1995 VOL 6 NO 1

Copyright© IJDP Ltd. The following pages are produced in cooperation and with approval of the International Journal on Drug Policy Ltd.

DRUG WORKERS AND THE DRUG MISUSE DATABASE

Tim Crabbe, Drug Misuse Research Unit, University of Manchester, UK

INTRODUCTION

Many of the data that support research findings in the drug field in the United Kingdom are based on the analysis of questionnaires which are routinely returned to the network of regional 'drug misuse databases'. These 'database forms' are completed by staff in a variety of drug services, whenever a drug user presents to the agency with a problem for the first time or after an absence of 6 months or more. They include a series of questions relating to the user's social background and drug use profiles. Although the questions are kept under review and were originally prepared in consultation with a whole range of interested parties, they are not chosen by the participants in the interview process. They belong to the academics who created the database system and, in this sense, reflect their values and interests, which may, or may not, be the same as those who ask and answer the questions (McDermott and Pyett, 1992, p. 136).

The significance of this consideration was apparent during the course of an assessment of the validity and quality of data held by one of the largest of these regional databases and of the understanding that agency staff have of how the system works. The study, which was conducted in the region covered by the former North Western Regional Health Authority, was part of a wider project concerned with the preparation of a UK national focal point for the European Monitoring Centre for Drugs and Drug Addiction. It was primarily interested in whether existing drug misuse reporting systems provide an accurate picture of the number and type of drug users presenting to drug services.

Satisfying the demands of this brief entailed the utilisation of largely quantitative methodologies, involving the collection of data from agency case notes and the distribution of structured questionnaires to service managers. The researcher's sociology background and experience of qualitative research techniques led to an additional awareness of the 'non-quantifiable' factors that might influence the quality of data collected. This appeared to be an equally valid area of research and one that was considered through the utilisation of observational research methods while based at each of four Community DrugTeams (CDTs) in the region, for a period of up to 2 weeks. As Jones and Power ( 1990, p. 13) have suggested. 'All too often such information database, some workers felt that clients were more comfortable if the interviews were conducted in a less formal manner, where general notes were taken and recorded on the forms later.

There was very little evidence of any systematic approach to this issue at any of the teams that were visited. It was more a case of workers using which ever approach to the completion of database forms that appealed to them and their perception of their clients' preferences. The practice of using the agency copy as the front sheet in the client notes meant that most workers preferred to complete the forms during the assessment interviews, recording additional information separately as appropriate. Those workers who completed the forms after the initial assessment readily admitted that this sometimes resulted in their failure to obtain all the information required by the database. However, they also felt that they got a more honest account from the clients during the interviews.

Difficult questions and honest answers

Although most of the data categories included on the database forms mirror the information requirements of the agencies in question, there are certain areas of questioning that remain problematic. The concerns expressed were not generally related to the experiences of the workers asking the questions, but to the discomfort of the clients as observed by those workers. Comments from members of the drug teams suggested that the questions on the database forms relating to occupation and legal status generated a feeling of unease that threatened to jeopardise the honesty of the responses:

'It's [the client's main source of income] so obvious that the clients can get a bit jumpy . . . a bit suspicious as to what the information is for.'

'It's a bit presumptuous to ask. A bit obvious really, but the only time they'll tell you they mug old ladies is when they can't have a script! The truth isn't important, they'll just tell you what they think you want to hear.'

Similar comments were made in relation to questions about the clients' legal status.

This situation did leave some workers feeling uncomfortable with the questioning process but most accepted it on the basis that interviewing clients was 'part and parcel' of the package of planning their treatment. The problem is rather tenuously linked to the database in any case, given that the database forms do not actually ask about main source of income. They ask whether the client is employed and in what occupation. This illustrates how the forms can act as a spring-board to further questioning that is of more relevance to the questioner.

The concern expressed about the degree of honesty of clients' answers to the questions referred to above is more widely recognised in relation to questions about the sharing of injecting equipment There is an overwhelming consensus of opinion that clients' responses to questioning about the sharing of equipment are notoriously unreliable. There is little doubt about the importance of the information and the need to gather it, but there is a great deal of concern about the timing and method of collection. As one worker put it:

'It's just too soon to ask those sorts of questions. They [the clients] just want a script and think they won't get one if they share [injecting equipment], so they just tell you what they think you want [to hear]. If these questions were left out and asked later on, when a treatment programme had begun, it would be more accurate.'

