ABSTRACT
Introduction
Target community
Instruments
Procedure
Target population
Validity and reliability of data
Results of the field-work research team
Induction in the community and methods for making contact with the target population: establishing contact with key informants
Sampling frame and data collection methods
Conclusions
Acknowledgements
Author: M. E. MEDINA-MORA , P. RYAN , A. ORTIZ , T. CAMPOS , A. SOLIS
Pages: 17 to 26
Creation Date: 1980/01/01
An intensive case-finding model of drug use is presented with respect to its advantages and limitations. Two approaches were used: ( a) direct, in which the questions concerning drug use were asked openly by the fieldworkers, and ( b) indirect, where the field-workers obtained information about drug use from general conversation with interviewees. The indirect approach was found to be very useful in contacting new drug users and in assessing the psycho-social context of drug use. This research suggests that the intensive case-finding method is suitable for a community where the majority of drug users form cohesive groups, while it was found to be less effective for the study of isolated cases.
Despite the considerable advantages derived through the application of the epidemiological model to the study of drug use, the issue of case identification continues to present problems. These difficulties occur partly because of the secrecy in which drug use is conducted, thus increasing the problem of contacting cases. Another difficulty arises from the mobility of the drug-using population, which increases the problem of applying traditional epidemiological survey techniques in fixed stable locations such as households, schools or treatment centres.
Studies using traditional epidemiological techniques in households and schools require an institutional infrastructure in terms of the availability of statistical data, trained personnel and data processing facilities which is often not available in developing nations. This consideration illustrates the importance of creating new epidemiological methods more easily applicable in developing countries. The kind of method that would appear to be necessary would be one that would allow an appreciation of the overall prevalence of drug use in a given community, particularly by those drug users who are homeless, not attending school or without a stable occupation.
Also, the method should, if possible, be capable of minimizing low response rates, and of assessing the psycho-social context of drug use. Such a method would also allow the possibility of assessing far more accurately the treatment needs of the studied population.
The development of such an ambitious project requires a detailed test of techniques and methods for its application in the community.
In 1971, Hughes and others [ 1] proposed a method of case identification with the above characteristics, to be used for the detection of cases of heroin addiction. The copping zone (place where heroin is bought, sold or used) was used in that study as the epidemiological unit for case detection. It was possible to use this approach since the copping zones were highly stable in time and were frequently visited by a majority of the addicts. Former drug users were used as epidemiological field team workers.
Clearly, the application of this method is based on the assumption that a sufficiently stable and sufficiently frequently visited copping zone is available for use as the epidemiological unit of observation and that informants can be located who are well acquainted with the target community and accepted by it.
An alternative method of sampling the user population when the copping zone approach is inapplicable has been developed by de Alarc6n [ 2] , who found that the snowball approach was an effective method for locating cases. So far as staffing was concerned, newly graduated professional psychologists were employed, and techniques were developed to contact the drug subculture.
Thus the focus for the present study was the review of methods of initial induction in the community, of initiating contact with the sampled group for drug users, and of gaining demographic, epidemiological and psycho-social data on the drug user group.
In this article, the intensive case-finding model will be evaluated in four main respects:
The field-work research team
The initial phase of induction to the community and the methods for making contact with the target population of drug users
The sampling frame with respect to the relative utility of the copping zone and snowball methods
Data collection methods, including the use of both participant observation and directed interviews
The zone selected for study was a suburb in the south-western area of Mexico City of low socio-economic level, with a total estimated population of 70,000, of which 65 per cent were aged 25 years or less.
Basic demographic and epidemiological data were collected by means of the WHO Short case Reporting Form. 1 Qualitative data concerning the psycho-social context were collected by means of a semi-structured questionnaire. Both instruments were used as guidelines and were filed after obtaining information in the community.
Two field-workers were hired, both graduates in psychology with previous experience of community work. 2 Professional staff had to be hired as no addicts in treatment or ex-addicts were available at the beginning. No street addict was found to act as field-worker, mainly because the drug users were well organized in small groups and no user was found who interacted with most groups of users in the community. The same was true for drug dealers. Drug supplies for a group of users were usually obtained by a group member who would buy enough for the whole group. Dealers were not looking for their customers and were, therefore, not acquainted with the large majority of users. The only resource available therefore was professionals who could be introduced to the target group by informers in the community.
