 1999c). In the absence of such data, it is difficult to know how large the health problem might be. Various stories of police and public settlements have been made in the media (see, e.g., Langley, 2000).

There is reason though, to think that the number of such cases is likely to be artificially low due to the advice given to officers and local Police Federation officials regarding the legitimacy of claims. A Memorandum written by the legal firm that represents officers states that they are no reasonable prospects for suing the police on the basis of the sprays being 'unsafe' (Care, 2000). The memorandum draws on the 1999 COT report and a 1983 study of CS to support this position. It quotes from the COT report that 'There was no evidence of mutagenicity, carcinogenicity or tetatogencity. It was considered that the available data did not, in general, raise concerns regarding the affects [sic] of CS spray itself'. However, no mention is made of the conclusion of the COT regarding the 'sparsity' of data on CS and MIBK as a spray. No mention is made of the recommendation of the report to establish information on delayed effects and the failure to introduce follow-up studies. Neither of the critical Porton Down reviews mentioned above are cited. Throughout the memorandum, CS sprays are referred to as CS aerosols and evaluated in terms of the potential effects of CS rather than CS sprays.

Certain steps have been taken to minimise the health effects posed by the sprays for the police. To familiarize officers with the effects of the sprays and to prepare them for cross-contamination, during training sessions officers are exposed to the sprays. Initial pre-trial training tests involved officers being sprayed directly in the face, as this is standard procedure for using the sprays against members of the public. After a number of serious injuries, the training policy moved from the policy of direct to 'indirect' exposure. The 1996 Guidelines state that exposure "should be achieved without spraying the irritant directly into an officer's face" (ACPO, 1996, p. 4). In practice, this means trainee officers walk on a patch of ground that has been sprayed. According to the Home Office Police Policy Directorate (1998), only a 3 percent rather than a 5 percent spray is used for the purposes. In addition, training sprays contain MC as a solvent rather than MIBK. While these steps might reduce hazards associated with the sprays, they also reduce the likelihood that possible problems associated with exposures to CS sprays would become obvious to central police management.

The possibilities for unintentional effects extend beyond those involved in the immediate encounter. Due to the lingering effects of CS, custody officers, medical practitioners, and others must take precautions in handling people and material exposed. An internal study undertaken by Sussex police force in early 1998 identified a number of problems experienced with the sprays in other forces including food contamination, hospital ward contamination and several individuals being taken off operational duties because of exposure to CS spray (Gregory & Knill, 1998).

As the discussion above indicates, nothing approaching the scheme for monitoring adverse reactions of pharmaceutical drugs exists in the case of CS sprays. The speed and type of advice given on decontamination procedures betrays the lack of urgency that characterises much of the control of the sprays. After nearly five years of use, guidelines issued by Police Scientific Development Branch in late 2000 suggested using sodium metabisulphite as a decontaminant for vehicles and buildings, a possibility first identified for buildings and humans in 1991 (Jones, 1991).

DISCUSSION AND IMPLICATIONS

In response to persistent criticisms about the potential and experienced effects of CS sprays, officials in the UK have given constant reassurances about their safety and the robustness of the mechanisms enacted. Much of the force of these reassurances has stemmed from making reference to the controls in place for monitoring the safety of pharmaceutical drugs. As has been argued in this paper though, statements to the effect that the sprays have been or are regulated in a manner similar to that of a pharmaceutical drug are highly questionable. These rely on dubious assumptions about the tests done for CS and MIBK a) on their own and b) not in a spray application form. Instead of a thorough research protocol, approval of the sprays was largely based on expert opinion that was later determined to be wanting by the same experts while claims about the safety of CS and MIBK together have been based on a 'sparsity' of information. The existing state of surveillance mechanisms in place foster little in the way of understanding about the risk for long term implications.

Policy choices are often made in situations of uncertainty and disagreement, where the costs and benefits of particular options are not always known in advance (Burch & Wood, 1989). What matters in such circumstances is whether organisations take proper initial precautions and learn from their experiences (see Lindblom & Woodhouse, 1993). The past and current regulatory regimes surrounding CS sprays in question afford little reason for assuming either condition is being met. The extent to which the more long term and severe effects related to multiple exposures will manifest themselves is uncertain.