Another more sporadic area of concern was related to the collection of information about clients' ethnicity. Some workers saw this issue as an irrelevance because of the fact that the area covered by their 'team' had an almost entirely white population, and in some cases workers chose not to record clients' ethnic group on the database forms at all. The point was also made by at least two workers that the information might be used in ways that would work against the interests of those communities that the data were originally intended to serve. This was rationalised on the basis that policy makers could potentially use the information to argue that drug problems were inherently and specifically related to the behaviour and culture of black communities.

These workers were reassured by the explanation that the information was collected to identify any under representation of ethnic groups among the client base of drug services, enabling policy-makers' plans to be influenced in ways that would make drug services more attractive to black and Asian clients. Although this explanation was welcomed it appeared that this was the first attempt that had been made to talk through the issue in this way with these particular workers. This revealed further insights into the factors that affect workers' attitudes towards the database by highlighting the importance of communication in the research process.

Nobody raised the issue of the client's anonymity and the threat to it that some suggest is posed by the requirement to collect the clients' initials, date of birth and gender on the database forms. Although this has been an issue for database workers and the 'trade journals' for some time (Davies et al., 1993; Jones, 1994; Mott et al., 1994) it was not an issue for those workers who actually had to request the information in the agencies that were visited. Furthermore there was no suggestion that the clients themselves were in any way uncomfortable about providing this information. Nor was there any suggestion that clients would be more willing to answer the other questions honestly if they had not provided these details initially.

Clients appeared quite prepared to provide their full names, addresses and date of birth on request. It seemed that there was an assumption that this information would have to be provided to obtain treatment of any kind. Although there was no suggestion that this was an issue for clients, it was not possible to consider whether these assumptions had had any impact on the willingness of would-be clients to present to services, given the location of the research in agencies to which clients were already presenting.

What are we doing it for?

Drug workers can be a pretty cynical bunch, particularly when they are faced with an increasing workload and declining resources. In this context, monitoring and evaluation are often seen as unnecessary and time-consuming burdens which are secondary to the needs of the client group and the day-to-day business of coping with the agency case load. This attitude was reflected in the 'banter' between workers which can be found in any drug service, but not in the teams' more serious assessments of the database's importance. The system was regarded as a valuable and necessary tool which provided important information for themselves and others.

Indeed the overriding concern in all of the agencies was associated with this feedback, rather than the information that is provided by the teams or the questions that generate that information. There is a pervasive, if cynical, sense of realism, which recognises the importance of monitoring in the context of the competition for funding allocations and appraisals of service provision. The regional database is seen as an accessible source of data in the monitoring process and one that places limited demands on the resources of the agencies. The problem with this situation is that the database was designed to measure 'the extent and nature of presenting drug problems'. It was not designed to facilitate the monitoring and evaluation of drug agency workloads, which is increasingly what purchasers and consequently, the agencies themselves are concerned with.

The information that is fed back by the database is consistently distributed and generally well received. But there are several 'bug bears' associated with the mismatch between its function and utilisation, as perceived or otherwise. First, the practicalities of a system which relies upon information being gathered and sent in by drug workers, employed at agencies throughout the region, in order to generate agency, district and regional reports, mean that it sometimes takes longer to get the information to the drug teams than they would like. Almost daily telephone calls from one of the teams, who were preparing the ir own annual report, to the regional database, enquiring about the availability of the annual agency report, gave testament to these concerns, and there is no question that the situation was causing a great deal of stress and frustration. Similar attitudes were found at other teams which seemed to be accentuated by a feeling that there was an inappropriate emphasis on 'getting everything right' ahead of 'getting the data out'.

The other concerns relate to the extent to which the data produced by the system reflect the workload of the agencies. The concerns are twofold. First, th definition of a new presenting episode as the situation when a client presents for the first time or after an absence of 6 months or more means that there is no way of reporting ongoing contact with clients. A more clients coming into contact with drug agencies are held in treatment the database can generate al increasingly misleading picture of how busy individual agencies are:

'It [the drug misuse database] gives an increasingly poor impression of the agency's workload. More and more people are now held in treatment and despite this being a key DH [Department of Health] aim, it's not reflected in the database figures. Indeed there are now misrepresentations, as DMD shows declining figures for this CDT but our workload is actually increasing as more clients are kept in treatment.'

