The issue of appropriate research designs for studying heroin use and treatment for heroin addiction is one of the more problematic in the field. The true experiment, with random assignment of subjects to treatment versus no treatment is generally considered the surest path to valid knowledge in both medicine and social science (Campbell and Stanley, 1963, the classic statement of the virtues of experimental design). With respect to the treatment of heroin addiction, there is a generally held belief that the patients' life situations have been rapidly deteriorating just prior to beginning treatment, and thus might be expected to improve over time even if treatment is not provided.

 

Despite the apparent desirability of using randomized experiments to assess treatment for heroin addiction there are a host of difficulties in doing so. Dole and Singer (1979) have discussed these ethical, practical, and theoretical problems in detail. An illustrative case is the one study in which random assignment was attempted, (Bale et al, 1978) but patients migrated away from the assigned treatment to the treatments of their choice. Thus what had started as a true randomized assignment study became confounded with patient self-selection—acceptance of the treatments assigned.

 

In view of the great difficulties in doing true experiments in the treatment of chronic heroin addiction, it becomes important to determine whether similar substantive findings are obtained with a variety of research designs. This paper will examine substantive results from three studies of chronic heroin addicts that differed greatly with respect to research design issues. The substantive areas to be considered are drug use and deaths. The three studies to be examined are a treatment follow-up study of methadone maintenance patients in New York City (Des Jarlais et al, 1981; Dole and Joseph, 1979), a true randomized experiment of methadone treatment in Sweden (Gunne, 1981) and an oral history study of elderly narcotic users (Courtwright, Joseph and Des Jarlais, in press).

 

TECHNICAL DIFFERENCES AMONG THE STUDIES

 

The research designs are undoubtedly the greatest difference among the studies. The Swedish study used a classic experimental design with subjects who applied for treatment being randomly assigned to either methadone maintenance treatment or to no treatment. The design was maintained over time, with the no treatment group not receiving any formal treatment. Comparisons were thus made between those who did not receive treatment with those who received continuous methadone maintenance treatment. The follow-up of the New York City study used a longitudinal design. Comparisons were made of the pre-treatment, during treatment and post-treatment periods for the same subjects. The oral history study involved life history interviews with a group of elderly (fifty-five years or older) narcotic users. To the extent that comparisons were made, they were between the subjects' behavior and the reported behavior of other narcotic users known to the subjects. The three studies can easily be ranked on a scale of "methodological rigor" from true experiment to pre-experimental (Campbell and Stanley, 1963).

 

The studies varied in other important aspects. The Swedish study had a relatively small sample size—seventeen in the experiment treatment group and seventeen in the control group. The New York follow-up study, in contrast, had a sample size of 1,500. The oral history study had an intermediate sample size of 50.

 

The social context in which the subjects lived varied in the three studies. Sweden is a socially homogenous country, in which the heroin addiction problem is relatively new. The New York City follow-up study had as its context the city that has been called "the heroin capital of the world." The study covered a time period (roughly from 1966 to 1975) when the city saw the highest incidence of heroin use in its history. The oral history study was also primarily based in New York City (all subjects were recruited while living in New York) but spans the time period from the 1920s to the present. It thus covers the wide variations in public policy, treatment availability, and heroin availability that have existed in New York City over the last 60 years.

 

There are several similarities among the studies that should also be noted. First, all three studies included only subjects who had applied for treatment at least once in their lives. They did not include persons whon only experimented with heroin or who managed to successfully self-treat problems associated with heroin use. Second, all studies covered relatively long periods of time. From a minimum of two years in the Swedish study to over sixty years in the oral history study. Finally, all three studies include periods in which the subjects lived in their "natural" communities. They were not confined to residential treatment settings or hospitals for the duration of the study. These similarities among the studies lead to some expectation of similarities in the substantive findings, and also serve to distinguish these three studies from other studies using the same research designs but different subjects, shorter time periods or more restricted settings.

 

The methodological differences among the studies are still sufficiently great that different findings could be easily attributed to the differences in methods. Two types of substantive findings will be examined across the three studies—drug use and deaths.

 

DRUG USE

 

The fundamental question in assessing the effectiveness of treatment for heroin addiction is whether the declines in heroin use that are typically seen during treatment (and to a lesser extent after treatment) "caused" by treatment, or would they have occurred without treatment being provided. The Swedish study, through the use of a random assignment control group, provides the easiest to interpret answer to this question. Heroin use in the methadone maintenance treatment group declined to essentially zero. Heroin use in the untreated group remained very high—twelve subjects were using heroin at the time of follow-up, two were in prison, two had died, and only one had managed to become abstinent without formal treatment (or incarceration).

 

In the New York City follow-up study, the question of the effectiveness of treatment in reducing heroin use is answered by comparisons of three different time periods for the same subjects: prior to, during, and after treatment. Figure 1 shows the heroin use for 528 subjects in this study during the study period and for whom complete drug histories were obtained. There is a clear suppression of heroin use during treatment with a frequent return to heroin use after treatment.

 

The drop in heroin use from the prior to the during treatment periods is too large to be plausibily attributed to other factors than the provision of methadone treatment. Similarly, the return to heroin use after treatment is too large not to be associated with the cessation of treatment.

The oral history study does not directly address the question of the impact of treatment on illicit heroin use. It does, however, cover a period from the 1920s to the 1960s when there was very little treatment available for heroin addiction. Thus the heroin use of during this period can be considered an estimate of heroin use in the absence of public treatment, for this admittedly atypical group of heroin users. In describing their use-of narcotics, the oral history subjects do not mention frequent or long periods of abstinence. Instead they describe their ability to obtain an almost continuous supply of narcotics, even during periods such as World War II, when the illicit supply was probably at its lowest point during the century. These subjects had what seem to be unusually good connections for obtaining narcotics in the illicit market, as well as a well honed ability to obtain narcotic prescriptions from private physicians.