In order to gain access to drug users, the first stage of the project was to establish contact with those agencies, institutions and individuals in the target community, which because of their official role and specialized knowledge were likely to have contact with the target group and were known and trusted by them.
Informants were asked to introduce the field-work team to drug users and to specify copping zones. The informants were classified as qualified or non-qualified according to the type of contact they had with the community. Following this classification, informants such as priests, school teachers, health workers, or shop owners were considered as qualified informers, whereas children and housewives were not included in this category.
In order to evaluate the screening ability of informers in the community, three different criteria were followed: (a) number and time of interviews held with each informant in order to gain their confidence, (b) number of drug users, and (c) comparison of the information given by informers with that obtained by field-researchers.
1 This form was developed by the WHO Research and Reporting Project on the Epidemiology of Drug Dependence in an effort to encourage comparability in drug use surveys of youth and can be obtained from the Division of Mental Health, WHO, CH-1211 Geneva 27, Switzerland.
2They were the co-authors A. Ortiz and T. Campos.
The second stage consisted of establishing contact with the target population. Two different sampling methods were tested: a snowball method and a "meeting-places method" in order to gain access to all drug users in the community. The first method required that each contacted drug user himself refer the field research team to all other drug users known in his particular social network. Therefore each member of the target population had a double function; he was both a subject of investigation and a source of referral [ 2] . The second method was based upon observing specific and stable meeting places, with definable geographical boundaries where drugs were consistently bought, sold and consumed [ 1] .
The third stage consisted of obtaining basic epidemiological data about the use of drugs and the evaluation of the psycho-social context in which drug use was taking place. Two different approaches, one direct and the other indirect, were used:
Direct approach, or structured interview. In this approach the field-workers presented themselves as researchers studying tile phenomenon of drug use in the community. Questions about the use of drugs were open and direct, and their help in gaining contact with other drug users was solicited;
Indirect approach, or participant observation. In this approach the field-workers also presented themselves as researchers but they manifested a more general interest in the problems of the community. The drug-use information was collected as one of the items of interest. This theme was introduced in the general conversation and, as a result, the speed of data collection was slow. The approach followed was that of participant observation in which the field-workers interacted with the drug users as members of the group. Contact with other users was made in the course of normal social interaction with the group.
Of the 123 drug users identified in a period of five months, 62 were studied. Only one of these was female, a former user who did not belong to any group. The ages of the users ranged from 14 to 35 with a mean of 23.3. The largest proportion (42 per cent) was 20-24 years of age. Most of them, 82 per cent, were single; 11 per cent were married and 6.5 per cent, separated. Their level of educational attainment was low: 6.5 per cent had never attended school, 35 per cent had spent 1-6 years at school; the mean number of years of completed schooling was 7.7 years. Only 11 cases were students or trainees, 27 per cent had been unemployed in the previous 30 days, 41 per cent had been employed fully or part-time, but jobs in most cases were neither permanent nor secure and were short-term in nature.
Alcohol and cannabis were the most frequently used drugs: 90 per cent were current users of alcohol and 81 per cent of cannabis. Inhalants were used by 31 per cent of the subjects, but only 14.5 per cent had used them during the previous month. Other drugs, such as hallucinogens and amphetamines, were used by only a small proportion of users. Most of them were long-term users with a mean of 7.4 years of use of drugs other than alcohol.
Whenever possible, direct corroboratory evidence was sought, i.e. each individual drug user was interviewed not only once, but on several occasions with the hope of observing direct evidence of drug use during the course of the interview. Behaviour such as open smoking of marijuana or inhaling of solvents or inviting others to smoke or inhale was taken as direct evidence of a current habit of drug consumption. Direct evidence of this nature was available in 23 of the 62 interviewed cases. Corroboratory evidence from friends of the interviewee was obtained in an additional 20 cases. In 19 cases there was only the self-reported evidence of those interviewed.
The field-researchers turned out to be very efficient. It would seem that forming a staff of professionals is a good way to get information in a context where no agency or person is working with a drug-user group or no former addicts or addicts in treatment are available to become members of the staff.