There is a general recognition that this is not the function of the database; indeed this had been explained to each of the CDTs visited on so many occasions that it was something of a standing joke. However, the database's status as the only independent, comparable source of information has led to a deep suspicion that purchasers and administrators do use the figures as the 'best indicator' of agency workloads in the determination of funding allocations. In this context there is strong support for the database designers' own suggestion that the database should move towards a system of 6-monthly reporting, regardless of whether clients have remained in contact during that period.

This, in itself, would not address the issue of how much work is put into different types of client with different types of problems. There is a concern that purchasers are becoming more interested in 'head counts' at the expense of the finer details of the work of drug services. Six-monthly reporting would not necessarily lead to any greater interest in the measurement of the scale and range of work done by agencies, on which DMD provides only limited information. Agency managers were equally concerned that this lack of measurement extended to the effectiveness of treatment provided:

'The database is limited to the first presentation and our plans are only assessed on a one off basis. There is no measure of the services we provide, ongoing contact with clients, the training we do and telephone contact. All these aspects of our work aren't recorded. Neither is there any measure of the outcomes and the effectiveness of the treatment we provide.'

There is a desire to see a shift in the balance information collected by the database to address these concerns, which seems to reflect an accommodation with the 'value for money' ethos which entered the field of drug research and policy in the 1980s (Dorn and South, 1989,p. 15):

'It [DMD] was OK in 1988 but now there is a need for outcome measurement as this is what purchasers want to know about. Some primitive goals should be set such as, "stop sharing in six months" or "drug free in six months" and then the achievement or otherwise of these goals assessed and reported. This would make the database more relevant to the drug team and purchasers.'

Although there is general agreement that some form of outcome measurements should be included on the database forms, which is shared by the researchers responsible for developing the system there are ongoing difficulties in relation to the balance of local and national needs which will be exacerbated by the further devolution of the NHS. There is no one clear perspective, just as there is no consensus towards the ways in which the information should be used:

'There is a need to consider things in a broader context incorporating outcome measures beyond such narrow definitions of "kept in contact". So what if someone's been kept in contact. What does that mean? We should be asking how has contact changed people's lives? . . . Everyone concentrates on treatment and prevention but you can't "cure" people of a drug problem, so maybe the database should concentrate on identifying trends in individual patterns of drug use . . . What are the typical careers of drug use? How long is it before speed freaks get into heroin? . . . This sort of information would be far more useful to the drug worker on the ground

These are more aspirations than criticisms though, and the general attitude towards the database among drug workers is positive, even if a little apathetic. The very fact that drug workers become most critical of the database when they have not received the latest reports demonstrates its importance to those who use it.

DISCUSSION: ATTITUDES AND INFLUENCES

The principal findings of the research project on which this article is based, which are summarised in 'The Research Context' above, provide an impressive vindication of the quality of the data held by the drug misuse database in the region that was considered. The cooperation that was offered by the drug teams that were visited and the value attached to the process of reporting to the database that was observed offered further evidence of the system's importance. Yet it would be wrong to ignore the concerns and weaknesses outlined above. Many of these concerns have nothing to do with the validity of the data collected or the effectiveness of the system that generates it, but they are important to the participants in it, and because these participants are making interventions in the research process, they are important to the future reliability of the system.

Many of the concerns expressed relate back to the boundaries to the research intervention discussed earlier and the degree of freedom that the interviewers have in that process. The concern that the database forms are having a powerful influence on the way that initial assessment interviews are conducted is the key barometer of this tension. As long as workers are comfortable with the information being collected by the database then there is little room for concern. Where there is discomfort with areas of questioning, or where the information needs of the agencies are at odds with the order and intensity of questioning demanded by the database forms, then there will be a potential for workers to become alienated by the whole process of data collection.

In this light, there is a need constantly to monitor and to review the questions that are included on the database forms and to recognise the need for flexibility in the collection of certain data items. There are a number of areas of questioning which workers are not completely at ease with. These issues need to be addressed so that the benefits of the information to the participants become clearer and those responsible for collecting the data have some ownership of the questions. Without this, there is a danger that the interviewers will not regard the data they are collecting as having any worth and so be less inclined to obtain the necessary information from clients when faced with any hesitancy on their part. It can also lead to a divorce between the interview process and the completion of the forms.

When information is collected during a separate interview and transferred to the database forms at a later date, it is likely that the questioner will impose more of his or her own values and interests on the process (McDermott and Pyett, 1992, p. 136). The questioner will be conscious of the need to collect the information required by the database forms but not bounded by its structure. This is liable to result in the collection of more information in the areas that the worker and the client are interested in, possibly at the expense of other data required by the database which may then be left out. The inability to record such additional information on the database forms, in spite of its relevance to the client's situation, may prove alienating to the worker who has collected it.