The characteristics of the CS spray debate are surprisingly similar to those found for Bovine Spongiform Encephalopathy (BSE) and variant Creutzfeldt-Jakob Disease (vCJD) in the UK. A public inquiry into BSE/vCJD (Phillips, 2000) found inadequate and untimely department response measures, unclear risk implications being stated from research, a lack of proper surveillance procedures, a non-overt decision-making process and an unshakeable belief human life was not threatened. As with the BSE, the absence of substantial indicators noting immediate and severe negative consequences has been taken as an indication of safety while potential critical signs have been marginalized.

In response to this and other scientific controversies in recent years, government agencies have reappraised the manner in which scientific endeavours should be regulated and how technical advice ought to be handled. Much critical commentary has been made of the British "culture of secrecy" and "bunker mentality" in relation to policy decision-making processes. In the past the overall British government system for the regulation of risky environmental and medical technologies was characterised by trust-based procedures. Arguably this situation limited the public accountability of officials (Jasanoff, 1991). According to government officials, this situation is said to have changed now. Former Trade and Industry Secretary Steven Byers (2001), for instance, called for open and transparent policy making with regard to risky technologies, where uncertainties are shared with the public.

There are major questions though about how such sentiments can be realised in practice. The case of CS sprays raises concerns about the 'public' character of knowledge. As indicated above, while decision-making procedures and information about the sprays were supposed to be made 'public', what counts as public is contentious. Even years after the initial introduction, the original justifications for adopting the spray are still unclear. Had the basis of CS sprays' approval been widely known when approval was given (as recommended by Himsworth, 1971), this would have no doubt led to further questions being raised as well as calls for more rigorous surveillance procedures than those enacted. The case of the sprays has been characterised by a slow trickle of information that has not been evaluated together in systematic fashion.

In response to persistent uncertainties and various health concerns, individual police forces such as Surrey, Hertfordshire, Northamptonshire and Sussex have initiated research into alternative incapacitants. Currently, the Sussex police force is leading such initiatives. In April 2001 it began trials of spray called PAVA, a synthetic form of oleoresin capsicum or 'pepper spray'. Despite the PAVA spray passing numerous safety tests carried out on advice from Porton Down, the Home Office and Police Scientific Development Branch have failed to approve the spray for the trials (Gregory, 2001). The debate between Sussex and the Home Office has hinged on the question of what it means to regulate something 'akin' to a drug. It should be clear that this a matter of some considerable debate.

While this paper has focussed on a particular incapacitant device in a specific setting, its implications are more general. As stated in the introduction, the last few years have seen increasing attention to non-lethal weapons as tools for resolving disputes about the use of force by the police. Despite such developments and the somewhat long standing use of some non-lethal weapons, there are no international standards or agreed procedures for determining the safety of such devices (Omega Foundation, 2000). Instead individual countries and police forces have been left to their own devices in determining what is safe or unsafe.

Given this international situation, devising proper procedures for the regulation of this weaponry is a daunting task. Systems for the regulation of risky technologies are not always transferable between different international contexts (Jasanoff, 1991). The limitations of the risk assessment and control procedures in place for chemical non-lethal weapons in the UK and Europe could hardly been said to be unique to these technologies (Klapp, 1992). Assessments procedures for risk in the control of industrial toxic chemicals and medical technologies, for instance, have been found wanting on many of the same grounds discussed in this paper. The reasons for this vary by country and are dependent on particular social and historical contexts (see Kammen & Hassenzahl, 1999).

Yet despite this diversity, arguably a central lesson that has come from the study of such innovations is the importance of the distribution of the burden of proof for safety (Thornton, 2000). Ensuring that the onus for safety in the first instance rests with manufactures is a basic step in improving the intelligence of policy-making processes (Monroe & Woodhouse, 1986). In relation to non-lethal weapons, this general prescription has application across a number of settings. In countries such as the US where there are a plethora of different companies marketing a wide range of chemical incapacitants and other weapons, police forces and control agencies are highly reliant on and persuaded by the (sometimes questionable) safety advice given by manufacturers (Allen, 2000; Doubet, 1997). In countries such as the UK that do not posses such an extensive private sector, decisions are often taken by particular regulatory bodies that are still reliant on inadequate commercial data about such products and their effects. One way of correcting this situation would be the introduction of further health and safety disclosure requirements for manufactures as part of a process of licensing weapons.