However, it should be emphasized that simply to be a professional is not in itself a sufficient qualification. The following additional skills and abilities are essential: the ability to interact as part of another subculture and the willingness to expose oneself to certain dangers and work in a flexible schedule which includes nights and weekends.
Age and sex were also important: young field-researchers could interact better with drug users of the same age. Male and female staff interacted differently with drug users. In the subculture of the target population, a female was found to come up against certain barriers, i.e. there were certain topics which were not mentioned in front of a woman. She was also protected by the group: consequently the group became jealous if the female field-worker tried to make contact with other groups. In this sense, the mobility between different groups may be more restricted for the female field-workers. This difference suggests that in case a decision has to be made, male field-workers are a better choice. No final recommendation can be made in relation to the usefulness of working together or individually in the community. There are certain circumstances where more than one researcher is seen as threatening, but on other occasions working as a team provides security. In any case, having more than one researcher in the field provides the opportunity of confirming information.
Another important finding is that more than one pair of field-workers are necessary to interact with rival groups of users.
A total of 66 informants were interviewed in the first phase. Forty informers were interviewed the first month of field-work and the other 26 were interviewed contemporaneously until contact with drug users was established.
A total of 10 initial contacts were made by means of key informants. Nine of these stemmed from qualified informants and only one from those who were non-qualified.
There were considerable variations both in the number and duration of interviews with informants. Similarly, there were considerable differences in the time which elapsed prior to making direct contact with the users. The closeness of the friendship the key informant had formed with the young drug user and the frequency of their meeting were particularly important factors in determining the speed with which contact was made.
Twenty-nine per cent of the places referred to by the informants were confirmed by the field-researchers. A further 29 per cent referred to by informants were not confirmed by the field-research team. Around 30 per cent of places detected by the field-research team were not mentioned by the informants. Tiffs does not imply that the meeting places not confirmed by the field-research team were not in regular use; it could be that they were used, but on an unpredictable and irregular basis.
Eighty per cent of the places referred to by the informants (both qualified and non-qualified) were open. The more hidden locations away from the main streets were known only by the users themselves.
When the informants were interviewed, they reported that pre-adolescents most frequently used inhalants, and adolescents marijuana. That corresponded with tire experience of the field-research team so far as adolescents were concerned. No cases of pre-adolescent drug use were discovered.
The main results of this phase suggest that rather than any type of informant being particularly useful or effective, what was of chief importance was to locate particular individuals who had a special knowledge or experience of the target group.
The best policy seemed to be to contact as wide a variety of informants as possible in the initial stages so as to maximize the chances of contacting a key informant who could introduce the field-work team directly to the target population. There was also no sure way of predicting which informant group was likely to prove the most effective. This was likely to vary with the particular community concerned.
It would seem that all key informants were very accurate in their identification of drug users. No false positives were detected by the field-research team. It is not possible to make any definite conclusion concerning the 61 identified but non-interviewed cases. It seems probable that the majority were likely to be users; the number of false positives was unlikely to be lfigh. However, it would be necessary to take on more field-research staff in order to find out.
Finally, it is worth adding that the non-qualified informants were particularly useful in the early stages of the study in orientating the field-research team toward the special needs and characteristics of the target community.
As the usefulness of the sampling methods is related to the method of inquiry, they will be presented in conjunction.
A total of 160 interviews were conducted in order to gather basic epidemiological data on the 62 drug users. The mean number of interviews per user was 2.58 for the indirect approach and of 2.0 for the direct approach. A more important difference was that in the indirect approach a fairly large amount of time had to be spent with users before the topic of drug use could be addressed. Also, in the direct approach no information about the socio-cultural context could be obtained.
The direct approach led to only nine cases while the indirect approach resulted in 53 cases in five months of field-research.
In both approaches the presence of the "introducer" was of vital importance as the only way of entering into contact with other drug users known by the first contact and to gather information. In some cases this introducer became one of the key informants; in other cases, it was one of the drug users who introduced the field-workers after checking them on several occasions and after a period of interacting with him.