A similar spiral of alienation' is possible in relation to those questions that drug workers feel are answered dishonestly by clients who are seeking treatment. However, these concerns come from a rather different perspective in relation to the collection of information relating to patterns of sharing of injecting equipment, caused by the perceived importance of the data. In this context drug workers are highly motivated in their desire to collect the information but feel that it is being requested and recorded at an inappropriate moment in the treatment process. It is important that such concerns are recognised, both because the workers that raise them have a specialist knowledge about the validity of the data they are currently collecting and because action will generate further confidence in the system among the interviewers that sustain it.

In a similar way the feeling that the regional database does not presently reflect the scale of agency caseloads, the nature of their work or the effectiveness of the treatment provided has potential consequences for the system. It has not been the purpose of the database to measure any of these things but the environment has changed since 1987 when the system was first established. An overwhelming proportion of agencies in the area covered by the former North Western Regional Health Authority now use the information provided by the database to assist with their own funding applications, and an equal number believe that the information is used by purchasers for similar purposes. In this context, unless there is some other widely recognised independent assessment of the work done by agencies, such as the University of Manchester's Drug Misuse Information System (DMIS - an agency based computerised client management system), there will be an incentive on the part of agencies to over-report to the regional database.

That is not happening at the moment, as is evidenced by the full results of the research on which this article is based, and this represents a good position from which to defend the system from the pressures discussed above. The regional databases represent the first systematic attempt to monitor the extent and nature of presenting drug problems in the UK and their success has fed an explosion in the desire for more information, monitoring and evaluation. These data are wanted for a myriad of purpose and in the absence of other reliable sources of information the data provided by the existing regional databases will be utilised. In this sense there is some thing of an information vacuum which needs to be filled before the participants in the research process namely those workers completing database forms face any negative consequences as a result of their involvement.

Attempts to fill this vacuum, which are already under way with the creation of DMIS and review of existing data items on the DMD forms, need to b accompanied by an awareness of the importance of striking the right balance between the needs of the system and the needs of those workers completing the forms. The database designers' practice of maximising participants' involvement in the generation

and piloting of new data collection instruments has been vindicated by the validity of the data collected. This approach should be maintained while recognising the importance of the individual agency's freedom to determine the extent and method of their own data collection process. Just as important as the workability of the system of data collection is the degree of commitment that the individuals involved have towards it, which means involvement, flexibility and feedback.

.

ACKNOWLEDGEMENTS

This article is based on work which was supported by the National Addiction Centre and funded by the European Community's REITOX Project. It could not have been written without the support and assistance of Michael Donmall, Director of the University of Manchester Drug Misuse Research Unit.

Dr Tim Crabbe, Drug Misuse Research Unit, University of Manchester, UK

REFERENCES

BarrioG, Caudia B, VicenteJ, SauchezJ, De laFuenteL, Grupo de Trabajo del Seit (1993). Evaluation de la Calidad del Indicator Tratamiento de Isistema Estatal de Information sobre Toxicomanias (SEIT). Revistade Sauidade Higiene Publca 67: 385-99.

Crabbe T, Donmall M (1994) . REITOX Project - EMCDDA Preparatory Work EpAniological anl Statisticalfunction: The DrugMisuse Database Validation Study. Manchester: University of Manchester Drug Misuse Research Unit.

Davies P, Jones M, Ghodse H (1993). In confidence. Druglink September/October.

Donmall MC (1990) . Towards a National database. Druglmk 5(2): 1>12.

Dorn N, South N (1989) . Drug research and policy in Britain: A contemporary history. IntematFonalJoumal of Drug Policy 1(1): 15.

Jones M (1994). Letters. Health Service Joumal June.

Jones S, Power R (1990) . Observation to intervention: Drug trends in West London. Lntemational Joumal of Drug Policy 1(2): 13.

McDermott F, Pyett P (1992). Minimising harm in research. International Soumal of Drug Policy 3: 136.

Millar T, Donmall M (1993) . Drugs in the North West: The extent and nature of presenting problems in North Westem RHA 1992. Manchester: University of Manchester Drug Research Unit.

Mott J, Caddle D, Donmall M (1994). Letters. Druglink January/February.