As social scientists examining the regulation of pharmaceutical drugs have argued though, licensing in and of itself does not guarantee adequate safeguards (Abraham, 1995). The regulation mechanisms must be backed with organisational control systems that make public the evidential basis of decisions as well as the criteria for making those decisions. Areas of medical uncertainty need to be defined and reduced. As with the monitoring of adverse reactions to drugs, creating open and rigorous systems for the 'post-marketing' surveillance effects of weapons is also necessary. In many cases, such as that of England and Wales, this will require institutional reforms in the way the use of force, injuries and compensation claims are monitored and made available to police officers and the public.

Without rigorous procedures for the regulation of such non-lethal weapons, there is the potential for quite a bit of negotiation over safety claims. As has been illustrated in this paper, in such situations the danger exists of unduly optimistic assessments of the potential effects of technology being made where scientific doubt is interpreted in favour of a liberal and poorly monitored deployment. Those social scientists and others concerned with the health implications of such technology would well be advised to maintain a certain degree of scepticism about the claims made. To counter the problems that arise from the poor state of existing regulations, there is a need for strict licensing of the production and approval of non-lethal weapons, one that is open to the highest levels of scrutiny.

REFERENCES

Abraham, J. (1998). Regulating the cancer-inducing potential of non-steroidal anti-inflammatory drugs. Social Studies & Medicine 46, 39-51.

Allen, T. (2000). Chemical cops. In These Times 3 April, 4-8.

Association of Chief Police Officers. (1996). Police trials of CS for self-defence in England and Wales - 1996 guidelines for use. London: ACPO.

Ballantyne, B., Gall, D., & Robson, D. (1976). Effects on man of drenching with dilute solutions of CS and CR. Medicine, Science & Law 16, 159-170.

Benford, D. (2001). Personal Interview. January 8.

Boateng, P. (1999a). Home Office communication to Matthew Taylor MP. POR/98 9/11/4 8 January.

Boateng, P. (1999b). Home Office communication to Matthew Taylor MP. POR/99 9/11/1 11 March.

Boateng, P. (1999c). CS sprays. House of Commons Hansard 27 January, London: HMSO, column 449.

British Broadcasting Corporation (BBC) News. (1999). Life-saver or deadly tool? BBC News, 8 July <www.bbc.co.uk&gt;.

Burch, M., & Wood, B. (1989). Public policy in Britain. Cambridge, MA: Basil Blackwell.

Care, A. (2000). Memorandum: CS briefing note to staff. December, London: Russell, Jones & Walker.

Clarke, C. (1999). Incapacitant sprays. House of Commons Hansard 30 November, London: HMSO, column 106.

Committees on Toxicity, Mutagenicity, and Carcinogenicity of Chemicals in Food, Consumer Products, and the Environment. (1999). Statement on 2-chlorobenzylidene malononitrile (CS) and CS spray. London: HMSO.

Coupland, R. (1996). The effects of weapons on health. The Lancet 347, 450-451.

Doubet, M. (1999). The medical implications of OC spray. Millstadt, IL.: PPCT Management Systems.

Euripidou, R. (2000). Personal Interview. Chemical Incident Response Service 14 November.

Gilligan A., & Evan, R. (1998). Warning on CS gas ignored by police. Independent 19 April, 1.

Gray, P. (1997). CS gas is not a chemical means of restraining a person. British Medical Journal 314, 1353.

Gray, P. (2000). Formulation affects toxicity. British Medical Journal 321, 46.

Gregory, P., & Knill, L. (1998). CS incapacitant: should Sussex police adopt CS incapacitant? Lewes: Sussex Police.

Gregory, P. (2001). Personal Interview. 3 April.

Gunningham, N., & Grabosky, P. (1998). Smart regulation. Oxford: Clarendon Press.

Hawkins, K., & Hutter, B. (1993). The response of business to social regulation in England and Wales. Law & Policy 15(3), 199-218.

Himsworth Committee. (1971). Report of the enquiry into the medical and toxicological aspects of CS. Part 2 Cmnd 4775 London: HMSO.

Home Office Police Policy Directorate. (1998). The management of health and safety in the police service: Volume 3. London: HMSO.

Home Office. (1999). Home office welcomes independent scrutiny findings on CS spray. Press release. 294/99 23 September.

Howe, S. (1997). CS alternative. Policing Today December, 10-11.

Hu H.,, Fine J., Epstein, P., Kelsey, K., Reynolds, P., & Walker, B. (1995). Tear gas: Harassing agent or toxic chemical weapon? JAMA 262, 660-63.