A total of 50 cases were contacted by means of the snowball method. The initial contacts were with four young drug users. These led to 45 detected cases through the indirect approach and 5 by the direct approach. Four snowball chains were followed, consisting of 12, 14, 6 and 13 cases. The four initial cases were contacted by means of key informants. The remaining 41 were detected with the help of introducers, drug users who had already been contacted and who were prepared to introduce the field-work team to further users. A further 8 cases did not lead to further contacts and so were not part of the snowball chain.
In only one instance, using the direct approach, was the interviewer presented to new users (by means of the snowball method). It is perhaps significant that all five were relatively young in terms of the average age of the target group (17 as opposed to 23.3 years); they were still attending secondary school and some were also working. They were not regular, chronic users, and none had had problems with the police. Perhaps they had less to hide than the older, chronic users and so felt less reluctant to pass on information about their friends.
Each group of drug users had its own "territory" within the community. Within its own territory, there were two or three meeting places the group regularly used. It never met with other groups in these places. There was no regular or specific time when the users would gather at a particular meeting place. Consequently, it was never possible to be sure when and where to meet the group. Certainly, the strategy of simply remaining in one meeting place until the group turned up would have resulted in a considerable waste of time.
As the results indicate, it proved much more effective to apply the snowball method with the indirect approach. There were two main reasons why the indirect approach was successful in facilitating further contacts. First, the field-research team were prepared to spend many hours, not only during the day but also at night and weekends, in getting to know the contacted group and gradually being accepted and trusted by them. Secondly, spending a lot of time with the group was also essential to make it possible to make new contacts. It was impossible to guess when the group might meet with other users. Consequently it was essential to be with the group as much as possible to maximize the chance of being on the spot when the group met with other users. It was also important that the users become familiar with the field-workers and accustomed to seeing them around.
In conclusion, we may say that the intensive case-finding model can be successfully applied to the study of drug use in a community if the majority of drug users form cohesive groups. The study of isolated cases would be more difficult using this method.
The applicability of this method is not limited by availability of informants such as drug users in treatment, ex-dealers or users who are fully conversant with the local drug-taking subculture. Professional workers, such as young graduate psychologists, can provide an excellent alternative. The informants perform two useful functions: first, introducing the field-work research team to the norms and values of the local community and drug subcultures; second, introducing the field-work team directly to the drug-user groups and facilitating the development of trust between the field-workers and the groups. However, it is not possible to conclude that one particular type of informant is necessarily more useful or effective than any other.
The availability of informants is particularly important in communities where no stable, frequently used copping zones exist to provide a suitable focus for observation. In these conditions, the snowball method proves to be extremely effective as a method of contacting drug users. This is particularly the case when an introducer is available who is trusted by the user groups; this facilitates a rapid extension of the contact chain. The presence of an introducer is important to maximize the probability that the users are prepared to serve as referral sources for other drug users in their social network, thus establishing the chain of contacts essential for the functioning of the snowball method.
Nevertheless, it needs to be emphasized that the presence of informants is not essential for the application of the intensive case-finding method. Leal, for example, was able to establish direct contact with a group of inhalant users [ 3] . However, in this case there was a stable, frequently used copping zone by means of which the users could be fairly easily detected. The existence of the zone was primarily determined by the distribution network of the preferred drug, which could only be purchased in a certain place.
In terms of specific methods of data collection, it would seem that the direct approach method, despite its advantages in terms of time and cost, has a major limitation in terms of the difficulty of making additional contacts.
The indirect approach method proved to be extremely useful, both in terms of contacting new users naturally in the course of the observation process and also in checking data reliability and in evaluating the psycho-social context of drug use.
Thus it can be said that the intensive case-finding model has proved a very valuable means of overcoming some of the limitations of traditional epidemiological methods, such as household and school surveys, especially in terms of giving more sampling coverage to "floating" populations, those with no fixed location in households, schools or work. The model also minimizes problems of low response or under-response by means of allowing the interviewee to develop a relationship of trust with the interviewer prior to giving information. It also allows the longitudinal study of changes in drag consumption patterns and in psycho-social functioning. It provides a fund of valuable information on which to base an appropriate and effective treatment programme. An additional advantage is that it does not require highly sophisticated statistical and computer processing (facilities which are not always easily available in developing countries).