 

When discussing the specifics of their use of narcotics, these subjects mention a controlled style of use, even though they were using on at least a daily basis. They emphasize that they would use their "regular" amount, that they "were not greedy," and did go on binges where they would consume all of the narcotics that they could obtain.

 

The oral history subjects are clearly an atypical group of narcotic users, but they do provide evidence against the hypothesis that narcotic addiction will decline in the absence of formal treatment (Snow, 1973; and Waldorf, in press) for evidence in support of such a hypothesis. The inference from the oral history study is thus consistent with the Swedish and the New York City methadone studies: illicit narcotic use will continue at very high levels in the absence of formal treatment among persons who have established a chronic pattern of heroin use.

 

DEATHS

Drug related deaths may be considered the ultimate form of substance abuse. From their different methodological perspectives, the studies discussed here provide similar insights into the relationships between heroin use and death. The Swedish study found that none of the persons admitted to methadone maintenance dies within the two year follow-up period, while two of the seventeen persons who were randomly denied methadone treatment died, for a death rate of six percent per year. This difference in death rates can only be considered suggestive, however, because of the small sample size.

A major advantage of large scale longitudinal studies is that they can provide stable estimates of death rates and estimates of risk factors. Table 1 presents death rates and causes of death for the New York City study. The during treatment death rate is 1.5 percent per year, while the post-treatment rate rises to 3.5 percent per year. The major difference in the causes of deaths is the opiate related deaths. The preponderance of these opiate related deaths came shortly after the subjects had terminated from methadone maintenance treatment. They had usually detoxified as part of termination and presumably had not developed sufficient tolerance to counteract the varying quality of illicitly obtained narcotics. The size of the sample in the New York City methadone follow-up study provides statistical confirmation of the treatment lowered death rates that were suggested by the random assignment Swedish study. It also provides the most probable reason for the excess deaths in the non-treatment period.

 

Since all subjects in the oral history study were alive at the time of interview, and there has not yet been sufficient follow-up to note death rates, the only evidence from this study about drug related deaths is again indirect. In some ways this study may provide the most interesting insights into the relationships between heroin use and death. Clearly the subjects of this study lived well beyond the normal life span for heroin addicts. A major focus of the study is to determine the reasons for this unexpected longevity. The analysis of the reasons for longevity is not yet completed, but two of the reasons that the subjects give for their longevity have already been noted—their ability to obtain regular supplies of narcotics and their self-regulated use of narcotics. The two other reasons that are frequently given by the subjects are their ability to minimize their involvement in the "street life" and the scrupulous care they took to insure the cleanliness of the needles they used for injecting narcotics.

 

CONCLUSION

 

The three studies briefly discussed here had similar subject populations—persons who had become chronically addicted to narcotics. Despite the great differences in the methodologies used in the three studies, they all provide strong support for the hypotheses that maintenance treatment both greatly reduces illicit narcotic use and death rates for these persons.

 

At a deeper level of analysis, the three studies all point to the desirability, and perhaps the inevitability, of narcotic maintenance treatment once illicit narcotic use has become established within a society. If narcotic maintenance treatment is not publicly provided within such a society, there will be a small group of chronic narcotic users who managed to create their own private maintenance programs. Their narcotic use will be characterized by a relatively continuous use of narcotics, in relatively carefully regulated dosages, and with great concern for hygiene. These persons will also seek to avoid the dangerous street subculture that surrounds much of illicit narcotic use. The creation of these self-maintenance programs is testimony to both the intractibility of narcotic use among such individuals, and, given their narcotic addiction, a remarkable set of coping skills.

 

If narcotic maintenance treatment programs are publicly provided, then more suitable substances for the most suitable narcotics for maintenance can be readily utilized, hygienic conditions much more easily achieved, and alternatives to involvement in the street subculture encouraged for much greater numbers of chronic narcotic users. Clearly, not all persons who use illicit narcotics nor even all persons who have become addicted for lengthy periods of time, are suitable candidates for narcotic maintenance treatment. But for the sizable percentage of illicit narcotic users who do need maintenance treatment, the greater availability of publicly provided maintenance will be not only a matter of reducing illicit drug use and involvement in the street subcultures, but also, literally, a matter of life and death.

 

REFERENCES

 

Bale R, Van Stone WW, Englesing TJJ and Zarcone VP: Preliminary 2-year follow-up results from a randomized comparison of methadone maintenance and therapeutic communities. In: Smith DE, Anderson SM, Buxton M, Gottlieb N, Harvey W and Chung T (eds) A Multi-Cultural View of Drug Abuse. Cambridge, Mass.: Schenkman, 1978

Campbell DT and Stanley JC: Experimental and Quasi-Experimental Design for Research. Chicago: Rand-McNally, 1963

Courtwright D, Joseph H and Des Jarlais DC: Oral histories of elderly methadone patients. Journal of Oral History (in press)

Des Jarlais DC, Joseph H and Dole VP: Long term outcomes after termination from

methadone maintenance treatment. Ann NY Acad Sci 362:231-238, 1981

Dole VP and Joseph H: Long term consequences of methadone maintenance treatment. Final report to National Institute of Drug Abuse, contract 5H81DA01778-02, 1979

Dole VP and Singer B: On the evaluation of treatments for narcotic addiction. Journal of Drug Issues 205-211, Spring 1979

Gunne L: The Swedish methadone maintenance treatment experience. Drug Alc Dep 1981

Snow M: Maturing out of narcotic addiction in New York City. International Journal of the Addictions 8:921-938, 1973

Waldorf D: Natural recovery from opiate addiction: Some social psychological processes of untreated recovery. Journal of Drug Issues (in press)