Jasanoff, S. (1993). Acceptable Evidence in a Pluralistic Society. In Mayo, D., & Hollander, R. (Eds.), Acceptable evidence. (pp. 29-47) Oxford: Oxford University Press.

Jason-Lloyd, L. (1991). CS gas - an indiscriminate weapon? New Law Journal 15, 1043-1045.

Jenkins, C. (1999). Pain reaction. Police Review 1 October, 18.

Jones, R. (1991). CS Gas: An antidote and decontaminant. Military Medicine 156(11), A6-7.

Jones, R. (1997). Are CS sprays safe? Lancet 350, 606.

Jowell, T. (1999). Department of health communication to Matthew Taylor MP. POH(3)5592/141 30 March.

Kammen, D., & Hassenzahl, D. (1999). Should we risk it? Princeton, NJ: Princeton University Press.

Klapp, M. (1992). Bargaining with uncertainty. New York, NY: Auburn House.

Kock, E. & Rix, B. (1996). A review of police trials of the CS aerosol incapacitant. Police Research Series Paper 21 London: Police Research Group.

Kossoff, J. (1998). Police now use CS sprays more than truncheons Independent on Sunday 9 August, 1-3.

Lancet. (1998). "Safety" of chemical batons. The Lancet 352, 159.

Langley, C. (2000). Blinded Pc sues met over CS raid farce. Evening Standard 22 June.

Lindblom, C., & Woodhouse, E. (1993). The policy-making process. 3rd edition Englewood Cliffs, NJ: Prentice Hall.

McVeigh, S., & Wheeler, S. (Eds.) (1992). Law, health and regulation. Aldershot: Dartmouth.

Morone, J., & Woodhouse, E. (1986). Averting catastrophe. London: University of California Press.

Omega Foundation. (2000). Crowd control technologies. Report to the Scientific and Technological Options Assessment of the European Parliament. PE 168.394 Luxembourg: European Parliament.

Parneix-Spake, A., Theisen, A., Roujeau, J.C., & Revuz, J. (1993). Severe cutaneous reactions to self defence sprays. Arch Dermatol 129, 913.

Phillips, Lord (2000). BSE inquiry report. London: HMSO.

Police Complaints Authority. (2000). CS spray: Increasing public safety? London: Police Complaints Authority.

Rappert, B. (2002). Constructions of legitimate force. British Journal of Criminology, (forthcoming).

Rice, P., & Jugg, B. (1994). A review of the toxicology of the riot control agent o-chlorobenzylidene malononitrile (CS gas). Salisbury: Chemical & Biological Defence Establishment.

Rice, P., Dyson, E., & Upshall, D. (1996). A review of the toxicology of methyl iosbutyl ketone and methylene chloride. Salisbury: Chemical & Biological Defence Establishment.

Rice, P., Jones, D., & Stanton, D. (1997). A literature review of the solvents suitable for the police CS spray device. Salisbury: Chemical & Biological Defence Establishment.

Richardson, G., Ogus, A., & Burrows, P. (1982). Policing pollution. Oxford: Clarendon Press. Thornton, J. (2000). Pandora's poison. London: MIT Press.

Schmid, E., & Bauchinger, M. (1991). Analysis of the aneuploidy inducing capacity of 2-chlorobenzylidene malonitrile (CS) and metabolites in V79 chinese hamseter cless. Mutagenesis4, 303-5.

Spellar. (1997). CS Gas. House of Commons Hansard 6 November, London: HMSO, column 316.

Thornton, J. (2000). Pandora's poison. London: MIT Press.

Trevisick, S. (1996). Dispatches: The truth of CS. London: Channel 4 and Liberty Publication.

Wadham, J. (1996). Preface. In Dispatches. London: Channel 4 and Liberty Publication.

West Midlands Health and Safety Advice Centre (WMHSC). (1996). CS incapacitant spray: Potential health hazards. Birmingham: West Midlands Health and Safety Advice Centre.

Wood, M. (2000). CS Gas. House of Commons Hansard 12 May, London: HMSO, column 1187.

Worthington , E., & Nee, P. (1999). CS exposure: Clinical effects and management. Journal of Accident and Emergency Medicine 16, 168-170.

Yih, J.P. (1995). CS gas injury to the eye. British Medical Journal 311, 276.