On the other hand, the intensive case-finding model does possess certain limitations. For example, the process of entering into contact with key informants and then with drug-user groups can prove time-consuming and relatively expensive. A possible way of overcoming this particular disadvantage would be to choose smaller communities. This would also have the advantage of aiding the evaluation of overall prevalence and increasing the probability of contacting all users.
It is emphasized that the population contacted by means of the intensive case-finding method were chronic marijuana users with an average age of 23.3 years who typically met in groups.
Although some of the informants contacted by the field-work team mentioned that the community also contained young pre-adolescent inhalant users, it did not prove possible to contact them. It is difficult to say whether there simply were no other type of users other than those actually contacted, or whether the intensive case-finding method lends itself to the detection of chronic users while excluding or ignoring others less known by the community.
This latter hypothesis is supported by data drawn from a household and school survey conducted in the same community [ 4] . It was discovered in both these populations that there were a greater number of occasional, experimental and former users than regular users.
It is possible that no inhalant users were detected by this method because their style of drug consumption is more secretive and thus less known by the community. If inhalant or other types of users tend to conduct their drug habits alone rather than in groups, it would be another factor that could limit the applicability of the intensive case-finding method. The same factor may account for the fact that only one female user was detected; perhaps female users have more secretive drug-taking habits and are thus less easily detected. Chavez and Soils, in a previous study in the same community [ 5] , reported that the drug-taking behaviour of female users was indeed more secretive, and that only the closest friends of the users knew of their habit.
An additional factor is that in numerical terms there probably are fewer female drug consumers; a number of previous studies of households [ 6] , schools [ 7] and communities [ 5] , all found a greater prevalence of drug use among males than females.
It would therefore seem necessary to introduce modifications in the intensive case-finding method so as to include those populations that the method, as currently constituted, tends to omit (female drug users, experimental or occasional users etc.). One possible area of modification lies-in the kind of informant used; it would be particularly important to develop closer contact with children. Increasing the amount of field-work time spent in the community and increasing the size of the field-work team would maximize its probability of detecting more secretive patterns of drug use.
Finally, given the limitations of time of the intensive case-finding method, it would seem advisable to devise alternative case-finding methods that, although they might be less effective in terms of gathering a rich store of longitudinal data, might be more effective in quickly and cheaply determining the existence of a drug-taking problem in a given community.
This paper is based on the results of a collaborative study of the WHO Research and Reporting Project on the Epidemiology of Drug Dependence which has been co-ordinated by the Division of Mental Health at the WHO headquarters in Geneva (MNH Medium Term Project 01/02/01).
Acknowledgement is given to the World Health Organization for its support, to Dr. Patrick Hughes, the Project Manager, and to Dr, G. Jarvis for their comments on the methodological aspects of this study.
This project was undertaken under the supervision of Dr. Carlos CampilloSerrano, Head of the Social Sciences Department of the Centro Mexicano de Estudios en Salud Mental CEMESAM).
P. Hughes and others, "The social structure of a heroin copping community", American Journal of Psychiatry, vol. 128, No. 5 (1971).
002R. de Alarcón, "The spread of heroin abuse in a community", Bulletin on Narcotics (United Nations publication), vol. 21, No. 3 (1969), pp. 17-22.
003H. Leal and others, "Estudio naturalistico sobre el fenómeno del consumo de inhalantes en niños de la ciudad de México", in Inhalación voluntaria de disolventes industriales (Mexico, Ed. Trillas, 1977), pp. 442-459.
004M. E. Medina-Mora, M.E. Castro and G. Terroba, "Drug use among youth population in Mexico", paper presented at the WHO Meeting of Collaborative Investigators, Research and Reporting Project on the Epidemiology of Drug Dependence (Penang, University SAINS Malaysia, 1979).
005M.I. Chávez, A. Solís and others, Drogas y pobreza (Mexico, Ed. Trillas, 1977).
006M. E. Medina-Mora, "Prevalencia del consumo de drogas en algunas ciudades de la República Mexicana; Encuestas de hogares", Enseñanza e Investigación en Psicologia, vol. 4, No. 1 (1978), p. 7.
007M. E. Castro and M. Valencia, "Consumo de drogas en México. Patrones de use en la poblacion escolar", Salud Pública de México, vol. 20, 1978, pp. 585-590.